THE UNIVERSITY OF ILLINOIS LIBRARY 650-7 W75re fto- 51 -65 Cop. 2 AGRICULTURE Digitized by the Internet Archive in 2016 https://archive.org/details/ruralprimarygrou5165kolb G?30 .1 \4 1 S 85 28 105 31 '] 56 13 95 39 i 52 8 66 20 5 39 6 71 17 ) and over 86 17 121 26 Total 76 18 137 40 0 15 3 28 11 1 14 3 14 5 2 15 3 20 2 3 9 2 17 10 4 11 5 17 5 5 5 1 16 1 8 and over 7 1 25 6 Total 169 50 267 90 0 22 9 33 21 1 13 10 36 [ 22 2 44 13 60 i 14 > 28 1 6 44 13 1 13 1 2 18 7 5 13 3 35 5 :> and over 36 7 41 8 Total... 117 38 162 76 14 6 24 19 l 12 7 21 17 2 15 10 19 11 3 15 4 27 10 4 16 1 21 5 5 15 2 12 5 > and over 30 ' 8 38 9 Tctai ' 1 26 10 32 21 0 1 2 3 4 5 6 and ovej 1 8 3 4 4 1 2 1 0 0 1 4 4 5 3 5 * . 3 8 3 3 3 4 2 3 3 40 Wisconsin Research Bulletin si Table IX — Continued Total 28 3 32 10 0 3 3 6 1 240 and over 1 3 0 2 2 2 3 0 1 1 3 1 0 4 2 4 7 0 5 1 5 2 0 5 3 6 and over 9 0 9 0 Table X. — Size of Farms of Owners and Tenants In Neighborhood and Non-Neighborhood Areas Occupancy I Size of farm in acres j Percentage and number of neighbor- hood and non-neighborhood farms Per cent Nur nber Neigh- borhood Non- neigh- borhood Neigh- borhood Non- neigh- borhood Total 100 100 535 867 Under 59 17 20 94 177 Both 60 - 119 41 41 219 357 120 - 179 29 28 155 238 180 - 239 7 6 36 53 240 and over 6 5 31 42 Total- — 100 100 416 630 Owners Under 59 18 22 76 137 60 - 119 41 42 169 267 120 - 179 28 26 117 162 180 - 239 6 5 26 32 240 and over 7 5 28 32 Total 100 100 119 237 Tenant Under 59 15 17 18 40 60 - 119 42 38 50 90 120 - 179 32 32 38 76 180 - 239 8 9 10 21 240 and over 5 4 3 10 striking contrasts, however, which would indicate either size of farm or occupancy as being decidedly characteristic of either sort of grouping. The educational purpose or function is very high in point of frequency in both the eastern and western parts of the county, Rural Primary Groups 41 having 31 primaries and 8 secondaries. The reason for this should already be clear from the discussion of the factors, which have tended to create and to maintain the rural groups. This educational function is one which has persisted longest in the changing processes of the smaller and earlier groups. There were some over twenty-five groups which show this as their sole function with the exception of an incidental social activity in connection with the school. The kinship consciousness factor exerted its power as a com- pelling purpose to a greater extent in an early day than at pres- Table XI. — Number and Size of Families Among Owners and Tenants In Neighborhood and Non-Neighborhood Areas Grouping Number of children per family Number and size of owner and tenant families Number of families Total number of children Owners Tenants Owners Tenants Total 1,046 356 3,534 915 0 150 79 0 0 1 120 72 120 92 Both 2 190 59 380 118 3 151 52 453 156 4 118 28 482 112 5 110 23 549 115 6 and over 207 43 1,550 322 Total 416 119 1,435 330 0 55 24 0 0 1 43 23 43 43 Neighborhood 2 85 28 170 56 3 56 13 168 39 4 52 8 218 32 5 39 6 194 30 6 and over 86 17 642 130 Total 630 237 2,099 585 0 95 55 0 0 Non-neighborhood 1 77 49 77 49 2 105 31 210 62 3 95 39 285 117 4 66 20 264 80 ,■) 71 17 355 85 6 and over 121 26 908 192 42 Wisconsin Research Bulletin 51 ent, yet as a possible third factor in the group life of today, it is worthy of consideration. By further analysis of the data assembled in Table IX an- other table was drawn, number XI, in order to set out the size of the families among owners and tenants in neighborhood and non-neighborhood areas. There appears to be a slight tendency for more children in the neighborhood than the non-neighbor- hood areas but the difference is not significant in size. As one would expect the tenant family is uniformly smaller. Both tenant and owners show heavy numbers in the two child per family class. In comparing the size of family and size of farm, there appears to be larger families of both tenants and own- ers when the extremes in size of farms are eliminated ; that is to say, there are larger families on the farms of the two classes, 60 to 119 acres and 120 to 179 acres. Table XII. — Summary Table Showing Average Children Per Family of Owners and Tenants In Neighborhood and Non-Neighborhood Areas Occupancy Average number ot children per family Non- Neigh- neigh- borhood borhood Total average - 3.3 3.1 Owners - — — _ _ Tenants - -- 3.4 3.3 2.8 2.5 Another trial with reference to the family was made in the summary table, number XII, showing the average children per family of owners and tenants in neighborhood and non-neighbor- hood areas. The comparison for owners and tenants is in favor of the owner, 3.4 compared with 2.8 in neighborhood and 3.3 compared with 2.5 in non-neighborhood. For both owner and tenant the neighborhood has a slight advantage, 3.3 as compared with 3.1 for the non-neighborhood. To get more directly at the kinship influence within the neigh- borhood itself, the following question was asked on the “family question card” in western Dane county: Rural Primary Groups 43 “To how many families within a distance of five miles are you or your family related by blood or marriage as near or nearer than second cousins ?” The results are shown in Table XIII. As soon as the classes of related families pass into and beyond that of 8 to 11 families, the number within the classes shows a rather decided increase in favor of the neighborhood as compared to the non-neighbor- hood area. In each of these classes extending beyond but in- cluding the 8 to 11 families approximately a third more is found in the neighborhood section. Table XIII. — Family Relationships as Near or Nearer Than Second Cousins Within Five Mile Radius of the Home of Neigh- borhood and Non-Neighborhood Families Number and percentage of neighborhood and non-neighborhood families Number of related families Per cent Number Neighbor- hood i Non-neigh- ! borhood Neighbor- hood Non-neigh- borhood Total 100 100 281 357 Less than 4 _ 33.4 44.3 94 158 4-7 28.0 30.1 79 108 8-11 18.4 12.0 52 43 12-15 9.2 6.4 26 23 16-19 3.6 2.5 10 9 20 and over 7.1 4.5 20 16 The kinship function is clearly predominant in at least one neighborhood in the east and two in the west. This cannot be construed to indicate that the people set about deliberately and consciously to promote family consciousness to the exclu- sion of all else, but there was evidence that its role is a central one in social affairs, group leadership and marriages. Chart I is designed to show this condition in the eastern neighborhood called Betlach-Dushack. xl This shows eight inter-related gen- eral families and 37 units of these families within the neighbor- hood and 30 units without the bound of this group but in nearby groups or villages. 11 The information regarding the relationships was secured from Mrs. Effa Dushack of Sun Prairie and the number of families in relation to the area was secured from the Rural Directory of Dane County. 44 Wisconsin Research Bulletin 51 CHART I RELATIONSHIPS IN THE BETLACH-DUSHACK NEIGHBOR- HOOD Family name with lines of relationships Total Number of families within neighborhoods and those nearby. 11 Rural Primary Groups 45 The place of the local government as a function of group life was found to be a minus quantity. This is not to say that local government is not a function of the rural group but no group discovered had this as its primary or secondary purpose or interest. The nearest approach to it was in Springdale town- ship where a very fine town hall is equipped with all social facilities as dining-room, auditorium and kitchen. There is a township community club, even, but all this is organized about the social rather than the governmental function. Nationality solidarity as a group function is not strong in the primary sense yet in the secondary it ranks highest. This can- not indicate that the group deliberately sets about to achieve solidarity in this particular aspect but its influence is there and many of these groups are distinctive because of this influence. Its lines are observed in marriage, in religion and in language. The religious purpose ranks high as a primary and low as a secondary, having exchanged places, as was, pointed out above, with the nationality factor when its original and present in- fluences are compared. The social function shows itself to be a strong factor in the secondary sense for its characteristic is that of a supplementer to other purposes considered of initial importance. In the non-functional classification appear the 26 groups which mustered in before under the “lack of factor” classification. Structure. The structural relationships are summarized in Table XIV. Discussions of each will be omitted here to be taken up again under “Forms,” when maps will be used to show comparative relationships. The structures when sum- marized in the order followed for the functions above are as follows : Economic association, legal-education organization, kinship ties, legal-political organization, nationality bonds, church congregation, social clubs, and remainder falling into the decadent classification. These structures from the stand- point of the numerical tabulations of those primary and those secondary follow rather closely the corresponding functions. This is what one would expect, though now and then devia- tions are to be found — when, for example, the nationality bonds are made to do service for getting an educational or a social purpose accepted. On the other hand there may be a 46 Wisconsin Research Bulletin 51 function without the structure to make it effective, or again, a structure of a certain character may make the function im- possible of attainment. Table XIY. — Groups Classified on Basis of Structure Structure of groups, ; primary and . secondary Eeo- | Legal- Legal- nomic educ. Kinship political Nation- Church Social Deca- Division of county associ- ation i organi- zation ties organi- zation ality bonds congre- gation club 1 dent Pri. Sec. Pri. Sec. Pri. Sec Pri Sec. Pri Sec. Pri. Sec. Pri. Sec. Total 1-; 5 35 2 2 2 0 0 2 32 32 3 10 17 26 Eastern 1 10 2 14 2 1 0 0 0 0 14 14 1 5 7 I 12 Western i 4 3 ! 21 c! 1 2 0 0 2 18 18 2 5 10 14 Group Institutions. Institutions should be considered at this point for in a real sense an institution is only the more or less fixed and rigid form of the social structure. Table XV sets forth in a classified arrangement, the important local social institutions of the rural groups. The school leads with 112, cemeteries next with 57, church and store tie for third place with 47 each. Farmers’ organizations are low, although many, but not all, of the dairy manufacturing plants owned and operated by a farmers’ co-operative association, are in- cluded. The open county postoffice now rendered useless by the mail delivery service was formerly an institution of rather unusual importance. It was a typical group institution as is shown by the fact that it was present in nearly one-third of the neighborhoods. Some one is sure to object that a cemetery is not a social institution. Very many of them, however, represent an active local cemetery association organized as an endowed corpora- tion with trustees, for the purpose of keeping alive the memor- ies of dead neighbors and relatives, as well as for making beautiful their places of burial. Such associations are real social structures and they are discharging a worthy function in many localities. When riding along the country road and a cemetery and church or school, or perhaps both, appear on Rural Primary Groups 4 7 Table XV. — Institutions of the Group Institution Total Eastern Western Total 356 158 198 Cemetery 1 57 • 28 29 Church - - - - 47 22 25 Creamery or cheese factory . _ __ 39 13 26 Farmers’ organization . 10 5 5 Post office (active) __ __ __ - 6 2 4 Post office (discontinued) 28 11 17 School (district) 112 52 60 School (parochial) . 6 1 5 Social center 4 0 4 Store or shop 47 24 23 the horizon, one may be sure of preliminary evidence of the existence of a group, if not at the present time then in an earlier day. In one neighborhood where such an association is active, the teacher of the district school, a half mile away, takes her children once a year to this place of memory. She points out names which are still household names in that neighborhood and tells over again the story of the early life of those settlers who forged out of the wilderness the civiliza- tion to which these children have now fallen heir. Form. — In this section effort will be made by use of the maps to show the form or shape which these primary groups have assumed. First the tabular lists will be introduced to show totals for the county. The table is number XVI and carries the pri- mary and secondary classification as did the tables for function and structure above. It is not to be expected that even a primary classification of a group as having the form of a school district, for example, means that the two are exactly coextensive but that the group tends more nearly to approximate this area than other comparable area. This becomes then, in a sense the objective measure of the structure of the group. The first, in order followed in the sections above, is the area corresponding to economic activity of the group and labeled the economic or trade area. The term “trade area” may be used here a little inaccurately since it has come to mean the area about a village trading cen- ter. Here, its use means the geographic extent of the eco- nomic activity of the group, be it retail grocery trade or a creamery or cheese factory patronage area. The summary 48 Wisconsin Research Bulletin 51 tabulation, table XVI, shows 12 groups following primarily their economic area in extent. This is a somewhat smaller number than are reported under the “Economic” classification for structure and function which seems due to the fact that the economic has not worked out or in some cases has lost its definite boundary for the strictly open country groups. De- scribed in another way, it means that the economic area has been of secondary significance or has fitted into other recog- nized areas such as church parish or school district. Figure 10 of eastern Dane county shows a regular village trade area map superimposed on the primary group map. These boun- daries were made from a study in 1917 and in some cases the lines had been smoothed in order to eliminate overlappings and to exclude certain small center areas. 12 Table XVI. — Groups Classified on Basis of Form Group forms, primary and secondary Division of county Trade area School district Family circle Town- ship Nation- ality settle- ment Church parish Social area With- out form Pri. Sec. Pri. Sec. Pri. Sec. Pri. Sec. Pri. Sec. Pri. Sec. Pri. Sec. Total 12 9 39 4 2 1 0 4 2 25 33 2 7 19 26 Eastern 7 6 19 4 1 0 0 0 0 11 15 0 2 7 12 Western 5 3 20 0 1 1 0 4 2 14 18 2 5 12 14 The only area recognized by this map corresponding to any of the primary groups under consideration is that of Deans- ville. Here the area is much larger than is shown on the base map. The conclusion one comes to as this whole map is observed is that the village or city trade area boundary is no respector of primary group lines. A trade line cuts through Norway Grove, through West Koshkonong, through East Bristol and so on. On the other hand, this disregard of group lines is more apparent than real. To show this point more directly Figure 11 for the western part of the country is intro- duced. It shows the results of the questions regarding trad- 12 Galpin, C. J., and James, J. A. Rural Relations of High Schools. Agr. Exp. Sta. Bulletin, University of Wisconsin, No. 228 (1918), p. 11. Rural Primary Groups 49 NEIGHBORHOOD TRADE AREA FIG. 10.— A VILLAGE TRADE AREA MAP These boundaries were made from a study in 1917. In some cases the lines had been smoothed to eliminate overlapping's and to exclude certain small cen- ter areas. 50 Wisconsin Research Bulletin 51 NEIGHBORHOOD TRADE AREA Neighborhood Boundaries ,1 , / Township tu tnu 71 ) tun FIG. 11. — trading areas of the west The trade lines do not always respect group lines, but in the cases of cutting over there is usually an area of overlapping. Rural Primary Groups 51 in g center which were asked on the “Family Question Card.” The trade areas are shown in red. The overlapping and the inclusion of one area within another are shown by systems of cross hatching and hatching, respectively. The trade lines do not always respect group lines, of course, but in many in- stances where they do cut over there is also an area of over- lapping. For example, Union Valley, Oak Hall, East Middle- ton, Acorn, and Mud Lake are typical of this tendency. There are exceptions to this, for Primrose is cut by the New Glarus boundary, Pine Bluff by Mt. Horeb, and Frenchtown by Belle- ville. Of the exceptions the first indicates the Swiss influence which is rapidly moving north across Primrose township. The second shows a local center’s former trade area abandoned and now divided between two larger centers though the other bonds of the group are still holding fast. The last instance of Belleville shows an interesting competitive game between very active farmers’ stores at Basco and Paoli and those of the larger center, Belleville. The latter has the decided advantage because of the new and improved roads, but the smaller centers are surprisingly vigorous in their command of the farmers’ trade. A number of smaller centers, not shown on the smoothed map, appear here as inclusions or as in the midst of overlappings, as, for example, Riley, Klevenville, Fitchburg, Marxville, Roxbury, and Springfield Corners. A farmers’ organization area map is presented under Figure 12, which carries the plottings of the answers to the question, “Where do you go for your farm organization meetings such as Equity, Farmers’ Clubs, or Shipping Associations, etc.?” This was used only in the western part of the county. The reports were rather scattering indicating no great strength of farmers’ organization, especially when considered from the standpoint of the local neighborhood units. In general the village head- quarters are the same as those for the trade areas, the high school areas and even to some extent the church areas. This would seem to point to the focussing of certain of the farmer neighborhood interests in the nearby villages. The next is the school district area. This comes most nearly approximating the primary group area for the tabulation shows 39 primary and 4 secondary classifications. When turning to the map comparison, Figure 13, one is not impressed at once with 52 Wisconsin Research Bulletin 51 NEIGHBORHOOD FARMERS’ ORGANIZATION MAP FIG. 12.— FARMERS’ ORGANIZATION AREA MAP Reports on this question indicate no great strength of farmers’ organiza- tions especially from the standpoint of the local neighborhood units. Rural Primary Groups 53 the coincidence of lines, for there appears to be a good deal of cutting across the group lines. This is true, particularly of the larger groups, where some other factor has been determining as for example, the religious functions in the case of the parish boundary. In other instances the central part of the group and of the district seem to- be the same, but their boundaries are different in certain details. One explanation is that those liv- ing near the edge of the neighborhood were less conscious of the grouping, especially in the case of the newer comers or renters and also in the cases where the group itself was slipping toward the decadent class. A number of the neighborhoods and the district correspond almost exactly, however, and there the group is the district and vice versa. Such cases are Gaston, Tippel, Ritchie, Blue Valley, Malone Valley, Hanerville, and Stone. For a comparison of the high school areas in relation to neighborhood groups Figure 14 is drawn. 13 These areas, it should be noted, are simply areas of influence or of patronage and not the legal high school area itself. It has been calculated that seven- eighths of Wisconsin’s area lies outside legal high school districts. 14 It is doubtful whether Dane county is an exception to this pro- portion. The maps now under observation clearly show that the rural neighborhoods in the strict sense of the term, do not have high schools of their own. The farmers’ children attend village and city schools by the permission and tuition system. The farm- ers do not have a voice in the management of these institutions. The family circle is too indefinite a unit to be of special value here. It is carried over to keep the classification intact. The two cases tabulated under the heading are Table Bluff, or Kahl Hill as it is sometimes called, and the Betlach-Dushack group. The tozvnship as a geographic unit receives some recognition in this measurement plan. There are four groups which tend to approximate the townships in which they exist, although for no functional or structural reasons. These are Ashton, Primrose, Roxbury and Springdale. All are essentially church parishes. Some tendency was noted among certain of the church bodies, especially the Lutheran and Catholic groups, to speak of their parishes in terms of the township area and name. 13 Galpin and James, op. cit., p. 9. 14 Ibid, p. 7. 54 Wisconsin Research Bulletin 51 NEIGHBORHOOD SCHOOL DISTRICT MAP FIG. 13.— THIS APPROXIMATES MOST NEARLY THE PRIMARY GROUP CLASSIFICATION Here are 39 primary and 4 secondary classifications. A number of the neigh- borhoods and the districts correspond almost exactly. Rural Primary Groups 55 NEIGHBORHOOD HIGH SCHOOL AREA vm/mm/t/M t ™ 7r7Tr> > ^ fEST DANE COUNTY, Z$v/ lines Neighborhood Boundaries / / Township High School V . Area High School / Pupil Vi t A Madison high / School Pupil £ FIG. 14. — HIGH SCHOOL AREAS AS RELATED TO NEIGHBORHOOD GROUPS These are simply areas of influence or of patronage and not the legal high school areas themselves. 56 Wisconsin Research Bulletin 51 NEIGHBORHOOD PARISH MAP Tnnnu n > n )n NEIGHBORHOOD \ Boundaries ' Township / Parish Boundaries v FIG. 15.— CHURCH PARISH BOUNDARIES The red lines represent the church areas tributary to the various centers rather than exact parishes of particular churches. Rural Primary Groups 57 NEIGHBORHOOD SOCIAL AREA MAP resT cou,,TY ^ Neighborhood / Boundaries M Township mmm / Overlapping / Social Areas / Small Social / Areas within j Larger Areas j Social Area / Boundaries " FIG. 16. — LITTLE SOCIAL LIFE CENTERED WITHIN THE NEIGHBOR- HOOD GROUP Much of the unorganized social life is taken for granted but there is evidence of the influence of the nearby village. 58 Wisconsin Research Bulletin 51 The nationality settlement , for the present day is also a very uncertain unit of measurement, in fact much more so than formerly because various local histories, biographies and private records use the term constantly to express geographic relation- ships. The neighborhood historic map pictures this tendency. As a primary unit it is negligible quantity, yet because of its strong structural power in the secondary sense it ranks rather high. No mapping is attempted. The next measurement is the church parish boundary. Figure 15 was drafted after a careful first hand study of the county had been made. Pastors, leading laymen, business men, farmers and bankers were interviewed with this one purpose in mind. This parish study was made entirely after the neighborhoods had been studied and mapped, in order to prevent the one pur- pose from influencing the other. Next to the school district comes the church parish as most nearly equalling the neighbor- hood boundary. In the large laboratory map, the parish boun- dary of each church was mapped separately but when super- imposed on the base map the lines became undistinguishable. Figure 15 therefore represents the church areas of the various centers rather than exact parishes of particular churches. Where there is but the one church at the center, whether village or open country, the boundary does represent the individual parish line. In the southwestern section especially, the two boundaries are frequently found coincident. Due to the mapping process re- ferred to above, there is conspicuous lack of overlapping which the laboratory map brings out rather strikingly. Even so, there are areas where two church centers show competition. Various margins and open areas also can be found where seemingly no church is having its influence deeply felt. Another of the area comparisons is the social area. This was obtained by the question, “Where do you go for social affairs such as parties, socials or dances, etc.?” The answers to this question are probably the least satisfactory and definite of those received, perhaps because of the indefiniteness of the question, but also because of the very uncertainty and lack of concen- tration of this phase of life itself. The results are mapped in Figure 16. The tabulation shows this area as represented by a local organization or club rather weak and vague except as a secondary choice. The map shows considerable overlapping but Rural Primary Groups 59 the surprising thing about the answers was that they indicated so little social life centered within the neighborhood group itself. There are a number of such reports to be sure, but much of the unorganized social life of the group is taken for granted and then again the evidence of the influence of the nearby village is rather striking. The ever present residue is here classed, as those groups with- out form. The boundaries to be sure appear on the maps, but there seems to be nothing distinctive about the lines in their relation to the structure or in their relation to the other recog- nized comparable units. One more area needs to be discussed in this section although there have been no corresponding functions or structures outlined. It is the communication-transportation area. It would seem difficult to overestimate the importance of this factor in its relation to these rural neighborhoods and espe- cially at such times as these, when the group movements and changes seem so rapid and sometimes so uncertain of direction. As a beginning for the study of this section, Fgure 17 is intro- duced, showing a country road map and an electric power line map imposed on the neighborhood base map. 15 The solid lines with the numbers in circles show the state trunk highway sys- tem as it was maintained for travel in 1921. The solid lines with the letters in the circles show the principle secondary highways which the county guaranteed to patrol in 1921. The broken lines show secondary roads maintained solely by the local government units. This road system clearly shows the city of Madison to be the hub for a wheel arrangement, as it is also for the railroad system. The village centers are seen as the secondary focal points in this transportation system. Although perhaps electric power lines cannot be classed under transportation in the strict sense of the term, yet for convenience the mapping was introduced here; but its justi- fication for consideration stands solely upon its importance with reference to any discussion of rural society. Unfortun- ately time did not permit of location of the various former tributary lines which tap these power lines at various relay stations. One line in the northwestern section was found, 13 The road map was drawn from maps furnished by the Wisconsin Highway Commission; the electric lines from maps furnished by Mr. Damon of the Wisconsin Railroad Commission. 60 Wisconsin Research Bulletin 51 NEIGHBORHOOD HIGHWAY AND ELECTRIC POWER MAP F1G . 17.— A COMMUNICATION-TRANSPORTATION t It is difficult to overestimate the gvouV movements the rural neighborhoods and especially at present wnen tnc gx and changes seem so rapid and often uncertain of direction. Rural Primary Groups 61 NEIGHBORHOOD MAIL SERVICE CENTERS ©Brooklyn ©CAMBRIDGE @ Cottace Grove ©Dane ® Oecrfielo ©De Forest © Edgerton © Macfarland © Madison © Marshall ©MOARISONVILLI ©Oregon ©Stoughton © Sun Prairie © Waunakee ® Windsor mt/antttu L. FIG. 18. — THE RURAL DELIVERY AREAS AS RELATED TO THE NEIGH- BORHOOD GROUPS The lines represent actual routes but not all the routes from a center are shown. They indicate rather the farm areas which have the same post office address. 62 Wisconsin Research Bulletin 51 however, with about 40 farmer patrons who used current for lighting, for power to run pumps, cream separators, power washers and to do the family ironing. To one, who has grown up on a farm in the days of kerosene lamps and lanterns, long handled pumps and wash boards, the change is certainly strik- ing. Local farmer companies are organized which build the tributary lines and connect with the larger power lines. The basis of selection of members of these local companies and their relation to these headquarter relay stations is the point for attention in this discussion. A second factor in this communication-transportation dis- cussion is that of the rural mail service. Under the discussion of the rural institutions, the tabulation number XV showed that at one time 34 of the primary groups had performed the function of local mail distribution. At the present time but six are charged with this duty and only two of these, Klevenville, Basco or Frenchtown as shown on the base map have more than a very small local delivery, that is, are starting points for rural routes. Figure 18 maps the rural delivery areas in their relation to the neighborhood groups. The lines represent actual routes but not all the routes from a center are shown. The purpose was rather to show the boundaries of influence of these centers in this par- ticular respect. The circles in the lines carrying letters refer to the center from which the route starts. The legend will assist in the identification of each. As in the case of the township lines, no respect is had for local neighborhood boundaries. It is a little difficult to understand by looking at the map just what lines or principles were respected, for a certain farmer has his post office address as Dane while his neighbor in the small local group, gets his mail from Waunakee. Marshall and Sun Prairie lend to confusion in another group, Basco, Verona and Oregon in still another. This seems to indi- cate that a local group of farmers are given their mail or post office headquarters without any reference to their convenience or choice. To the city business man this would be a serious handi- cap ; to the farmers it is often an inconvenience for it leads to confusion when others try to locate his place of business by use of this post office address. It is also a source of difficulty when a local group tries to effect an organization of some kind and finds that its lines of communication are not centralized but scattered. Rural Primary Groups 63 NEIGHBORHOOD TELEPHONE SERVICE FIG. 19. — RURAL TELEPHONE CONNECTIONS Only the areas surrounding' the various exchanges are indicated for these are the controlling factors in communication. 64 Wisconsin Research Bulletin 51 The next line of communication is the rural telephone. Only the eastern part of the county is shown on the map, Figure 19. The danger of confusion of detail was present again so only the areas surrounding the various exchanges were indicated, for after all these are the controlling factors in this matter of communica- tion. A final area of influence in the matter of communication is the local press. Figure 20 shows the areas of circulation for the va- rious local publications. This agency has within its grasp pow- erful means for influencing the thought and action of rural people. This local press in addition to centering attention in the village or city, contributes directly to the local consciousness of the coun- try groups. It carries in its news columns the “personals” and the local happenings of various groups under the captions of their own neighborhood names, even down to the detail of telling which of the neighbors had chicken dinner with Mr. and Mrs. John Doe and family last Sunday. ■V n . Rural Primary Groups 65 NEIGHBORHOOD PRESS MAP FIG. 20. — AREAS OF CIRCULATION FOR LOCAL PUBLICATIONS This agency has within its grasp powerful means for influencing the thought and action of rural people. 66 Wisconsin Research Bulletin 51 PART V THE PRIMARY GROUP AND RURAL ORGANIZATION Investigation and study should end in action. Organization is the watch word of the day. Therefore, what is the relationship of these rural primary groups to the whole field of organization? In order to be as free as possible from confusion, let it be re- called that by the primary or neighborhood group is meant that first group beyond the family which has social significance and which has some local consciousness of unity; that by community or trade area in the geographic sense is meant a group of farms with a trading center. The primary group is essentially a psy- chological thing for the study of which objective standards and geographic measurements have been used. The community or trade area may or may not be a significant social group although it is usually. This study has been concerned primarily with the strictly open country groups in their relation to the individual, to other such groups, and to the groups centering in the villages or cities. The Findings Summarized It has been suggested that going out in an agricultural district to find groups was like trying to find “the beaten tracks on the open deep .” 16 No such difficulties were experienced, for every- where there was evidence of group life and activity. After a few days in the field, enough to whet one’s sense, the social “water sheds” could almost be recognized as one passed along the road stopping every now and then to visit or to ask a few non-committal questions. One hundred and twenty-one groups of rural people have been represented within neighborhood geographic boundaries and they have beside the criterion of local consciousness of unity, the one common objective characteristic of a group name. This name may stand for as many different things as there are groups or it may Sims, N. Li., The Rural Community , p. 141. V Rural Primary Groups 67 stand for a number of differing bonds within the one group. For the sake of getting at bases for discussion, these various possible meanings were classified under seven heads termed functions, as follows: economic, educational, kinship, local government, na- tionality, religious, and social. It was found that 95 of the groups could be located within this classification and 26 for one reason or another were practically without designating factors, that is, were decadent or as it might be expressed, they were present-day names but for traditional groups. About each village or city was found a grouping or focusing of the open country life toward this center. Sometimes a rural consciousness as distinct from the village was found and sometimes, especially with the smaller centers, there seemed to be a blending over until no boundary could be distinguished. These groups owe their original existence to a number of dif- ferent factors or their combination such as topography and original vegetation,' nationality bonds, religious purpose, the mi- gration from a common place of residence and economic or social purposes. Again these groups are changing things. Only 18 were found in which changes were not easily recognizable. These changes are partly to be explained on the basis of shifting popu- lation, modification of institutional arrangements and leadership, and in improvements in means of communication and trans- portation. Conclusions Significant for Rural Organization Organization Plans Must Recognize Rural Primary Groups. The first significant conclusion which may be drawn from the study is that there are real rural groupings in the open country and that these groups are of such importance as to render any scheme of organization questionable which deliberately leaves them out of consideration. These group lines or “mixing” lines, as Super- intendent Thompson calls them, cut across arbitrary townships and county lines, appear within trade area boundaries and tend to approximate other recognised lines in the following rank as judged by frequency: school district, church parish, trade area, and social area, with family circle or nationality settlement prac- tically without meaning at the present time. The groups are not the results of accident but of achievement, sometimes a conscious process and sometimes not. The social product, unity or solidar- 68 Wisconsin Research Bulletin 51 ity, has come about by adaptations which include both intimate association and conflict. This social product cannot be lightly set aside. Even though it has been shown that nearly half the groups have lost their original characteristics, yet the majority which remains far outweighs those which have gone. They are the ones which have accommodated themselves to the changing needs. They have expanded their areas; they have organized their resources. Evidence is not lacking to indicate that the farmer wishes his own groups to count, wants his interests recognized and even demands that he be given every advantage which organi- zation can bring. Many Rural Primary Groups Are Too Small For Efficient Service. The exponents of better methods for farmers’ co- operative enterprises have been advocating these many years that one of the prime essentials for success is “sufficient volume of business.” They say that this is necessary in order to make pos- sible certain specialization which in turn is needed to give the maximum of service at a minimum of cost to the individual mem- ber. This principle applied to creameries means, they say, 100 to 150 families depending on the individual farm production in order to reach about 200,000 pounds of butter fat. This principle applied to churches according to the Ohio Rural Life Survey shows that for 19 counties only 48 per cent of the rural churches with a mem- bership of 101-150 grow, while 59 per cent with memberships of 151-200 and 76 per cent with memberships of 201 and over are growing. 17 This means, in order to get into the class where growth can be assured, that upwards of 50 families would be needed if the size of family were five as the figures show it is for Dane county. George S. Dick of the Wisconsin State Department of Education would apply this same principle to the educational service of the rural community. He says that at least 100 to 150 children are needed to make possible an efficient consolidated school which carries with it the possibility of a local high school. In fact, he argues that the high school feature is one of the biggest reasons for consolidation. This would mean at least 75 to 100 families. If the 26 neighborhood groups are excluded which were classed as decadent, we find that the balance averages a little short of 30 families per neighborhood. If these 26 are left in as they should be, since the total includes the families within their bound- 17 Vogt, “Introduction to Rural Sociology, p. 315. Rural Primary Groups 69 aries, the average becomes slightly over 21 per neighborhood. As the average size of farm in the county was about 148 acres, 100 families would require a land area of a little less than half of the area of an average township. The groups mapped, how- ever, average considerably less than a quarter of a township. The conclusion seems rather inevitable then, that a neighborhood self- sufficiency scheme of rural organization is not a present day pos- sibility, especially when such a large number of small groups are found which tend to bring the averages down below the point of efficiency. Other Primary Groups Render Distinctive Services But of Limited Number. Approaching the case from slightly different angle, suppose one were to attempt another classification of the groups represented on the base maps, this time not simply to rule out those which gave evidence of lacking any creating or holding factor, but on the positive basis of distinctive service rendered by the group as a group. This means not such service or function as would be performed in a more or less off-hand manner whether the group were effective or not, such as the township elections, the running of the district schools or the operation of a country store, or garage, but rather such unity of activity in which the group as a group works as a present day conscious force. A classification on such a basis involves personal judgment to be sure rather than statements from members of the group in ques- tion. Such a tabulation, however, has been attempted and is presented under Table XVII. The caption headings are shifted a little to include communication-transportation services such as Table XVII. — Groups Classified on Basis of Distinctive Present Day Services Performed By Group Action Number of groups and services performed Section of county Num- ber Eco- nomic ( trade) Educa- tional Reli- gious Social Farmer organ- ization Com- muni- cation, trans- porta- tion ! Two or more services Total 60 20 | 28 35 22 13 4 40 East _ 31 11 14 16 10 7 3 20 West 29 9 14 19 12 6 1 20 70 Wisconsin Research Bulletin 51 telephone central, rural mail route headquarters, or local good roads association, while family and nationality factors were dropped. This tabulation shows a total of 60 neighborhood groups performing services as follows: economic (trade), 20; educa- tional, 28; religious, 35; social, 22; farmer organization, 13; communication-transportation, 4. Those groups performing two or more of these services or functions totaled 40. There were 5 groups which performed four of the services, but none more. These services are distinctive and must be recognized as such, especially when they are performed by groups large enough to render them efficient from the standpoint of sufficient volume of business. This was the case with the majority of those groups within the distinctive service classification. Nevertheless the evidence shows that the neighborhood group does not render a sufficient number of different services to meet the needs of its own members. This does not imply that such a group is useless but rather that its organization plans must extend beyond its own borders as well. Table XVIII — Village and City Centers Classified on Basis of Services Rendered to Rural Primary Groups Number of centers and services rendered Section of county Number Eco- nomic (trade) Educa- tional (high school) Reli- gious Social Farmer organ-J ization Com- muni- cation , trans- porta- tion Total of 5 services Total * 29 19 29 29 23 1 27 23 East 1 15 ! 15 i 10 15 15 11 1 13 10 West 14 i 14 9 14 14 12 i 14 12 The Village is the Farmers’ Service Station. Turning now to examine these village and city centers in their relationship to the rendering of this same set of services to farmers outside their limits, Table XVIII shows a total of 29 such centers with the city of Madison excluded, which render the services as follows: economic (trade), 29; educational (high school), 19; religious, 20 ; social, 29 ; farmers’ organization headquarters, 23 ; communi- cation-transportation, 27. There was a total of 23 centers which Rural Primary Groups 71 rendered five out of the six services classified. The groups and centers included in these classifications are now presented in map form under Figures 21 and 22. The active rural service groups are shaded and the 23 centers rendering five or more of the services are circled. This procedure should indicate sufficiently that a minimum of five services or functions seems necessary to the life of a rural group or even a rural people not recognized as a group. It is evident that all of these services cannot be rendered by the rural primary groups themselves as they have been defined here since only a little over 50 per cent of those shown classify in this distinctive service class and less than 65 per cent of these performed two or more of the necessary services. This may have been a long way around to come to the statement that for pur- poses of rural organization the village or city center must be in- cluded in any plan which involves the farmer group. Therefore, Village and Rural Groups Must Federate. But now that the village and open country groups are tied together in this rural relationship of the six services, the question comes as to how the organization of this relationship shall be effected. From figures 21 and 22, where the active services of the rural groups and their village centers are shown, it becomes evident that there are at least three relationships which must be worked out. They are : first, the relation of the distinctive service per- forming groups to the village centers rendering the five services and of these rural service groups included in this village influence area to each other ; second, the relation of this village and service group area, which now can be called the “community” area, to the non-functional groups and to the non-grouped areas, as well as to the village centers of smaller service ; and finally, the inter- relation of the larger “community” areas themselves. Active Rural Groups and the Village Center Should Form a Community. The first relationship is that of the distinctive service groups to the village centers to the end that together some plan of organization may be effected bringing into play the forces of a larger social unity, which may be called for purposes of geographic location, the community-service area unit. It must be said first of all that this larger unit is not to swallow up the active primary rural groups nor to subject their freedom or au- tonomy to any arbitrary centralized power. People are different. One group cannot satisfy all the demands made by such differing 72 Wisconsin Research Bulletin 51 NEIGHBORHOOD VILLAGE CENTER MAP Service Relationship FIG. 21.— ACTIVE SERVICES OF RURAL GROUPS AND THEIR VILLAGE CENTERS The larger community unit must not swallow up the active primary rural groups. Rural Primary Groups 73 NEIGHBORHOOD VILLAGE CENTER MAP Service Relationship FIG. 22. — RELATION OF GROUPS AND VILLAGE CENTERS SHOW NEED FOR NEW SERVICES Now that village and open country groups are tied together in rural rela- tionship of six services, the question comes as to how the organization of this relationship shall be effected. 74 Wisconsin Research Bulletin 51 elements of a whole community area. The young people have certain demands to be met by organization and so do the women and children. All do not have the same inclinations or ideas regarding social activity or religious expression and so all the interests can seldom be served in one grouping. But there are certain other things which all in such a community may have in common. The point is that organization must come through group activity and group processes. To be sure there are those who fear the organization and recognition of such group conscious- ness but this appears to be the way of progress, for as Doctor Ely says in speaking of the development of legislation regarding private property, “for legislation always represents actually exist- ing forces. If any section of the community does not stand for an actually existing force, it is not and cannot be represented by leg- islation .” 18 Probably as good a statement of the point in question as can be found is made by Miss Follett. “Every group once be- come conscious of itself instantly seeks other groups with which to unite to form a larger whole. Alone it cannot be effective . . . The reason we want neighborhood organization is not to keep people within their neighborhoods but to get them out. The move- ment for neighborhood organization is a deliberate effort to get people to identify themselves actually, not sentimentally, with a larger and larger collective unit .” 19 A practical illustration may be of service. Last fall at Arena the farmers and villagers met as a Parent-Teachers Association for a community program with a grain, fruit and vegetable exhibit and contest. The meeting was in the township high school. Village and country folk were on a par in its management and control. Nevertheless the farm- ers were still farmers and villagers were still villagers and the agricultural teacher had been wise enough to build upon the fact for there were farmer clubs grouped out about the district schools and it was these groups which were competing in the exhibits. The village people had their group organization and also competed in the classes which matched their interest. This same principle lies at the bottom of the Federated Clubs of Walworth County. The farmers are organized into local groups about certain activ- ities and interests of their own. The business men in the service centers are organized about their own interests. Twice a year all these groups assemble in a common meeting but they gather as 18 Ely, R. T., Property and Contract, Vol. I, p. 366. 10 Follett, M. P., The New State , p. 249. Rural Primary Groups 75 groups, once in the summer for a social and patriotic occasion and once in the winter for serious conference upon the common interests and problem of the county. This matter of the larger community organization of the various interests of the rural life must be brought to the attention of the people themselves and especially to those who have been desig- nated leaders of the schools, churches, welfare agencies and the social activity. It must come to the attention and action of legis- lative groups. Some have suggested that the neglect of this prin- ciple led to many of the difficulties in France after the revolution when the country was arbitrarily divided without respect to those groupings already existing. Progress will come as the organiza- tion of those lines of interest which are most simple and upon which all can agree most easily and quickly, are attempted first. For as Professor Cooley says, “All progress must be based upon conformity to what is.” The Non-Grouped Areas Need Organization. The second relationship under discussion is that of this community-service unit, made effective by the federation of the various group inter- ests, with that area not included within neighborhood boundaries. First of all, it should be observed that there is more than coinci- dence about the fact that there is an ungrouped area about each village center as shown on the base maps. Life and activity are centered in the village, which dominates the institutional life as well. Attention has been called to the same tendency as shown by a school study in Iowa County. 20 In every case, except where the factor of topography intervened, the schools immediately sur- rounding the village center are smaller than those outlying. The responsibility and the tendency of the center to this immediate non-grouped area, then, seems to be quite clear. It must center in the village. Next in order are the groups still more or less recognized as groups but not in the distinctive sense which was made the basis for the ranking as shown on the maps, figures 21 and 22. Here the group bonds are slipping, the local consciousness is fading due to some factor or combination of factors of change outlined pre- viously. Here the problem is to make a substitution for the van- ishing group. There appears to be more than a chance significance 20 Merritt, E., and Hatch, K. L., Some Economic Factors Which Influence Rural Education in Wisconsin, Agricultural Exp. Sta. Research Bulletin 40, University of Wisconsin (1916), p. 31. 76 Wisconsin Research Bulletin 51 in the expression that it is quite as serious for a man to be without a neighborhood or community as to be without a country. There were powerful forces at work upon the individual in many of those early settlements which have been described. It was here that the boy and girl learned their lesson in society, learned that rights were matched with duties with reference to this group life, that privileges were paired with responsibilities. This is not a matter of sentiment at all, it is a question of group control, of group standards and loyalties and of co-operative experience. A case in point is of a young and well-educated couple just start- ing to farm on what was the site of the old Hyer’s Corners Inn. They were called upon and this neighborhood idea was discussed from every angle. The question card had come to their home and they -were at a loss to know how to answer it. Here was their sit- uation : They received their mail and went to farmer organization meetings in Dane village; they traded in Waunakee; they went to church in Martinsville ; they sent their milk to a creamery at Springfield Corners, and they were within the school district to the north known as Elm Grove. The question is not raised as to which of these bonds should be cut or that any loyalties should be weakened, but it is contended that this scattering of interests is obviously impairing the efficiency of this family unit. It was a unit by itself but looking for a larger group relationship. The automobile has been one factor in this whole situation also. The early acquisition of the machine has excited a rather exagger- ated sense of freedom, for where five miles was formerly a limit, now it is twenty-five, and more. The automobile must be “domes- ticated” and made to do service for neighborhood and for commu- nity as well. Again, families were found who were rather on the “fringes” of the recognized groups, casting their lot in with the group when some activity turned to their advantage but keeping aloof from complete identification. In some cases these “fringes” of families would also appear in certain “non” areas with reference to church, social activity, or farmers’ organization work but in these latter arrangements, the maps fail pretty largely to show the relationship since when the larger boundaries were located many families not reporting are entirely surrounded by those reporting the center for which the lines were drawn. Certain designations, however, for these non-grouped families were frequently caught in more or Rural Primary Groups 77 less casual conversations such as, “They don’t keep the Sabbath,” “They are free thinkers,” “That is a mixed marriage,” (meaning that people of two church groups had married, resulting in their not identifying themselves with either group completely), “They don’t belong anywhere,” and “Oh they are — (Naming a nationality different from that of the group in question). We don’t mix with them. They are alright but they seem to handle things differently than we do.” A group of farmers was encountered on the road and in the course of the conversation were asked where they did most of their trading. This seemingly touched off a line of discussion not new to them for they began to protest at the treatment and prices they had encountered in the two centers between which they had evidently been oscillating in their trading. Finally they were asked whether they had ever tried carrying their protests as a group to the group of village business men of whom they were complaining and seeing if some understanding could not be reached. They had not tried this to be sure because they had developed no sense of neighborhood grouping in relation to other groups, not to say a sense of community group life. In many instances, therefore, the organization activity with respect to these non-functional groups or non-grouped areas will have to be the process of realignment and integration. As has been intimated earlier, many of the groups may be expanded to include people whose differences are no longer sufficient to keep them out. Such is the case with Hanerville and Dunkirk, with Gaston and Pierceville, with Happy Valley and Erbe Valley, the First Luther and Mud Lake, and so on. In other cases it should be the organization of new neighborhood units either where the old have ceased to render services needed at present, or where there has been a complete lack of group activity for a long time. Oregon and Fitchburg townships are the illustrations of the first, and Dane and Verona townships of the second condition. Where there are “fringes” as they have been called, of families which do not seem to fit anywhere, patience and tact will be required. Their interests and that of their children must be searched out and ap- pealed to until finally a sense of the group life will reappear for it is there even though hidden for the time. Finally in the matter of the relation to the village centers, not rated in the five service class, a problem is raised which 78 Wisconsin Research Bulletin 51 will have to be postponed until another year when the character of the services rendered from the standpoint of this organized community group can be studied. It is probably not too much to say, however, that before so many years, centers now giving only very limited service will have to improve their kind and number of services or stand up and -show cause why they should exist. Inter-Community Organization is Necessary for Administra- tion Purposes. The third relationship under this section re- mains. It is that of the organization arrangements of the various integrated community service groups themselves. The very brief word on this relationship would seem to be carried in the Wal- worth county story told above and which could be repeated for various other counties in the state. The county under present arrangements at least seems to be the administration unit for a number of rural organization enterprises, such as the County Agricultural Representatives, the Home Demonstration Agents, the Farm Bureau, the County Supervision of Schools, the County Nurse System, and the County Y. M. C. A. plan. Some scheme of federation of the local work or community units can serve as a decided advantage, and this will come as the community rela- tionships of group and groups are understood and become the basis for social progress. Practical Implications For District Schools and High Schools. The district school has been distinctly the institution of the rural neighborhood. With the neighborhood it must adjust itself to changing conditions. Many districts are face to face with the problem of maintaining an efficient school with limited resources and with only a few chil- dren in attendance. Buildings and equipment also need replacing. The inevitable answer is consolidation. For the high school the implication passes over into an open challenge for in the real sense of the word, the farming group does not have a high school. Country boys and girls go to high schools but their parents do not participate in its management since their farms lie beyond the legal boundaries of the district. One superintendent, not of Dane county, would determine upon the distribution of the schools by comparing the number of chil- dren with the square miles of area in a county, and then establish an arbitrary ratio. The present study would indicate that such a Rural Primary Groups 79 procedure would be fatal to the best interests of education. Group lines should be taken into account. The community high school, where the area is not too great or the center too large, comes nearer the ideal. The farmer’s voice is needed in such a school to give it the vocational emphasis which is required, for unless the farm boy or girl can learn to love and respect his chosen work he cannot stay on the farm. For Farmers’ Organizations. In Wisconsin many of the national farmers’ organizations have not taken the neighborhood nor the community units into consideration in their organization plans. The Farm Bureau Federation is using the township for its unit. One organizer argues that it was the best unit because it was the easiest to follow and because “it took in everybody.” The farmers’ organization maps presented above indicate, however, that the farmer considers the area surrounding the meeting places as the unit. Moreover, as one travels over a township said to be organized, not township, but group organization is found. The fact that one group is organized may perforce, mean that another in the same township is not organized. The two groups may not be compatible; they may be divided by natural barriers; the “traveled” roads may lead in opposite directions. The great assets of group loyalty, mutual confidence, and local leadership — all so- cial products of group life — are sorely needed by farmers’ co- operative organizations. These can best be secured by utilizing the neighborhood and the community. For Village Commercial Associations. Village and city business men have long since learned some of the lessons of group activity. Sometimes it may not have been from choice, but ne- cessity. Many farmers not initiated into the ways of “organized business” are suspicious of such associations and some maintain that they have grounds for such suspicions. Whatever may be the fact in the case, it seems hardly probable that the farmer will be won over by being offered individual membership in these business men s associations. Nor does it seem likely that he will respond to efforts directed toward the organization of farm bureaus as ad- juncts to Chambers of Commerce, as was recently attempted by one city center. Group activity of his own is rather the method which the farmer is at present disposed to employ in effecting his bargaining and in presenting what he believes to be his rights 80 Wisconsin Research Bulletin 51 before legislative bodies. All this by no means precludes the pos- sibility of the business man and farmer working together; it rather enhances such a possibility. When both are organized they can meet on equal footings. What is more, as has been shown, each is indispensable to the other. The farmer must have his service station. The business man must have his clientele. This village commercial association has then a real place in the life of the larger community unit. It has a real chance to perpe- trate its own interests by placing its services at the disposal of the farmer group and by promoting, not exploiting, the resources of the landed area round about. For the Rural and the Village Church. The case of the church is not easily stated. It seems apparent that in some in- stances the open country churches are losing ground. The ten- dency seems to be toward a centering in the small hamlets and villages. In the county, however, a number of Lutheran and Catholic churches offer very striking exceptions. In the larger centers a discriminating distinction was often made between a farmers’ church and a business and professional man’s church in the same city. The laboratory map indicating the parish boundary for each church, together with the location of the pastor’s resi- dence, presents a formidable problem of overlapping areas, of inefficient administration and of neglected fields. These smaller centers and their rural constituency are face to face with a lack of economy and in many instances with a sheer extravagance, both in the distribution and size of their church plants and in the inefficient services of their untrained clergy. No one who is fa- miliar with conditions at all would argue for some idealistic plan of complete amalgamation of the church bodies, for the reasons outlined under the discussion of group differences, yet the eco- nomic pressure if nothing else, must drive certain closely related groups to a rearrangement of their work along lines of neighbor- hood and community organization. For Social Welfare Agencies. The county will doubtless be the administrative unit for social welfare work in rural districts for some time to come. It is about the smallest unit which can maintain a financial budget of sufficient size to allow the employ- ment of professional services. When it comes to the working unit, however, whether it be on the professional or the voluntary Rural Primary Groups 81 basis, the neighborhood and the community group arrangements must be taken into consideration. As in the case of farmers’ or- ganizations, local loyalties are needed. Group opinion is essen- tial to the establishment of certain standards of health, for ex- ample, as well as for the enforcement of various laws, even in such matters as compulsory school attendance. Headquarters may be maintained at the county seat for the influence of such centers is marked, yet the worker will have to cast off its urban atmosphere when she goes into certain rural groups if she would have her services effective there. Many agencies are looking to the rural field but thus far they have been looking through city glasses and have seen only the larger community centers. The smaller groups need not so much of professional help, perhaps, as they do the discovery and training of local leadership. There is need for recreational activity also for through this door will be welcomed other forms of health and welfare work. Local cen- ters need accommodations and buildings, which even the local farmer can feel are his very own. In a word, then, social and welfare work will succeed as the local units of society are rec- ognized and are made to feel the responsibility of being the source of the stream of social life. Growth to be continuous must be an indigenous growth with roots planted deep in local soil. CONTENTS Page Part I. — Rural Primary Groups and Their Discovery 3 Importance of rural groupings 5 Rural primary groups defined 5 How groups were discovered 6 Mapping the groups 10 Part II. — Genesis and Tendencies of the Groups . 11 Historical and present day groupings compared 11 Influence of topography and vegetation 17 Population and nationality factors 22 Source of group names 29 Part III. — Group Changes and Processes 31 Group changes - 31 Group processes 33 Part IV. — Function, Structure, Form 37 Functions — economic, educational, kinship, government, nationality, religion, social 37-45 Structure 45 Institutions 46 Form — Trade area, school district, family circle, town- ship, nationality settlement, church parish, social area i 47-65 Part V. — The Primary Group and Rural Organization.... 66 The findings summarized 66 Conclusions significant for rural organization 67 Practical implications for social institutions 78 Research Bulletin 52 BUWRsm of «- SEP 3 July, 1922 The Development and Winter Injury of Cherry Blossom Buds R. H. ROBERTS AGRICULTURAL EXPERIMENT STATION OF THE UNIVERSITY OF WISCONSIN MADISON, WISCONSIN CONTENTS Introduction . . 1 Status of the Problem 2 Development of ti-ie Blossom Buds . 3 Methods .; 4 Relation of Stage of Development to Killing 8 Winter Killing of the Blossom Buds 15 Factor of Susceptibility 19 Practical Application ... . ....19 Hardiness of Other Meristematic or Embryonic Tissues....20 Summary . 22 Literature Cited 23 The Development and Winter Injury of Cherry Blossom Buds With a Consideration of the Observed Hardiness of Some Meristematic Tissues W INTER KILLING of the blossom buds of the sour cherry (Prunus Cerasus) may be of sufficiently general occur- rence to cause a crop failure, either on individual trees or throughout entire cherry producing sections. There is usually serious injury in some orchards and some districts each year. The failure to secure a crop of cherries in Wisconsin is more apt to be due to this cause than to any other common trouble. Appar- ently sweet cherries (Prunus avium) are not commercially profit- able because of winter injury to the blossom buds; the trees grow well, but the blossom buds are usually killed during the winter and the trees are, consequently, unfruitful. Studies were begun in 1915 upon the prevalence, occurrence and nature of winter killing of the fruit buds of the sour cherry. The results of the observations made during the first two years, upon some varieties of this fruit have been reported previously (11). At the time of that report it had been determined that in general, when unequal killing occurred in different varieties, in different trees of a variety, in different parts of a tree or in differ- ent blossoms of a bud, the extent of the injury seemed to bear a direct relation to the degree or stage of development which the incipient blossoms within the buds had attained at the beginning of winter; the more advanced the blossoms the greater the likeli- hood of injury. The work done since 1917 tends to verify this idea as to the relation between the stage of development of the blossoms and the extent of the injury due to low temperatures. As a result of the present work, also, it is thought that the reason has been found to explain why the stage of development is closely associated with winter injury of the blossom buds. Since the injury occurs in definite regions of the blossom and since these areas of primary injury are almost wholly confined to cells which have a large central vacuole, it is concluded that this vacuolated condition of the cytoplasm of cells in these regions renders them more readily susceptible to killing (12). 2 Wisconsin Research Bulletin 52 If the stage of development of the blossoms and the presence of a large central vacuole are both related to injury, it should be expected that vacuolation would be found to be characteristic of a particular stage of development. This is in general the case, although some notable exceptions have been observed. For ex- ample, blossom buds may have very much enlarged flower parts and yet have but few cells with large vacuoles in the cytoplasm. Such buds are relatively very hardy. In other words, suscepti- bility to injury is more closely related to the conditions of the cytoplasm of certain cells than to the relative morphological devel- opment of the blossoms. The appearance of a large central vacu- ole is taken as evidence of the approach of maturity of the cells. The rate of maturing would be materially affected by the nutri- tional and growth conditions of the trees. This fact gives promise of success in attempts to reduce the injury through cultural means. • Status of the Problem A variation in relative hardiness of the blossom buds has been found in various types of fruit producing plants. Among the re- ports are those of Miiller-Thurgau (8) and of Gladwin (4) upon grapes, of Whipple (14) relative to apples, of West and Edlefsen (13) upon peaches, and of Goff (5) upon cherries. Garcia and Rigney (3) have shown that apple blossom buds become progres- sively more tender as spring development takes place. This is a common experience and, in fact, gives rise to the usual idea ad- vanced to explain the reason for winter killing of fruit buds. This idea is clearly stated by Neilson (9) when he says : “The buds (on peaches and cherries in Ontario) are killed by extreme cold or through starting into growth in mild weather and being subsequently frozen in cold snaps There are indications that the blossom buds develop slightly throughout the winter whenever growing temperatures prevail. And yet, in Wisconsin at least, the development of the blossom buds during the winter season due to intervals of warm weather is not commonly a direct cause in rendering the cherry blossom buds more susceptible to winter injury. Whatever injury may occur appears usually relatively early in the winter season. For example, initial injury occurred during the nights of December 7 and 8 in 1917, January 3 and 4 in 1919, and on January 4 and 5 in 1920. There had been no “warm spell” in any one of these seasons to start the blossoms into growth previous to these dates. Development of Cherry Blossom Buds 3 Chandler* has suggested that the duration of the winter rest period particularly in southern parts of the United States may have much to do with susceptibility to killing, at least in the case of peach buds, because the buds pass the rest period and start into early season growth in time to be injured by spring frosts. The occur- rence of injury in Wisconsin appears to be little affected by ques- tions of winter development, unless it is through unknown changes induced by low instead of relatively high temperatures. Goff (5) suggested that the mechanical protection of the bud scales may play an important role in bud hardiness since he observed that the larger buds in which the scales are somewhat separated, are more tender than the smaller buds, which appear to be more closely covered and better protected. From artificial freezing experiments, however, Chandler (2) found that the scales apparently have no such protective function, for buds with the scales removed are as hardy as normal buds. Goff likewise con- sidered that the position of the buds on the tree bears an impor- tant relation to hardiness. He found the most extensive damage to occur “on the most exposed parts of the tree, as the ends of the branches” and stated that “the centrally located buds (in the trees) were decidedly less injured than the outermost ones.” On the basis of the present studies during the past several years it seems probable that the important factor concerned in winter hardiness of blossom buds is not the protection afforded by posi- tion, but rather the less advanced state of development of the buds on the spurs in the shaded inner portion of the trees as com- pared with those on the terminals. It would seem, therefore, that more emphasis should be placed upon the state or condition of the blossom tissue of the buds and the susceptibility of such tissue to injury, than upon the external and mechanical conditions sur- rounding the buds. Development of the Blossom Buds When it became apparent that the extent of winter killing of the blossom buds of the cherry is directly related to the degree of their development at the time when excessively low tempera- tures are experienced, the question arose as to what might be the relation between the time of initial differentiation and total devel- *Chandler, W. H. Cornell Agr. Exp. Sta., N. Y. Lecture Notes 1918. 4 Wisconsin Research Bulletin 52 opment. It was realized, however, that the rate of expansion after differentiation might be more closely correlated with the final development attained at the beginning of winter, than might the time of initiation of the blossoms. A study was subsequently made of the seasonal development of blossom buds of Early Rich- mond cherries at Madison during the year 1917. It had been found, as previously reported, that there are marked and consistent differences in the extent of the killing of buds, directly related to the length of the seasonal growths on which the buds are located. The hardiest buds are on spurs (growths of less than a half inch long), somewhat more tender buds are borne on terminals of 7 to 8 inches in length, whereas the most tender buds are on terminal growths 2 to 3 inches in length. Fur- thermore, on the terminal growths there are differences in the degree of bud hardiness ; the buds near the tip are most hardy, the buds near the base are somewhat less hardy, and the buds along the mid-portion of the growth are least hardy (Figure I, [3]). The following series were used in making a study of the development of the “hardy” and “tender” buds. A. Buds on spurs (0 to % inch in length), on 34 to inch spurs, on 1 to 2-inch terminals, and on 7 to 8-inch terminals. B. Buds at the base, center and tip of 1 to 2-inch, 4 to 5-inch and 7 to 8-inch terminals. Methods The buds collected during the season of active vegetation were fixed in the field. The dormant buds were taken to the labora- tory and the tip of each bud was cut off before fixation in order to secure more rapid penetration of the fixing agent. Chrom- acetic fixative was used. The material was stored in 70 per cent alcohol until the time of examination. It was not necessary to imbed most of the material since free- hand and hand microtome sections mounted in glycerine were satisfactory for the study of the stages and of the rate of develop- ment of the buds. Approximately 2,500 buds were examined. The following system was used in determining the average de- velopment of the blossoms : A number of buds of a series were sectioned and glycerine mounts of the sections were made. Dur- ing the examination of these preparations, a chart was constructed Development of Cherry Blossom Buds 5 FIG. 1 — BLOSSOM DEVELOPMENT (1) Diagrams showing the pistil and anther development of blos- soms collected October 15, 1917. (2) Sketch showing appearance of two blossoms within sectioned buds. (3) From three to five blossoms form in each blossom bud. Most of these were winter-killed along the cen- tral and lower parts of the 4 to 6 inch growths in 1917. (4) Unequal development of the blossoms occurs within the same bud. This results in unequal hardiness and killing. 6 Wisconsin Research Bulletin 52 Figure I [1] to show the range in development of the vari- ous buds on a certain date. Comparisons were usually based upon the appearance of the floral parts in samples collected on the early dates and upon the relative size and form of the parts in samples collected later. The pistils and anthers in the more advanced blossoms were carefully observed for the appearance of sporogenous tissue. After having prepared a chart as illus- trated in Figure 1, additional buds of the same series, usually to a total of fifty, were cut longitudinally and examined with a bin- ocular microscope. This provided a rapid but satisfactory method for comparing the relative stages of development of the blossoms (Figure 1, [2]). The blossom development was noted and re- corded in comparison with the diagrams on the previously con- structed chart. Following is a typical record of a series showing the number of blossoms with a development equal to that repre- sented in each respective diagram: Diagram 1, 1 blossom; 2, 10 blossoms; 3, 27 blossoms; 4, 12 blossoms; 5, 0 blossoms. As the greater number of blossoms had a development corresponding to that shown in diagram 3, a section of a blossom at this stage of enlargement was used to make a drawing to represent the series. It was after this manner that the camera lucida drawings in Fig- ures 2 and 3 were obtained. The time of differentiation and the relative rate of subsequent development of the blossom buds of Early Richmond at Madison are shown by Figures 2 and 3 and Table 1. TABLE 1.— COMPARATIVE DEVELOPMENT OF CHERRY BLOSSOM BUDS DURING 1917 (Lowest rating shows 1 greatest development) Length of growth— Central used for comparison buds Buds along 4-5 inch terminals

; 6, x No. of hill9 8 8 8 8 8 5 1 w Av. no. tubers to hill 10.7 6.62 8.87 8.12 11.1 4.6 2.0 Av. wt. per tuber__ 3.03 8.9 6.43 6.2 3.9 .64 1.5 Soil temperature °C. 11 14.5 18 21.5 25 28.5 30.5 > a H No. of hills 8 8 8 8 8 8 7 w Av. no. tubers to hill 4.33 10.0 7.0 9.16 15.5 14.5 5.5 Av. wt. per tuber 3.8 7.02 12. 85 9.25 4.71 1.99 .34 •One and two plants were removed from 12° and 15° C. tanks respectively at the time the data was taken on the plants growing' at the other tempera- tures in the experiment ; these data are represented by the numerators while the denominators are the data for the remaining plants grown at 12 and 15° C. which were removed eighteen days after the removal of the plants from the higher temperature tanks. The numerator figures are comparable with the figures recorded for each of the temperatures above 15°. 28 Wisconsin Research Bulletin 53 Underground Parts T ubers Variations in soil temperature seem to have less consistent in- fluence on the number of tubers produced per hill than on any of the other host activities thus far observed. However, there does seem to be a tendency for the greatest number of tubers to de- velop at 15° C. and at 25°-28° C. There is a reduction in the number of tubers per hill at the intervening and the extreme tem- peratures. This curve tends to be the reverse of all other curves representing the plant activities observed in this work. The size of tubers is influenced considerably by variations in soil temperature. During the period involved in the various tank experiments, the largest tubers were produced at soil tempera- tures ranging from 15° to 22° C. with the optimum at or near 18° C. However, when plants are allowed to develop beyond the period of the main experiment, as was the case with the plants held at 12° and 15° C. in Experiment II, the tubers progress rap- idly and increase in size as shown in Table V. It will not be sur- prising to find that the optimum soil temperature .for mature tuber development is somewhat lower than just indicated, when a complete series can be carried through to maturity. Observations have not been made to determine the soil tem- peratures at which tubers commence to set first, but it is believed that this takes place at the temperatures which favor early sprout- ing and the emergence of the sprouts from the soil. The indica- tions are that the optimum temperature for the early setting of tubers is somewhat higher than the optimum for a yield of large tubers. The shape of tubers is influenced to a considerable extent by Table VI — The Ratio of Width to Length in Tubers Grown at Different Soil Temperatures. These data are based on measurements made on the whole population of Experi- ment V. 1 Soil temperature *C. 11 14.5 18 21.5 25 28.5 , 30.5 Ratio of tuber width to length. 1:0.9 1:1.02 1:1 1:1.12 1:1.14 1:1.25 ' l:l.6 1 The Irish Cobbler seed used in this work was typical of the variety which tends towards the globular shape. Most of the tubers were prac- tically circular in the cross section intersecting the stem and bud axis at right angles, and for this reason width is designated as a single factor in Table VI. Influence of Soil Temperature on Potato Scab 29 soil temperatures as shown in Plates I, II and IV. At low soil temperatures the length of Irish Cobbler tubers is less along the stem-bud axis than along the transverse axis, while at the higher temperatures the stem-bud is much the longer and tubers tend to become egg or pear shaped, as shown in Plates I, II and IV. Ac- tual measurements made on the population of a complete temper- ature series show that the ratio of the width to the length of tubers grown at 11° C< is about 1 :0.9, while at temperatures near 30° C. this ratio approximates 1 : 1 .6. The ratios developed at the intervening temperatures gradually approach the latter ratio as the temperature rises forming a rather regular curve as is shown by the data in Table VI. Fitch (1) has noted that under certain conditions potato tubers tend to elongate and develop the pear shape. He associates this condition with drouth, the “running out” of seed stock and sea- sonal conditions. While factors other than soil temperature may influence the proportional dimensions of tubers, the results herein recorded show definitely that soil temperature is an important factor in determining tuber shape, a point which is of economic interest to growers who produce exhibition seed stock. It is a matter of common observation that lenticel development on potato tubers is stimulated by certain moisture conditions. In these experiments where soil moisture was kept approximate- ly uniform, it has been found that lenticel character has also been influenced considerably by soil temperature. These organs have not been conspicuous on tubers developed at low soil tempera- tures, but have become large and protruding at the high tem- peratures. An evident suggestion is that their relative develop- ment may be associated with respiratory metabolism. The influence of soil temperature upon the chemical composi- tion of the tuber has not been a matter of direct inquiry in con- nection with this work. It may be assumed that the composition is so influenced and a type of evidence that bears upon this de- serves record. In Plate V is shown a complete temperature series of tubers grown in uninoculated soil. These tubers were photographed after storage in 70 per cent ethyl alcohol for five months after their removal from the soil. The tubers pro- duced at the low temperatures are jet black in color, whereas those grown at the medium temperatures show practically no dis- coloration. At 30° C. the tubers show a slight darkening, but to 30 Wisconsin Research Bulletin 53 Table VII — Data Showing the Influence of Soil Temperature on Certain Host Developments. These data are the average of the results from experiments II, III, IV and V. All data are based on determinations made at the close of each experiment. All weights are in grams and linear measurements in centimeters. Soil temp, degrees O.* H 12 14.5 . 15 18 21 21.5 24 25 27 | 28.5 27 30.5 Av. No. of tubers in inoc. hills 8.2 12.5 11.2 12.5 12.8 14.4 10.2 Av. Wt. of tubers per hill 22 42.7 50.0 54.2 51.0 25.0 9.5 Av. Wt. per tuber 2.55 5.44 : 7.0 6.0 4.2 2.1 .80 Above ground parts Av. Wt. of green tops per hill (Exp. IV and V) 55.1 77.0 82.2 89.0 87.0 70.0 50.0 Av. No. of stems per hill (Exp. II, III, IV and V) Av. height of stems per hill 2.8 3.6 4.2 4.6 4.0 3.21 2.8 Exp. II, m , IV and V) 22.0 23.0 22.0 23.5 25.0 30.2 14.0 Av. diam. of stems (Exp.V)—' Av. No. days for plants to 7.1 8.5 7.9 7.7 6.7 3.9 j 3.3 come through soil (Exp. Ill, IV and V) 23.0 17.5 14.5 ! 12.8 12.8 16.5 24.6 a less degree than is the case with the tubers produced at the low temperatures. The black coloration suggests a melanin relation involving the reaction of various enzymes, tyrosin and other com- plex proteins. While this variation cannot be interpreted at this time, it is evident that the chemical composition of at least the surface tissues of these tubers was influenced by the soil temperatures. The question naturally arises as to what relation such changes may have upon susceptibility or resistance to scab infection or development. These are evidently matters deserving further investigation. Stem bases and stolons These parts of the potato plant seem to be greatly modified by variations in soil temperature, as noted previously by Richards (8). At the high temperatures they become relatively large in diameter and fleshy in nature. The stems are very much larger below ground than they are above the soil line, whereas at the low temperatures the reverse is true, the underground stems being more slender in proportion than the above ground stems. As the temperature advances toward the higher limits of endurance the basal portion of the main stem as well as the stolons evidently * The soil temperatures given are those used throughout all the ex- periments. Since there was so little difference between the tempera- tures used in Experiments II and III, and those used in Experiments IV and V all data in this table were averaged in the usual manner. Influence of Soil Temperature on Potato Scab 31 assumes a storage function. This may be correlated in some de- gree with an inhibition of normal tuber development and conse- quently with a disturbance of translocation processes. As in the case of the tuber, lenticel development is very marked on stems and stolons at high temperatures and much reduced at low ones. Above Ground Parts Richards (8) has published upon data which are essentially in harmony with those obtained by the writers concerning the re- sponse of the above ground parts of the potato plant. The results herein reported, however, cover somewhat longer periods of de- velopment and a larger number of experiments than those which he discussed. Stems Soil temperature greatly influences the length of time required for germination and the emergence of the sprouts from the soil as shown in Table VII. Sprouts emerge first in soil held from 21°-24° C. and in general emergence is earlier at the high than at the low temperatures. During the early life of the plants the height of tops is directly correlated with the time of emergence. A soil temperature of about 27° C. soon becomes the optimum for these organs with a sharp decline in the curve above this tem- perature. Later, however, these plants at the higher soil tempera- tures commence to slacken their growth and come to maturity, whereas the plants at the low temperatures come on slowly and finally surpass plants grown at the high temperatures. In general the life cycle of the potato plant is short at high and long at low soil temperatures. The green weight of “stems” is also influenced by variations in soil temperature, but this factor seems to vary more or less di- rectly with the number of “stems” produced at the various tem- peratures. Apparently the optimum soil temperature for the pro- duction of green weight and that for the production of number of “stems” per hill is at or near 21° C. The two curves differ, how- ever, in that the green weight curve tends to show a wider opti- mum range than the “stem” curve. Both curves decline grad- ually from the optimum point. The development of “wings” on the “stems” is much reduced and the swelling of the nodes is much increased by high soil tern- 32 Wisconsin Research Bulletin 53 peratures and in some cases aerial tubers and auxiliary branches develop at the higher temperatures. The leaves of the potato plant also show variations due to soil temperature. In general the ratio of width to length in leaflets follows the same trend as is the case with tubers. At the low soil temperatures leaflets are wider in proportion to their length, whereas they are longer than wide at the high temperatures (Plate III A, B). At these latter temperatures leaflets are decidedly lanceolate in form, but they tend towards the round type at the low soil temperature. While these observations have dealt primarily with the evident characters, chiefly morphological, it is realized that the more im- portant modifications may be those occurring in the physiological processes, especially in nutrition and including food translocation or storage. This is indicated not only in the nodal swellings and aerial tuber developments noted above, but also in the influence of soil temperature upon chlorophyll, the foliage tending in gen- eral to a deeper green at the lower soil temperatures and a lighter color at the higher. We have not the data to go far in the discussion of the details as to the relation of even this single variable factor, soil tempera- ture, to the normal development of the potato plant and its tuber- ization processes. At the same time it is evident that an adequate discussion of these must include some consideration of their in- terrelation with the other variables in the environment, especially air temperatures and light. While these observations relate primarily to the host plant, their significance must not be overlooked from the standpoint of the disease. As pointed out earlier, the evidence at hand indi- cates that potato scab as a disease is influenced by the conditions of tuber development. These include the suggestion that rapidly growing tubers may scab more severely than slow growing ones, and that there may be such differences in the chemical composi- tion of the tubers developing under different conditions as may influence their relative susceptibility to infection. The rate of the tuber’s development, as well as its composition is, of course, cor- related, with the metabolic and growth processes of the rest of the plant. It is obvious, therefore, that the ultimate understanding of the influence of soil temperature or other variable factors upon the development of a disease like potato scab is conditioned upon further studies concerning the relation of these to host as well as / to parasite. Influence of Soil Temperature on Potato Scab 33 SUMMARY 1. The development of potato scab, caused by Actinomyces scabies, is evidently influenced by several environmental factors. An attempt has been made to secure evidence as to such possible influence of soil temperature. 2. Five series of experiments have been conducted in green- houses using the “Wisconsin tank” method. These have included seven gradations of soil temperature ranging from about 11° to 30.5 C. In all cases under this method the aim is to maintain the other soil conditions including moisture alike and approximately constant, and in each experimental series all the plants are ex- posed to the same aerial conditions. 3. The results show that under these circumstances the devel- opment of the scab disease is influenced by soil temperature. 4. The disease developed at all soil temperatures, 11° to 30.5° C,. but was comparatively slight at either extreme. 5. The optimum soil temperature for scab development as measured by the number of scabby tubers was found to be about 23° C., while the optimum for the percentage of the total tuber surface scabbed was a little lower, about 20.5° C. The conclu- sion reached is that all things considered 22° C. may be accepted as about the optimum soil temperature for scab development, where the “Wisconsin tank” method is used. 6. A field trial was also conducted in which three gradations of soil temperature were maintained during the season of tuber development. Of these the highest, approximately 25° C., proved to be most conducive to scab development. 7. Field observations seem in general to accord with the results obtained by experiments. They indicate that potato scab is com- paratively more prevalent in regions having high summer tem- peratures than in those of lower temperature, and also that in the same district in Wisconsin the disease development is greater dur- ing hot than during cool summers. It is to be noted, however, that such observational data is relatively meager and is to be con- sidered only as suggestive. Attention should be directed to this question by other observers. 8. As bearing more definitely upon this matter it is found that in the leading Wisconsin potato districts the mean temperatures 3,4 Wisconsin Research Bulletin 53 for July and August during the hottest midsummers approximate those found in our experiments to be the optimum for scab devel- opment. 9. Examinations of the scabby tubers from the controlled tem- perature experiments as well as of samples from commercial sources show that the scab lesions tend to be segregated upon the “stem end” portion of the tuber. 10. The evidence from the soil temperature tank series shows that such segregation is less evident at or near the optimum tem- perature for scab development and more apparent at the extreme temperatures. 11. The influence of temperature upon the development o£ the disease must obviously bear a relation on the one hand to effects upon the parasite and on the other hand to effects upon the host. 12. The available evidence indicates that the scab parasite as an independent organism is favored by relatively high temperatures, whereas the potato plant functions better in general at relatively low temperatures. It is, however, noteworthy that the influence of temperature upon the different potato organs is not uniform and that it varies also with the stage in their development. 13. The influence of temperature upon the prevalence of the parasite in the soil may be cumulative from season to season whereas the influence upon the host is immediate and temporary. 14. The immediate relation of temperature to the development of scab seems to be more closely correlated with its influence upon potato tuber development than with that upon the growth of the parasite. 15. It seems evident that a satisfactory interpretation of the relation of soil temperature to the development of potato scab must await on the one hand more critical study of its relation to tuber development and on the other hand a fuller knowledge of the details as to tuber infection and subsequent scab development. LITERATURE CITED (1) Fitch, C. L. 1914. Identification of potato varieties. Off. Pub. Iowa State Col. Agr., v. 12, no. 33 (Ext. Bui. 20), 32 p., 25 fig., 1 col. pi. (2) Giissow, H. T. 1917. The pathogenic action of Rhizoctonia on potato. In Phytopathology, v. 7, no. 3, p. 209-213, 1 fig. Influence of Soil Temperature on Potato Scab 35 (3) Jones, L. R. 1905. Disease resistance of potatoes. U. S. Dept. Agr., Bur. Plant Indus. Bui. 87, 39 p. (4) Jones, L. R. 1922. Experimental work on the relation of soil temperature to disease in plants. In Trans. Wis. Acad. Sci., Arts and Letters, v. 22, p. 433-459, pi. 33-37. (5) r — and McKinney, H. H. 1919. The influence of soil temperature on potato scab. In Phytopathology, v. 9, no. 7, p. 301-302. (6) Orton, W. A. 1913. Environmental influences in the pathology of Solanum tuberosum. In Jour. Wash. Acad. Sci., v. 3, no. 7, p. 180-190, 3 fig. (7) 1916. Report on potato diseases in Bermuda. In Rept. Bd. Agr. Bermuda, 1914-15, p. 13-15. (8) Richards, B. L. 1921. Pathogenicity of Corticium vagum on the potato as af- fected by soil temperature. In Jour. Agri. Research, v. 21, no. 7, p. 459-482, 5 fig., pi. 88-93. Literature cited, p. 481-482. (9) Shapovalov, Michael. 1915. Effect of temperature on germination and growth of the common potato-scab organism. In Jour. Agr. Research, v. 4, no. 2, p. 129-134, 1 fig., pi. 15. (10) Smith, J. Warren. 1911. Correlation. In Mo. Weather Rev., v. 39, no. 5, p. 792-795, 4 fig. ( 11 ) 1919. The effect of weather upon the yield of potatoes. In Potato Mag., v. 1, no. 10, p. 11-14; no. 11, p. 15-17; no. 12, p. 7, 16-17, 27 ; v. 2, no. 1, p. 16-17, 33-34. 23 fig. References, p. 34. (12) Verrill, Addison E. 1901-02. The Bermuda Islands. In Trans. Conn. Acad. Arts and Sci., v. 11, pt. 2, x p., p. 413-[957], 245 fig., pi. 65-104. Also separately issued, 1903. (13) Weiss, Freeman, and Orton, C. R. 1922. Progress notes on potato wart disease investigations. In Phytopathology, v. 22, no. — , p. (14) Wollenweber, H. W. 1920. Der Kartoffelschorf. Arb. Forsch. Inst. Kartoffelbau, Heft. 2, 102 p., 11 fig., 2 pi. (partly col.). PLATE I. PART OF THE TUBERS FROM EXPERIMENT I SHOWING THE INFLUENCE OF SOIL TEMPERATURE ON THE DEVELOPMENT OF POTATO SCAB. Note the severity of the infection on the tubers grown in the soil held near 24° C. and the tendency towards tuber elongation and the de- velopment of pear-shaped tubers at the higher temperatures. Ky w i r 14.5 ( i e c *t e -17— .002 3-23— .01 3- 30— .005 4- 27— .002 5- 2— .002 7-13— ..002 3-2 intra- venous y 2 of 24 hrs agar growth swine organisms 3-23 3 immature dead fetuses 10 4-27—0 4- 29—0 5- 3— .002 5-10 — .002 5-23— .002 4-27 intra- venous y 2 of 24 hrs agar growth swine organisms 4-20—100.8 4-25—101.8 4-27 — 101 . 5 4- 29—100.0 ! 5_ 2—100.6 5- 3—100.0 5-24 Bloody discharge containing hair, meat shreds, etc. 6 4-27—0 4- 20-0 5- 3— .002 5-10 — .002 5-23— .002 0-11— .002 7-1? — .002 4-27 intra- venous y 2 of 24 hrs. agar slope swine organisms 4-27—102. 4-28—101.4 1 4-29—104. 4- 30—102.4 5- 2—100. ! 5- 3—100. ? Was not seen to farrow, but prob- ably did 9 ! 4- 27- JQ 5- 3—0 5- 18- 0 5- 23— .002 0- 1— .002 6- 11— .002, 6- 23— .002 7- 18- .002 | 5-18 intra- venous y 2 of 24 hrs. agar slope bovine organisms 5-18—100.6 5-19—101. 5-20-101. 5-22—100.4 6- 8 Six normal pigs 8 4- 27—0 5 - 3 — 0 5— 18 — 0 5- 23— .002 6- 1— .002 6-11— .01 6- 23— .002 7- 18— .002 5-18 intrar venous y 2 of 24 hrs. agar slope bovine organisms 5-18—101. 1 5-19—101.4 5-20-100.6 5-22—100.6 6- 9 Nine normal pigs 5 4- 27—0 5- 3—0 5-18—0 5- 23—0 6- 1— .002 6-11—0 6- 23— .002 7- 18— .002 5-18 intra- venous y 2 of 24 hr. agar slope bovine organisms 5-18—101.2 1 5- 9—100.6 5-20—100.8 5-21—100.4 6-10 Three normal pigs 2 j 4- 27—0 5- 3—0 5-18 — 0 5- 23— .02 6- 1— .02 6-11— .02 6- 23 — .005 7- 18— .005 5-18 intra- venous y 2 of 24 hr. agar slope bovine organisms 5-18—101.2 5-19—100.6 5-20—102. 5-21—101. 7-14 Seven normal pigs 1-27 — 0 5- 18—0 i 6- 23 — 0 7- 18-0 Con- trol not inocu- lated i 6-7 j Seven normal pigs 8 Wisconsin Research Bulletin 55 TABLE I — Continued No . Dates of and of i reactions to Gilt blood tests Date inocu- lated Method of inoc- ! ulation Kind and amount of inoculum Dates and Date of 1 records of partu- temperatures rition Remarks 7 4-27—0 Con- 6-17 No evidences of a-18 — 0 trol not abortion infection 6-23-4) inocu- could be found in 7-18—0 lated the sis dead fetu- ses 4 4-27—0 Con- 6-12 .Five normal pigs 5-18 — 0 trol not 1 5-23 — 0 inocu- 6-23—0 lated 7-18—0 the four gilts inoculated with porcine strains. These four gilts all farrowed normally on the 21. 22, 23, and 57th day after having been inoculated. This was an average of 30.7 days which should be contrasted with the average of 24.7 days for the lot described in the preceding paragraph. Their pigs were alive and active at birth. Like the gilts in the other lot, they continued to react with agglutination test fluids, prepared from both porcine and bovine strains, until the experiment was brought to conclusion about three months after it was inaugurated. Gilts 12, 7, and 4 were controls on the two lots already men- tioned. They never showed evidence of infection by the agglu- tination tests. Two farrowed normally, while one farrowed six full-grown dead fetuses. Efforts to isolate abortion bacilli from the fetuses were not successful. This fact combined with the failure of the sow’s serum to react indicates that abortion infec- tion did not exist in the dam and that fetal death was due to some other cause. In order to dispel the thought that the bovine strains used in this experiment may have lost their pathogenicity and virulence, reference should be made to the data on the Guernsey heifer, mentioned below, which was inoculated with these identical strains. It should be stated in this connection that Connaway reports having succeeded in producing abortion in pregnant sows by feeding laboratory cultures from both bovine and porcine sources. It would have been better if the gilts inoculated with the bovine strains had not been so far advanced in pregnancy. In order to overcome this objection it is planned to repeat this experiment at the first opportunity. An Experimental Study of Infectious Abortion 9 Susceptibility of Heifers to Porcine and Bovine Strains Two heifers were bought for the purpose of determining the susceptibility of cattle to abortion bacilli of both porcine and bovine origin. These heifers had been bred shortly before they were purchased, but the exact breeding dates could not be ascer- tained. Their blood serum was tested and found to be free from all evidence of infection with the bovine abortion bacilli, as may be seen from Table II, to which reference should be made for results of subsequent serum tests. TABLE II— DATA ON HEIFERS INOCULATED WITH PORCINE AND BOVINE STRAINS OF B. ABORTUS. Name of Heifer Date Method inocu- of inocu- lated lation Kind of inocu- lum Date of abor- tion Incuba- tion period Agglutination tests (111 8-27 1 9-3 9-7 ! 9-9 1 1 J 9-11 9-121 12-16 1 1-9 1 ? 1 8 1 E d o\v 8-30-21 Intra- venous 4 porcine 10-27 strains 58 days^ 0 0 .002 1 1 ! ' .002 .002 l .0O5‘ .02 rnsey 9-7-21 Intra- venous 4 bovine 11-5 strains 59 daySj 0 0 0 6 1.005' .002 .005 .005 . 005 1 1 j: The solid yellow heifer was given an intravenous injection con- sisting of a suspension of pooled live porcine abortion bacilli representing the four different strains already mentioned. The dose was equivalent to the growth developed on one 24-hour pork-agar slope. The Guernsey heifer was treated in exactly the same way with the exception that four bovine strains of the or- ganism were used to inoculate her instead of the porcine strains. Within a few days after inoculation each heifer gave a clear- cut agglutination reaction, which may be interpreted to mean that she became infected, although a similar reaction would have oc- curred if killed cultures had been employed. Although a smaller quantity of blood-serum of the heifer inoculated with the porcine strains caused complete agglutination than was the case with her fellow, the agglutinins did not persist in this animal so long. These heifers aborted 58 and 59 days, respectively, after hav- ing been inoculated, so the incubation period of the disease was practically the same in each animal. In other words, one heifer was as susceptible to the abortion organisms recovered from 10 Wisconsin Research Bulletin 55 swine as the other was to the organisms recovered from cattle. Figure 2 is a reproduction of a photograph of the fetus aborted by the heifer that received the porcine abortion bacilli. Attempts were made to isolate B. abortus from the fetuses of both heifers by inoculating guinea pigs with stomach contents. The two guinea pigs injected with material from the solid yellow heifer’s fetus never reacted with the agglutination test and showed no lesions of infection with abortion bacilli when killed. The two guinea pigs, that were injected with stomach contents of the fetus aborted by the heifer which had been infected with bovine strains, both reacted several times to serum tests for spe- cific abortion bacilli antibodies. When these pigs were autopsied no visible lesions were noted and suitable culture media seeded with spleen pulp remained sterile. WAYS BY WHICH ABORTION IS CONTRACTED AND TRANSMITTED Bovine abortion has been produced experimentally in cows by introducing the bacterium abortus into the mouth, into the vagina, Fig. 2.— ABORTED FETUS OF EXPERIMENTALLY INFECTED COW The cow aborted 58 days after she had been injected intravenously with a suspension of abortion bacilli of porcine origin. An Experimental Study of Infectious Abortion 11 and into the udder. Therefore, it is reasonable to believe that under natural conditions cattle may become infected in these dif- ferent ways. The actual isolation of the infectious organism from aborted fetuses, infected placentas, and from the first milk (colos- trum) of sows, ’support the belief of experienced breeders of swine that sows acquire and transmit the infection in the same ways that cows do. ‘The following experiment was carried out to learn whether contaminated feed may be a source of infection for swine. Effects of Feeding Abortion Bacilli to Gilts A pen of four open gilts, designated in Table III as series 1A, were selected for this experiment. They were fed cultures of the aforementioned strains of abortion bacilli of porcine origin be- fore being bred. Previous to being so fed these gilts were tested and found free from any trace of infection. Reference should be made to the table for their subsequent serum reactions. All become reactors within two weeks after eating the abortion bacilli, but gilt 8 failed to react six w r eeks later and continued to be a negative reactor thereafter. A week to a month after eating this infectious material all had come in heat and had settled to one service by negative reacting boars. Gilt 5 farrowed four living pigs 105 days after conception ; gilt 6 five living pigs 113 days after conception; gilt 8 eight living pigs 115 days after conception; gilt 7 aborted five dead fetuses 79 days after being bred. Assuming that the cultures used in this experiment were virulent, as seems probable from results that followed their use in other experiments to be described, it is rea- sonable to conclude that these four sows became infected with the abortion bacilli taken into their digestive canals with food and drink. Since only one of these four actually aborted, it is evi- dent that normally gilts exhibit considerable resistance to, or have an appreciable amount of immunity against, porcine abortion bacilli when introduced per orem before conception has occurred. 12 Wisconsin Research Bulletin 55 Effects of Injecting the Bacilli Into the Blood Although infection seldom occurs under natural conditions di- rectly by way of the blood stream, it seemed desirable to attempt to produce it in this way, so that a check might be had on the vir- ulency of the cultures used in the feeding experiment just de- scribed, and to study any reactions which might develop in the experimentally infected swine. The animals used are designated in Table III as series IB. They were the four virgin gilts numbered 1, 2, 3, and 4 respec- tively. The data show that all gave negative agglutination re- actions before being inoculated and positive reactions eight days afterwards. They all continued to react for at least 18 weeks, with the exception of gilt 4, which ceased to react after the 15th week. This pen of gilts was inoculated intravenously before they were bred with a suspension of pooled cultures of porcine strains pre- pared in the manner already described. They were mated with negative reacting boars at the first heat period that subsequently occurred. It should be noted that gilt 1 was bred on November 25th and again on January 10th, after which date she appeared to have settled ; but on April 7th she was noticed to be in heat again. There is a decided probability that she aborted early in pregnancy, although she was not observed to have done so. This belief is based on the fact that she was a persistent reactor with the serum tests. Gilt 2 aborted on the 69th day of pregnancy and gilt 3 on the 83rd day, showing that the inoculum was virulent and that these gilts were susceptible to infection with porcine strains of abor- tion bacilli introduced intravenously. Gilt 4 farrowed seven lively pigs 116 days after conception occurred. This gilt may just as logically be said to demonstrate that the strains of porcine abortion bacilli used to infect her were avirulent or that she had a high resistance to the infection, or both. It is admittedly impossible to determine which of these two diametrically opposed statements is correct. The preponder- ance of evidence is in favor of the former, as one gilt either failed to conceive or aborted, two aborted, and only one carried her fetuses to maturity. TABLE III— COMPLETE RECORDS ON SWINE INOCULATED BY VARIOUS METHODS. An Experimental Study of Infectious Abortion 13 5-18 eupioa autAoa i 1 ° t ! 1 ! 1 ° 1 : ! 1 ! 1 i 1 : 1 1 l 1 o auioioj auiAog i i S J o i ] | i ! 1 1 l i l 1 1 n i • © § i ss & euiDioa g g i ■8 8 euiAoa § © © LO 1 8 8 - l T © s rH euioioa o o o o 1 « 3 s3 a+5 a O 3 -g ft « 3 ° 3s a O cultures TABLE III (Continued).— COMPLETE RECORDS ON SWINE INOCULATED BY VARIOUS METHODS. Wisconsin Research Bulletin 55 Agglutination tests 5-18 aupiOct euiAoa 1 - 1 i 1 i Mill! 1 : 1 i 1 i i i 1 M i © Si ei aupjOti 1 i II 1 i lilil: auiAog; .005 O © 1 1 1 1 l i I i i i i 9UIOJOJ to 8 © O © © © auiAog to o O L. 1 © 1 . 1 © l 1 © 1 © - 2-23 auiojoj 1 OJ O i- ;• 1 © 1 1 © | 1 1 © 1 © outAoa; 1 to 8 © i IV 1 © 1 e> L 2 auiojoj ! s i O .005 © © ~ . auiAoa 1 1 to 1 *■ 1^ I to |8 © S 1 • 1 ~ © rH auioiOcf 1 i 1 -i 1 : 1 i i i 1 i auiAog; 1 IS 1 © 1 © 1 • 1 ■ i • i § I • i 1 © 8 6 auiDjoj .002 § © O © auiAoa 1 1 1 CM 8 • 1 © © s aulDjOd § © 71 71 © 1 l© O i 9uiAog j (M g Ol 8 © © © i 1 © ! 2 1 rH 9U!DJO^[ | auiAog j i i i I'i 1 i ' O © ° i ° 1 ° ° Days preg- nant 2S O rH ! 1 LO s 1 • So to Date of farrowing and number in litter 3-8-22 3 weak pigs 4-1-22 7 pigs | ! ! i kl elipH 1-8-22 .4 dead pigs 04 04 ft ! 04 Pi 04 Date of breeding and boar used Jh (M a A o SPQ Method of inoculation and kind of inoculum Intravenous porcine cultures Intravenous porcine cultures Preputial porcine cultures 1 1 Date of in- ocula- tion 04 8 04 6 rH 6 2 1 Sex 1 * 1 § i I 1 1 ___ 1 No. of ani- mal 1 00 h* 00 I 04 1 7 1 Ci CO CO Series No. 2 | 1 O 1 ~ 1 | 1 1 1 TABLE III (Continued).— COMPLETE RECORDS ON SWINE INOCULATED BY VARIOUS METHODS. An Experimental Study of Infectious Abortion 15 Agglutination tests cc ) aupaoj | ©niAog 1 i 1 ! 1 : 1° 1 ! ! i 1 i i 4-10 eupjog | i • | ! | ! 1 i euiAog i i i i i 1 1 1 1 1 1 ! 1 J L 2-23 3-22 oupjLog; 1 ^ js 1 © 1 © 9U[AO,0 ® \% 1 1 © i 1 o OUIOJOJ o i I© 1 o _ eupoa h 8 O 1 © 1 o © ampjoa © © 1 1 © o aupog L o 0 © 11-29 aapjog | j ! 1 I i i aapog © © j o h . 10-28 eupiog © © .005 o eupog © © jl o 10-19 aupaog © i © 1 aupiog .002 © 1 i © © eurAog 1 1 © 1 L ! L « ! 90IMOJ i 1 i l-i 1 i 1-1 CJ) 3 9'JlAOa 03 O • Is s LO 8 § © 10-28 auioiog; S © 03 8 .002 i sutAoa © O i § js CT> s i oniDiog 1 "ir5 8 1 © |S © auiAog 1 .002 ? § s .005 10-11 anpiog 9UIAOg j 1 1 © ° 1° o © © >ays preg- nant • rH I; : 1 Date of farrowing and number in litter 2-6-22 aborted 5 dead pigs 2-23-22 aborted 4 dead pigs 4-18-22 7 pigs 3-11-22 6 pigs | • Did not farrow Date of breeding and boar used 11-2-21 11-22-21 Boar 21 12-0-21 Boar 21 12-28-21 Boar 22 12-22-21 Boar 21 11-15-21 Boar 21 10- 13-21 Boar 22 11- 15-21 and 12-9^21 Boars 21 & 22 Method of inoculation and kind of inoculum Subcutane- ous porcine cultures (killed) Subcutane- ous porcine cultures 1 (killed) Subcutane- ous porcine cultures (living) Subcutane- ! ous porcine cultures (living) Subcutane- ous porcine cultures (living) Date of in- ocula- tion Ol O Si 3 03 1 3 O? rH 3 rH 3 X 10 188 1.38 1.43 96.5 97.1 July Alternate leaves off — off node 1 4 79 1.71 1.61 106.2 100.0 24 Alternate leaves off— on node 1 4 78 1.67 1.61 103.8 100.0 10 Wisconsin Research Bulletin 56 Table VI. — Summary Plum Defoliation Data, 1919. (Transposition of Table V.) Treatment Date Number trees Number branches Nodes Buds check Percentage normal number buds Percentage growth at time of defoliation No. Av. buds Alternate leaf off— off node June 16-17 | 7 27 303 i .12 ! 1.45 8.3 56.3 June 25-26 ! 7 20 320 i .47 1.45 32.4 77.8 July 1- 3 | 3 17 302 1.16 1.77 65.6 92.3 July 10 7 11 204 1.16 1.45 1 79.9 95.3 July 24 1 4 79 1 1.71 1.61 106.2 100.0 Alternate leaf off— on node June 16-17 7 27 302 | 1.08 1.45 ; 74.5 56.3 June 25-26 7 20 320 1 1.55 1.45 106.9 77.8 July 1- 3 3 17 301 i 1.76 1.77 i 99.3 92.3 July 10 7 11 201 1.38 1.45 ; 95.2 95.3 July 24 1 4 78 1 1.67 i 1.61 ! 103.8 I 100.0 One-third leaf blade off... June 16-17 5 7 153 .97 : 1.38 70.3 59.7 June 25-26 5 6 174 ; 1.16 1.38 ! 84.0 76.5 July 1- 3 1 1 1 42 1 2.15 2.05 ; 104.8 88.6 July 10 3 4 167 1.57 1.76 1 89.2 90.4 One-half leaf blade off June 16-17 1 5 13 272 1 .56 1.38 40.6 58.8 June 25-26 5 7 199 1.06 i 1.38 76.8 77.0 July 1- 3 1 1 40 1.82 2.05 ; 88.8 88.1 July 10 3 4 153 1.38 i 1.76 | 78.4 94.2 Two-thirds leaf blade off June 16-17 5 7 155 .24 1.38 17.4 63.1 June 25-26 5 6 172 ; .59 | 1.38 ; 42.7 86.0 July 1- 3 1 1 42 ! 1.64 2.05 80.0 84.5 July 10 8 4 157 1 1.43 1.76 81.2 94.4 Midrib cut at center.. June 16-17 4 4 71 .81 j 1.41 57.5 51.5 June 25-26 1 1 25 ! 2.28 2.05 111.2 ; 81.2 Alternate midrib cut at base— off-- June 16-17 4 5 51 .47 1.49 31.6 58.3 June 25-26 i 5 6 92 1.17 | 1.38 94.8 83.5 Alternate midrib cut at base— on. . June 16-17 i 4 5 51 1.36 ! 1.49 91.2 58.3 June 25-26 5 6 92 1.47 1.38 106.6 83.5 Alternate 2 leaves off— off node June 25-26 3 49 • 49 1.57 31.2 ! 78.6 July 1- 3 ! i 1 18 1.39 2.05 ! 67.8 91.8 July 10 ! 6 10 190 1.11 1.43 77.7 97.1 Alternate 2 leaves off— on node June 25-26 3 3 50 1.45 1.57 92.3 78.6 July 1- 3 1 1 18 2.22 2.05 1 108.2 91.8 July 10 ! 6 10 18 8 1.38 1.43 96.5 97.1 Alternate 3 leaves off— off node— June 16-17 I 5 6 69 .24 1.40 17.1 64.2 June 25-26 ! 2 2 36 .39 1.52 25.7 78.2 Alternate 3 leaves off— on node— June 16-17 5 6 ' 69 1.21 1.40 1 86.3 , 64.2 June 25-26 2 2 36 1.72 1.52 113.2 1 78.2 Leaves off lower half branch— off.. June 16-17 4 4 42 .07 1.41 5.0 i 61.2 June 25-26 3 3 36 .58 1.52 38.1 ! 83.2 Leaves off lower half branch — on__ June 16-17 4 4 43 1.03 1.41 : 73.0 | 61.2 June 25-26 3 3 36 1.59 1.52 104.6 1 83.2 Leaves off upper half branch — off_. June 16-17 1 3 3 33 .46 1.53 [ 30.4 ; 77.8 June 25-26 3 3 39 .87 1.52 1 57.3 83.5 Leaves off upper half branch— on. . June 16-17 3 3 31 1.34 1.53 1 87.6 77.8 June 25-26 3 3 35 1.75 1.52 1 115.1 83.5 Right side leaf blade off— off June 16-17 5 5 97 .30 .67 i 44.8 63.8 June 25-26 1 5 6 i 168 .60 .71 1 I 84.5 77.4 July 1- 3 1 2 70 1.13 1.03 i 109.7 95.0 Right side leaf blade off— on June 16-17 ! 5 5 97 .56 .67 1 83.6 63.8 June 25-26 1 5 6 168 .65 .71 1 91.6 77.4 July 1- 3 1 2 69 1.07 1.03 103.8 95.0 Left side leaf blade off— off June 16-17 1 3 5 103 .19 .69 27.6 58.1 June 25-26 1 1 1 29 1.00 1.03 ! 97.1 84.8 July 1- 3 ! l 1 32 .88 1.03 85.4 100.0 Left side leaf blade off— on... June 16-17 1 3 5 101 .30 .69 43.5 58.1 June 25-26 ! i 1 29 1.17 1.03 113.6 84.8 July 1- 3 i 1 32 .97 1.03 94.2 100.0 Effect of Defoliation Upon Blossom Bud 11 It will be noted that in 1919 the defoliation work did not completely inhibit blossom bud development as was true in 1917. This is not because of differentiation having been in- itiated before the leaves were removed as probably occurred in 1918, but because there was a marked second period of bud formation. It was noticed August 8, 1919 that some buds were forming at defoliated nodes and that a third and even a fourth bud had appeared on some untreated nodes. These were clearly the result of a second period of differentiation. Prior to this time there had been no blossom buds at the nodes which were defoliated early in the season. There was again, as in 1918, a marked relation between the removal of a fractional part of the leaf blades and the number of blossom buds formed. Cutting one-third off the leaf blades on June 16 reduced the formation of buds to 70.3 per cent of that on check branches, taking one-half of the blade off gave 40.6 per cent and removing two-thirds of the blade resulted in but 17.4 per cent of the usual number of buds. The influence of removing the leaves upon bud formation became less and less as the season progressed. Defoliation after terminal growth was completed seemed to have little effect upon bud formation. This may probably be more in- cidental than of causal significance, however, as blossom bud differentiation on short spurs is certainly not as early as the cessation of terminal growth, although it is somewhat earlier than on the longer more vegetative growths. Results of 1920 Tests Alternate leaves were removed from a number of branches June 18 and 19, 1920. 4 This resulted in practically complete in- hibition of blossom buds. Exclusive of two limbs (branches number 32 and 33) which had practically completed their growth at the time of defoliation, there was but one bud formed at 307 defoliated nodes as compared to 280 buds at 300 nodes of untreated branches. The data are given in Table VII. 4 A. L. Schrader of the Department of Horticulture carried on this work. 12 Wisconsin Research Bulletin 56 Table VII . — Plum Defoliation Data, 1920. Alternate Leaves Removed June 18, 1920 Tree | Branch number | ON OFF GROWTH inches Branch number \ CHECK GROWTH inches | Nodes Buds Average Nodes Buds Average Length when defoliated Total for season Pctge. when defoliated 03 V 'O o Buds o fcfl c3 f-t 0> > < Length when defoliated Total for season Pctge. when defoliated 1 4-14 15 8 4 .5 8 0 .0 15.75 25.0 13.0 17 It | 6 .4 ! 14.75 23.0 63.1 16 8 2 .25 8 0 .0 15.5 28.0 55.3 21 21 7 i .35 16.5 25.0 65.9 18 10 8 .8 10 0 .0 23.0 41.5 55.4 19 10 7 .7 10 0 .0 17.0 32.5 52.3 i 20 9 6 .67 9 0 .0 17.5 28.25 32.0 22 10 4 .4 10 0 .0 18.25 24.0 76.0 — 6-10 24 12 18 1.5 lz 0 .0 18.5 31.5 58.7 25 25 39 1.56 16.25 26.5 61.3 26 13 15 1.15 13 0 .0 21.25 41.25 51.5 28 25 33 1.32 18.5 33.75 54.8 27 12 14 1.17 12 0 .0 19.75 34.0 58.1 31 25 32 1.28 15.5 28.5 54.4 29 13 18 1.38 13 0 .0 19.0 30.25 32.8 34 15 ! 17 1.13 9.25 19.25 48.2 30 12 17 1.42 12 1* .08 17.5 26.75 36.5 38 20 ] 32 1.6 16.0 26.5 60.4 32 9 13 1.44 9 4*1 .44 13.75 14.5 94.8 1 .. 33 8 12 1.5 8 4*| .5 11.0 11.5 95.7 36 12 16 1.33 12 ! o .0 19.25 28.75 36.9 ! 37 10 15 1.5 lOj 0 1 .0 17.0 32.0 53.1 7-15 39 11 5 .45 11 0 .0 15.5 41 20 12 .6 12.75 23 O 55 5 40 8 2 .25 8 0 .0 11.5 23.75 48.4 50 15 6 .4 12.75 19.75 64.6 42 9 0 .0 9 0 .0 16.5 24.0 68.7 48 9 2 .22 9 0 .0 13.25 18.0 73.3 :::: 9- 1 51 8 3 .38 8 0 .0 12.0 26.0 46.2 53 20 9 .45 11.0 20.5 53.6 52 9 6 .67 9 0 .0 11.5 23.25 49.4 55 15 13 .87 9.75 21.0 46.5 54 8 4 .5 8 0 .0 10.0 19.0 52.6 57 15 15 1.0 11.25 23.25 48.4 56 8 ! 2 .25 8 0 .0 10.25 23.5 43.6 61 20 17 .85 9.75 27.0 36.1 58 8 7 .88 8 0 .0 9.75 23.25 41.8 59 9 i 7 .78 9 0 .0 12.0 19.5 61.5 60 9 5 .56 9 0 .0 12.0 20.0 60:0 62 9 ! 3 .33 1 9 0 .0 12.0 20.0 60:0 L 9- 3 64 9 5 .56 9 0 .0 12.0 22.0 54.5 63 15 12 .8 10.0 19.75 50:6 65 8 5 .63 8 0 .0 9.75 18.0 54.2 68 15 7 .47 13.5 23.5 58.0 66 8 5 .63 8 0 .0 11.0 18.5 59.4 73 20 23 1.15 12.0 25.5 47.1 67 10 3 .3 10 ! 0 .0 12.75 22.75 56.0 69 10! 12 1.2 10 0 .0 15.0 38.5 38.9 72 9 ; 6 .67 9 0 .0 12.5 27.25 45.8 74 9 6 .67 9 0 .0 15.5 25.0 62.0 Total 1 4-14 6 55! 31 .56 55 0 .0 107.0 179.25 59.7 2 35 13 .37 31.25 48.0 65.1 6-10 9 1011138 1.37 101 9 .09 157.0 250.5 62.7 5110 153 1.39 75.5 134.5 156.2 7-15 4 37! 9 .24 371 0 .0 41.25 65.75 62.7 2 3 b 18 .52 25.5 41.75 61.1 9- 1 8 68 37 .54 68 l o .0 89.5 174.5 51.3 4 70 54 .79 41.75 91.75 45.5 9 - 3 7 63 42 .67 63 0 .0 86.5 172.0 50.3 3 50 42 .84 35.5 68.75 ,51.7 Total— 34 324 257 324 9* 481.25 842.0 57.2 16 300 280 .933 209.5 384.75 55.4 Av .794 7 028 * Small buds. Results of 1921 Tests Alternate leaves were removed from about 200 branches on June 15 and 16, 1921 in order to secure material for chemical analyses. Also three quarters of the blades of all leaves were removed from 20 additional branches on June 18. The results of these treatments upon blossom bud formation are shown in Table VIII. Effect of Defoliation Upon Blossom Bud 13 It will be noted that : 1. Data collected on July 7 showed almost complete inhibition of bud formation. The formation of buds at defoliated nodes took place at a second period of differen- tiation. At this time there seemed to be a stronger ten- dency for late buds to form at the alternate untreated nodes of the partly defoliated branches than on the check branches. 2. The growth of the branches was approximately 7 5 per cent complete on June 15-16 as compared to 55 per cent on June 18-19, 1920. This early cessation of growth may have a direct relation to the abundant formation of buds at the second period. Most of the buds on the branches with three-fourths of the leaf blade removed were near the base of the branch and had probably started to differentiate before defoliation was done. Table VIII. — Summary, Plum Defoliation. June 15-18, 1921. I Tree Treatment Date Number branches ON OFF Percentage normal Percentage | growth when treated Nodes Buds Av. i Nodes Buds* Av. “On” node “Off” node 6-10 j Alt . leaves off i 7- 7 41 549 824 1.5 555 39 .07 82.4 3.8 74.9 (Alt. leaves off 10-14 ! 73 904 1,695 1.87 91C 615 .68 102.8 37.4 78.4 1 Check 10-14 28 694 1,264 1.82 100.0 67.3 1- 7 Alt. leaves off 10-13 79 941 1,499 1.59 953 294 .34 89.8 19.2 ' 74.2 Check 10-13 18 418 741 1.77 100.0 66.6 4-15 ; 4 blade off 10-14 21 661 228 .35 32.1 70.8 Check ! 10-14 6 180 196 1.0S 16676 | 66.6 * All buds at defoliated nodes were noticeably smaller than those at non-defoliated nodes. It was observed that there was apparently quite a difference in the growth conditions of branches that had many and few blossom buds, especially at the defoliated nodes. This differ- ence is expressed in Table IX., and shows that the branches forming many buds were not only more nearly through grow- ing at the time of defoliation but also made less total growth. While the longer growing period of the latter probably re- duced the tendency to. blossom bud formation as evidenced by the development at the “on” nodes, there were probably, also, some buds on the shorter growths which were beginning to 14 Wisconsin Research Bulletin 56 differentiate at the time of defoliation as the period of for- mation is somewhat earlier on shorter than on longer growths. Table IX. — Compartsion op Growth of Branches Forming Many and Few Blossom Buds at Defoliated Nodes. Type Number branches Average number buds Percentage of total growth ; at defoliation! Average total length “On” node|“Off” node Many buds at defoliated nodes.. No buds at defoliated nodes 31 37 2.11 | 1.22 1.32 1 0.0 86.4 68.5 59.7 cm. ! 72.5 Effect of Defoliation Upon Blossom Bud 15 Summary- Some of the results secured in the course of the defoliation studies follow : 1. Removal of all or a part of the leaf blade had a marked inhibiting effect upon blossom bud formation if done early in the season. Buds which formed at defoliated nodes were noticeably smaller than the average. 2. When alternate leaves were removed the number of buds formed at undefoliated nodes was somewhat reduced. While the effect of defoliation was largely localized at the node treated, it would appear from this that there was also a mass or cumulative influence upon bud formation. This is also indicated by the formation of buds at defoliated nodes during the second period of differentiation. 3. Cutting the midribs transversely near the base of the blade had a marked inhibiting influence upon bud develop- ment, Fig 5, G. 4. Removal of different quantitative fractions of the blade had a marked direct quantitative effect upon blossom bud for- mation. 5. Removing one-half of the leaf blade by a cut parallel to the midrib reduced bud development to. about the same de- gree as removing half of the leaf by cutting completely across it transversely. 6. The spurs developing from the defoliated nodes are much smaller and shorter than from the other nodes. It is suggested that the difference in nitrogen reserve is a large factor in giving this condition. 30-7 / 75's / ljLS . ' r . / r. . Research Bulletin 57 October, 1923 ^SHTBFJLUSCSS > iNQV ^ 3 1923 The Fishy Flavor in Butter H. H. Sommer and B. J. Smit Agricultural Experiment Station of the University of Wisconsin Madison Contents Page Occurrence and prevalence of fishiness . 1 Review of the literature 2 Summary of the historical review 8 General plan of the experimental work 9 A study of the conditions that favor fishiness 9 Experimental .. 11-12 Review of literature 13 Acidity and fishiness 13* Salt and fishiness „ 14 Overworking and fishiness IS Metals and fishiness 16 Pasteurization and fishiness 16 Summary of experiments and literature ..^ 17 A study of the cause of fishiness in storage butter ..... 17 Amount of lecithin in dairy products 18 A study of the conditions under which lecithin yields trimethylamine 19 Trimethylamine determination 20 Preparation of lecithin 21 Experimental ~. — 21 Trimethylamine and fishiness ......; 24 Trimethylamine in butter 26 Fishiness in lecithin-added butter 27 Biological agencies and fishiness I....... 29 Decomposition of lecithin by bacteria 30 The production of trimethylamine from skimmilk and casein . 31 The production of trimethylamine from hydrolyzed and un- M hydrolyzed lecithin 34 General discussion . 38 The role of the various factors concerned in the development of fishiness 40 Summary 45 Bibliography ... 47 The Fishy Flavor in Butter H. H. Sommer and B. J. Smit* W HEN BUTTER is fishy, a distinct fish odor and flavor can be detected which is usually described as resembling that of mack- erel, salmon, or herring. Fishiness in butter is detected by even the most casual consumer. Butter is scored mainly on flavor, and when an off-flavor as offensive as a fishy flavor is present, the score is at once reduced from “extra” to a “second” or a “poor second” with an accompanying reduction in the price. This reduction may range from 3 to 5 cents per pound so that on a 1,000 pound churning the loss may amount to something between thirty and fifty dollars. Occurrence and Prevalence of Fishiness Fishiness is hardly ever found in fresh or comparatively fresh butter. It is primarily a storage defect, occurring in cold storage and export butter. On the basis of the score of the fresh butter it is impossible to predict which samples will become fishy, as some of the highest scoring fresh butter may become fishy during storage. It may occur at temperatures as low as — 15° F., or at times it appears after the butter is taken out of storage and is in the consumers’ hands. — — How to Prevent Fishiness 1. Avoid making butter from high acid cream. Use swdet cream or reduce the acidity of the cream by neutralizing. 2. Avoid excessive salting. 3. Avoid overworking. 4. Do not allow the cream, especially sour cream, to come into contact with poorly tinned iron or copper utensils. 5. Pasteurize the cream preferably at 145 °F. for 30 minutes. Fishiness is found in butter exported from Australia, New Zealand, and South Africa, and it is known to occur in various parts of the United States. There are creameries where the fishy flavor appears year after year ; among these are some that are regarded as using the most approved machinery and methods, and are managed by competent butter makers. In Australia the prevalence of fishiness has been especially marked since a large part of the output is exported over a long distance. It has been * This experimental work was done by B. J. Smit under the direction of H. H. Sommer. It was submitted by Mr. Smit in fulfillment of the thesis re- quirements for the degree of Ph.D. from the University of Wisconsin. The manuscript was prepared by Mr. Sommer. 2 Wisconsin Research Bulletin 57 estimated that this off -flavor in export butter has reduced the value of this colonial butter on the English market by thousands of dollars annually, causing losses that amount for a period of ten years to over $5,000,000. On this account this flavor is known in Australia as “Australia’s costly taint.” Similarly the South African export butter is suffering losses on the English market. About 75 per cent of the butter that deteriorates markedly during storage shows signs of fishiness; it is the most common of the storage flavors. Some writers are of the opinion that fishy flavor causes greater losses in butter than any other one defect. Review of the Literature The first reference to this class of flavors in butter is that of Storch ( 1890) ^ He attributes “oily,” “fishy,” and bitter flavors to the practice of ripening cream before churning. In oily butter Storch always found large numbers of certain bacteria present which were absent in non-oily butter. However, he could not produce oily butter by inoculation with cultures of these organisms. Kirchner ( 1891 ) 2 notes the presence of fishy and oily flavors in old butter, and attributes them partly to the feeding of oil cakes and partly to abnormal decomposition of the butter fat by bacteria that entered the cream through faulty handling. Weigmann (1891, 1892) 3 , 4 attributed this class of flavors to iron dis- solved in the cream from poorly tinned containers. He produced this class of flavors by adding small amounts of iron lactate to sour cream. Oliver ( 1894) 6 attributed fishiness in butter to the absorption of the odor from fish kept near it. O’Callaghan ( 1899) 6 believes that Oidium lactis is the cause. duRoi ( 1900) 7 attributes it to the litter collected from the woods. Harding, Rogers, and Smith ( 1900) 8 report a highly disagreeable fishy flavor in milk from a healthy cow. Inoculating other milk with bacteria from the fishy milk failed to reproduce fishy milk. Piffard ( 1901 ) 9 suggested the ability of the salt to absorb odors and flavors as an occasional cause of fishiness. He also suggested that the un- desirable, often fishy odor of water to which the cows have access, might be imparted to the milk and cream. No experimental data are given. Harrison ( 1901 ) 10 and his staff in their study of the effect of bacteria commonly found in milk on the flavor of butter, separated some twenty dif- ferent species, made starters, and inoculated pasteurized cream with them. Among other flavors they were able to obtain a fishy flavor. Thus Harrison attributes the development of fishiness to the presence and growth of unde- sirable bacteria in the cream. O’Callaghan (1901)^ reports cases of fishiness in fresh butter made from sour cream. Examination of the butter showed the presence of large numbers of Oidium lactis spores. He also found during two years obser- The Fishy Flavor in Butter 3 vation that, of a large number of fishy samples of butter examined, all contained Oidium lactis. As a remedy he advocates pasteurization at 168° F. or above. Willoughby (1903) 12 states that in Norway where the cows are fed to a large extent on fish and fish meal, the milk may acquire a fishy flavor if the fish used are not entirely fresh. O’Callaghan (1904) 13 gives the cause of fishy butter as Oidium lactis working jointly with the ordinary organisms that cause the souring of milk. Gray and McKay ( 1906) 14 found a pronounced fishy flavor in butter made from old cream and salted high. High salting did not cause a fishy flavor in butter made from sweet cream. O’Callaghan (1907) l5 states that in all cases where butter has been re- ported on as fishy on its arrival in London, the bacteriological records show that the butter from the factory has been marked as containing numer- ours spores of Oidium lactis. He also states that the salt cannot be the cause of fishiness because many a consignment of unsalted butter has turned out fishy in London. Kirchner (1907) 16 thinks that tallowy and fishy butter result from the further decomposition of oily butter, and that these defects are produced by acid forming bacteria and molds or by fat splitting enzymes of the milk. He thinks that pasteurization, the use of a pure culture, and the careful handling of the milk are sure remedies. Sommerfeld ( 1909) 17 contends that most of the butter faults have their origin in the souring of cream. The oily taste of butter, this writer has found, is often accompanied by a fishy taste. He states that the fishy taste apparently is due to the formation of trimethylamine. According to his experience, fishy flavor results when the cattle graze on meadows that are frequently flooded, and also when the butter salt is somewhat high in magnesium salts. Rogers (1909) lg is not of the opinion that trimethylamine is the imme- diate cause of fishiness, nor that Oidium lactis is the general cause. He states that many lots of fishy butter have been made in which Oidium lactis has been known to be absent both in the butter and the cream. Bacteriologi- cal inoculations with bacteria from creameries where fishy butter was pro- duced failed to produce fishy cream. Bacteriological examinations of the milk from farms where fishy butter could be made revealed no unusual varieties of bacteria connected with the production of the off -flavor, neither did a botanical survey of the farms as to water supply and the flora of the grass in the pastures reveal any clue. Rogers, however, found that in all cases where butter became fishy, it had been made from cream high in acid, either developed naturally or added artificially, though high acid cream did not develop fishiness uni- formly. He further finds that overworking of butter produced fishiness with a reasonable degree of certainty. In such butter the oxygen content decreased materially during storage. He also shows that butter made from sweet cream does not become fishy, and that pasteurized cream with starter 4 Wisconsin Research Bulletin 57 added but without ripening, seldom, if ever, becomes fishy. He offers the opinion that the fishy flavor is caused by a slow spontaneous, chemical change to which acid is essential and which is flavored by the presence of small amounts of oxygen. Rahn, Brown, and Smith ( 1909) 19 think that, since butter made from high acid cream frequently shows no trace of fishiness, the development of the flavor cannot be explained by Rogers’ chemical theory alone, because they think in all sour cream butters all the factors concerned are present and would therefore expect fishiness to appear uniformly in all such butters. Thomson (1911, 1913) 20 , 21 sums up his work and observations on fishi- ness in butter over a period of ten years (1898-1908) as follows: — Fishiness was found to be very pronounced in butter manufactured in districts where the herbage was rank and immature. In districts where the herbage was well matured only a faint indication of fishiness was found. Nearly all the fishy butter was manufactured from cream ripened to a high degree of acidity. All the butter was found to be free from fishiness before refrigeration. The fishy flavor appeared to be of equal strength throughout the affected butter. Butter made from one churning, a portion of which was brined and dry salted, was unaffected with fishiness after refrigeration, while the other portion, that was dry salted only, developed a fishy flavor. Preservatives appeared to check fishiness. Cultivations made from very fishy butter did not give colonies productive of fishiness. Butter stored at the temperature of 25 to 32° F. was very much more fishy in flavor than a second piece of the same butter stored at 5° F. A fishy flavor will develop in both salted and unsalted butter, but more commonly in the salted product. A high degree of bacterial purity in the products used in the manufacture of butter did not prevent the butter from developing a strong fishy' flavor. Although fishiness is much less prevalent in pasteurized butter, the flavor may be found in it in a very pronounced form. Butters containing from 14 to 16 per cent moisture appeared to be more susceptible to the taint. Fishiness was pronounced in butter made from cream containing colostrum. The writer concludes that fishiness in butter is largely a chemical change and that the production of the flavor is favored by a high degree of acidity and salt and a temperature of storage between 30 and 40° F. Weigmann (1911) 00 attributes the fishy flavor to the abnormal working of butter and also to the use of salt high in magnesium. Davis (1911, 1912) 23 , 24 found that the samples of butter which scored fishy were made from cream high in acidity and salted over two and one- half per cent. The Fishy Flavor in Butter 5 Rogers, Thompson and Kiethley ( 1912) 25 demonstrated more conclusively still the freedom from fishiness in butter made from unripened cream and the tendency of butter made from ripened raw or pasteurized cream to become fishy in storage. Reakes, Cuddie and Reid ( 1912) 2Q found no significant differences in the bacterial flora of fishy and high grade butter. Plugs of fishy butter put into high grade butter failed to cause fishiness, and plugs of high quality butter inserted into tubs of fishy butter retained their desirable flavor. These investigators concur with Rogers lg in his chemical decomposition theory. O’Callaghan ( 1912) 27 maintains that Oidium lactis working in conjunction with the ordinary lactic acid producing bacteria is the real cause of fishiness in butter. Steinhoff ( 1913 ) 2g states that fishy flavor is one of the most common flavors in, storage butter. He cites one case of a large creamery turning out uniformly fine butter when fresh, but when placed in storage this butter developed fishiness three to five years in succession. Rogers, Berg, Potteiger and Davis (1913) og showed that iron and copper salts have a decided effect in causing the development of fishiness. They showed that enough iron and copper could be absorbed by the sour cream from rusty cans and exposed copper linings of vats to cause a marked change in the flavor of the butter. Orla-Jensen ( 1912) 3Q states that certain yeasts can give butter a fishy flavor. Rogers ( 1914) 31 points out that an oily and metallic flavor may precede the fishy flavor. He excludes the theory of the bacterial origin of fishiness, because of the high salt concentration and the low temperatures at which the flavor develops, and because preservatives did not prevent it in his work. He excludes enzymes as a possible cause on the grounds that high pasteur- izing temperatures do not necessarily prevent fishiness. Synder 31 thinks that oversalting develops the fishy flavor in butter. Klein ( 1914) 33 thinks that the class of oily flavors in butter results from the action of bacteria. Ernst (1914) 34 believes that milk may become fishy from feeding fish meal to the cows and through pasturing cows on marshy fields which have been inundated. Rogers ( 1914) s5 states that fishiness rarely occurs in unsalted butter, and that the salt possibly furnishes certain conditions which are essential to the development of the flavor. He was unable to produce fishiness under winter conditions. Lewkowitsch ( 1914) 36 reports that fishy butter is met with in Norway, being obtained from the milk of cows fed on fishmeal. Hunziker (1915) 37 has found that extremely "high pasteurizing tempera- tures, such as 185° F. and higher, when used on sour cream, may cause poor butter with a disagreeable oily taste suggestive of fishiness. He found this especially true in summer when the cows are on green pasture. 6 Wisconsin Research Bulletin 5 7 Fleischmann ( 1915) 3Q states that fishy and oily flavors appear only in sour cream butter, and that they are caused by bacteria. Dyer ( 1916) 39 concludes that fishiness is caused by the slow oxidation of one or more of the non- fatty constituents of butter. He found that the decrease in the oxygen in the butter and the development of fishiness were parallel and were directly proportional to the quantity of acid present. O’Callaghan ( 1916) 40 still maintains that Oidium lactis is the responsible factor and that pasteurization and disinfection of creameries, stables, etc., are the only remedies. O’ Callaghan (1916) 41 has found that 75 per cent of the butters that deteriorate to any extent show an advanced fishy flavor at the end of about 3 months in storage. All such butter had been made from unpasteurized cream. He cites an instance where pasteurized and unpasteurized butter was stored by a certain company, and only the unpasteurized butter became fishy. Washburn and Dahlberg (1917) 42 found that salt hastened the deteriora- tion of butter at low temperatures of storage, that salted butter was more likely to turn fishy during storage than unsalted butter, and that there appeared to be a tendency toward a progressive development of the flavor through metallic to oily and finally to the fishy flavor. Hammer (1917) 43 reports the isolation of an organism from a can of fishy evaporated milk. He was able to reproduce the fishy flavor in milk, cream, and in evaporated milk, but he was unable to produce it in butter by direct inoculation or by inoculating the cream before churning. He con- cludes, that the organism is not of direct importance in the development of fishiness in butter since the organism is unable to grow in sour cream butter or in salted sweet cream butter. Klein (1917) 44 thinks that a fishy taste in milk may result from contact with rusty vessels or from imperfect rinsing of soap powders from the vessels. He is of the opinion that the flavor in the milk is not due to feed such as fishmeal or marshy pastures.. Ericson ( 1918) 45 finds that the fishy flavor is more prevalent in butter made from ripened cream than in sweet cream butter. He believes, how- ever, that the starter is not the cause if it is used correctly. Washburn and Holmes ( 1918) 46 did not find any fishy butter when it was made from cream pasteurized sweet and starter worked directly into the butter. Bouska and Washburn ( 1918) 47 have experienced that cream used in buttermaking must not be very sour otherwise the butter will often become fishy in storage. Rogers (1919) 4S says that in over 5,000,000 pounds of Navy butter made under the supervision of the Dairy Division from sweet cream of good quality, pasteurized and churned sweet without starter, not a single pound developed fishiness in six to eight months storage. Washburn ( 1919) 4n states that the most annoying problems of the butter industry are the metallic and fishy flavors. He states that fishiness is related The Fishy Flavor in Butter 7 to old and sour cream and to rusty cans, but that the exact cause is not known. Supplee (1919) 50 was able to extract trimethylamine from fishy butter; and by working trimethylamine salts of fatty acids into normal butter in small amounts he was able to produce the fishy flavor. He thus considers trimethylamine as the real cause, and lecithin, which he was able to isolate from butter, as the source of the trimethylamine. He found fishiness most frequently in salted butter made from ripened or artificially acidified cream. He found fishiness less prevalent in the pasteurized samples, and concludes from this that a biological factor is the fundamental cause. Hunziker ( 1920) 5l agrees with the findings of Rogers and Dyer, that high acid cream, overworking and the presence of metallic catalyzers are important factors in the development of fishiness. McKay (1920) 52 cites an instance to show that butter made from pasteur- ized and neutralized cream can be kept two years without any signs of fishi- ness, while unpasteurized and unneutralized cream butter became fishy. He is of the opinion that small hand separators not properly cleaned during extremely hot weather may be responsible for the fishy flavor. Brown ( 1920) 5g advocates the pasteurization and neutralization of the cream to eradicate fishiness in butter. He cites an instance of a factory storing two consignments of butter, one was pasteurized, and one unpas- teurized. The unpasteurized butter became fishy after 4 to 5 months, while the pasteurized butter showed no trace of the flavor. McKay (1920 ) 54 states that he receives reports reading essentially as follows : — “Made two churnings of butter from the same vat of cream : one churning has developed fishiness, the other churning is all right.” McKay concludes that the difference cannot be attributed to organisms, but to the method of manufacture. He attributes it to overworking. Cusick ( 1920) 55 studied the phosphorus compounds in cream and butter. He showed that pasteurizing acid cream caused a lower phosphorus content in the butter; in sweet cream, pasteurization had no such effect. He also showed that the organic phosphorus, which includes lecithin, is changed to inorganic phosphorus during storage, and that this change is more rapid in salted than in unsalted butter. He states that in the development of fishiness there is an appreciable loss of organic phosphorus^ (lecithin). He concludes that the breaking down of lecithin through the solvent action of salt water and lactic acid, gives off trimethylamine, which causes the fishy flavor. Brown, Smith and Ruehle (1920) 56 found that fishiness develops more readily in unpasteurized butter, and conclude from that, that a biological agent is concerned in the development of fishiness. They were unable, how- ever, to isolate any bacteria directly responsible for the off flavor. Cusick (1920) 57 was able to produce fishy butter from cream inoculated with pure cultures of Bad. ichthyosmius. He found no fishiness in the unsalted inoculated samples. During storage the bacteria decreased rapidly in both the salted and unsalted samples, but in the salted samples that 8 Wisconsin Research Bulletin 5 7 were fishy the decrease was only gradual. Cusick thinks that lecithin is broken down to a form that is used as a pabulum by bacteria. Hamilton ( 1921 ) 58 attributes fishiness in butter to a variation in the cold storage temperatures from day to day. Hardy ( 1921 ) 59 regards pasteurization as the most effective preventive of fishiness. McKay and Larsen ( 1922) 60 state that fishy flavor in butter causes greater losses in butter than any other one defect. They sfate that butter from high acid cream invariably becomes fishy in storage, and recommend partial neutralization of sour cream. Fryhofer (1922) gl believes that fishy flavor has its origin in various factors or a combination of factors. One of the factors on which there is general agreement is high acid. Summary of the Historical Review A review of the literature reveals the fact that there are a number of theories, hypotheses and opinions advanced to explain the occurrence and development of the fishy flavor in butter. Many of these opinions are not backed by any definite and exact experimental evidence. A number of miscellaneous causes have been suggested in the literature. A number of writers have attributed the fishy flavor to the feed, e. g. oil cakes (Kirchner 2 ), fish meal (Willoughby 12 , Ernst 34 , Lewkowitsch 36 ), marshy pastures (Ernst g4 , Sommerfeld 17 ). Other causes that are mentioned are : — The absorption of the odor from fish stored near the butter (01iver 5 ) ; derived from litter collected from the woods (duRoi 7 ) ; derived from the salt which may have absorbed the odor (Piffard 9 ) ; de- rived from stagnant drinking water having a fishy odor (Piffard 9 ) ; result- ing from the imperfect rinsing of soap powders from the vessels (Klein ). These suggested causes cannot be considered as general causes, and in most cases it is doubtful whether they can cause fishiness even in specific instances. The two main theories for the development of fishiness are, first, that it is caused by biological agencies, second, that its cause is purely chemical. The biological agencies that are mentioned as causes are yeasts, Oidium lactis , bacteria, and bacterial enzymes. (For a more detailed summary see page 29 “Biological Agencies and Fishiness”). The theory that biological agencies are the cause of fishiness is not by any means free from objections. The main objections that are raised against the theory are : ( 1 ) Inoculation experiments with the organisms supposed to cause fishiness have in general been unsuccessful ; (2) Pasteurization does not prevent fishiness absolutely; (3) Fishiness is favored by a high salt content, and develops at low temperatures, conditions which would inhibit the activity of microrganisms. Largely because of these objections some of the investigators (Rogers 18 , Thomson <>0 , 21 , Reakes, Cuddie and Reid 26 , Dyer 39 , hold the theory that The Fishy Flavor in Butter 9 fishiness in butter is the result of a spontaneous chemical change. The main objection to this theory and in favor of the biological theory is that pasteurization reduces the prevalence of fishiness even if it does not eliminate the malady entirely. Thus the controversy between these two schools of thought must be considered unsettled. While there is no agreement on the agency that causes fishiness, there is quite general agreement on the factors that favor the development of fishiness. A high acidity in the cream is one of the factors on which there is general agreement. Other factors that are well supported by experi- mental evidence are high salt, overworking, and the presence of the salts of metals such as iron and copper. The evidence presented also shows that pasteurization of the cream helps to eliminate fishiness. The identity of the fishy substance in the butter has not been established beyond all doubt. The most definite theory is that the fishy substance is trimethylamine and it is formed from lecithin. Rogers l8 objects to this theory on the basis of an experiment in which he obtained the fishy flavor in the distillate from fishy butter acidified with sulphuric acid, which, he concludes, eliminates basic substances, such as trimethylamine, from con- sideration. Thus the identity of the fishy substance and the agencies that produce it must still be considered unsettled questions. General Plan of the Experimental Work In our attempt to determine the origin and the cause of the fishy flavor in butter the following* plan of procedure was followed : 1. By means of experiments and a critical review of the literature deter- mine the factors and conditions that favor the development of fishiness. 2. On the assumption that lecithin is the mother substance of the fishy flavor, determine the conditions under which lecithin will yield trime- thylamine. If lecithin is the mother substance of the fishy flavor, then the conditions that favor the development of fishiness must also cause the greatest yield of trimethylamine. 3. Isolate trimethylamine from fishy butter. 4. On the assumption that lecithin is the mother substance determine whether butter to which lecithin has been added will become fishy more readily than ordinary butter. 5. From the study of the decomposition of lecithin determine whether the action is chemical, bacteriological or enzymatic. A Study of the Conditions that Favor Fishiness In order to determine the conditions that favor the development of fishi- ness in butter, the various factors suggested in the literature were studied as indicated in the following outline : Sweet cream, 400 pounds, was divided into ten lots of 40 pounds each as follows : 10 Wisconsin Research Bulletin 57 Lot 1 — Raw, unripened. Lot 2 — Raw, naturally ripened, acidity 0.4%. Lot 3 — Raw, naturally ripened, acidity 0.6%. Lot 4 — Raw, naturally ripened, acidity 0.6% neutralized to 0.25%. Lot 5 — Pasteurized, unripened. , Lot 6 — Pasteurized, starter ripened, acidity 0.4%. Lot 7 — Pasteurized, starter ripened, acidity 0.6%. Lot 8 — Pasteurized, starter ripened, acidity 0.6% neutralized to 0.25%. Lot 9 — Pasteurized after natural ripening to 0.4% acidity. Lot 10 — Pasteurized after natural ripening to 0.6% acidity. Each of the ten lots of cream was prepared and handled so as to con- form to the conditions as stated in the above outline. The ripening was done by keeping the cream at a temperature of 70° F. until the desired acidity was reached. In the starter ripening a good freshly prepared starter was used. In neutralizing lots four and eight sodium bicarbonate was used. The churning of the cream was done in an earthenware hand churn, after the cream had been cooled down to 40° F. It was churned to the granular size, the buttermilk drawn off and the butter washed with well water pre- viously cooled to 40° F. The butter was then worked in the churn with a ladle until the excessive moisture had been worked out. During the various operations of ripening and churning, the cream and butter were kept out of contact with metallic surfaces as much as possible. The cans used were new and well tinned with no exposed iron surface, the churn was earthen ware and glass jars were used for storing the butter. The butter made from each of the ten lots of cream was divided into seven parts and treated as follows : Part 1 — Unsalted. Part 2 — Medium salted, 1.5% salt. Part 3 — Highly salted, 3.0% salt. Part 4— Highly salted, 3.0% salt, plus 0.1% ferric oxide. Part 5 — Highly salted, 3.0% salt, plus 0.1% iron lactate. Part 6 — Highly salted, 3.0% salt, plus 0.1% tin lactate. Part 7 — Highly salted, 3.0% salt, and overworked. Each of these seven parts prepared from each of the ten lots of cream was divided into two parts ; one of these was stored as — 10° F., the other was stored at 35 to 40° F. The samples were scored by three judges at intervals of about a month and the different flavors carefully noted. The results are given in Tables I and II. TABLE I.— FLAVORS DEVELOPED IN THE EXPERIMENTAL BUTTER AT 35 TO 40° tment of the butter Treatment of the butter 3.0% salt plus overworking Slightly oily Slightly oily Oily Oiiy $ Slightly fishy K Oily and fishy W Fish ^ -a = *‘ 5B >i«q| > >> o Y FLj >>>>>. ooo Oily < Oily O Very oily Oily § Oily Very oily \jj UTTEE >> oo 13 > Slightly oily Slightly oily Tallowy Slightly oily Slightly oily Oily Oily Oily Very oily >-* 3.0% salt plus 0.1% tin lactate Slightly oily Slightly oily Oily *I!0 3.0% salt plus 0.1 % iron lactate Oily Slightly fishy Fishy Fishy Very fishy Very fishy Very fishy Very fishy Very fishy Very oily Fishy Fishy & tallo’y Oily Very oily Fishy Metallic & oily Slightly fishy Tallowy Oily Slightly fishy Tallowy Metallic & oily Oily Tallowy Very metallic Metallic & oily Tallowy Metallic & oily Oily Tallowy 3.0% salt plus 0.1% ferric oxide Metallic Metallic & oily Slightly fishy Slightly fishy Fishy Fishy Slightly fishy Fishy Fishy Oily Oily Tallowy Oily Oily Tallowy Oily Oily Tallowy Oily Oily Very oily Oily Oily Oily & tallowy Metallic Oily Oily Oily Oily Oily & Tallowy 3.0% salt Slightly oily Slightly oily Oily Oily Oily Very oily Very oily Very oily Slightly fishy Slightly oily Slightly oily Oily Oily Oily Oily Slightly oily Slightly oily Oily Oily 1.5% salt Slightly oily Slightly oily Slightly oily Slightly oily Slightly oily Slightly oily Oily Oiiy Oily : : : c3 <13 u h Unsalted Months storage ■<*<000 ■^CDOO rtOOO rj<000 00 rj<«£>00 ^tocc ■tfOOO n> 2? fao 5 & fafa oo ►.fa oo fao 5 fa fafa OO _fa o fa£ So .BP Tn >> fao O u, > fa'o 03 > j 3 •s c 'c 3 b §5 o° CO £►>►> ooo ►> o fa£ SO 00 fafa oo jjfaj SS^ .2P.2P Wfi %* o° co fa far* •s j ® 3&s So ^ fa "o ° J fa,2°S <3ta ► > -a ® fa s* 5 fa 2*5 fefa * > -a ►>►> oo . $ o oO - 2 - 2 H 4> 03 ss 2? ' *>>$ ons o ^Ort 3 h — fa* 60^ i 3 ® «3 £ sf -a ^ 2 « fa s&S ooS .Sfa fao £ § jl >?fa fa'o <* 3 jl -.S'S aS**S WkoX es-° o.o co fa JT ns o *5 £►><* S « fa ■*>3 -as cS .fa >,' H> sf* -a fa mj M ® ^fa fa far* o «— i cs »a fa —> 6 DJ ^3- K* j > OO I OOtJ aa* o o 2 ? SSo 3*t H 33 ■*s£ •SfaS 3£h or *5-g £ •H *2 3|S 03 ^ .2 fa 3 fa| C 3 •— * O S h s>fa *'°! °|g x*fa faO £ S ft fafa OO fa fa-3 S fa O Sh 0> > fa £ JO fafal3 ooj fa o fafa oo fafac oo £ fa fao OO u fafafa ooo fafafa ooo fa fafa-o OO u fa_fa oo fafa M.OC fafa OO x Sr «2 O O SS The Fishy Flavor in Butter 13 From Tables I and II we can make the following observations: 1. Fishiness appeared earlier and more often in the samples stored at the higher temperature. 2. True fishiness was found practically only in the unpasteurized samples (lots 1, 2, 3, and 4). 3. Pasteurization after ripening (lots 9 and 10) seemed to be more effective in preventing fishiness than pasteurization before ripening (lots 5, 6, 7 and 8). Lots 6 and 7, pasteurized and then ripened, showed fishiness in the samples containing 3.0% salt and iron lactate, while the corresponding lots, lots 9 and 10, pasteurized after ripening, showed no signs of fishiness. 4. The high acid samples, lots 2 and 3 especially, developed fishiness most rapidly. 5. None of the unsalted and low salted samples developed fishiness. 6. Only one of the samples where acidity and salt alone were active developed a slight fishy flavor. 7. The presence of iron oxide and iron lactate combined with high acid salt caused the most distinct and greatest number of fishy samples. The iron lactate was more active than the iron oxide. 8. Tin lactate did not cause fishiness. 9. Overworking showed a slight tendency to aid the development of fishiness. 10. Neutralization to a low acidity before churning was effective in re- ducing fishiness. Lots 4 and 8, neutralized to 0.25% acid, showed less fishi- ness than the corresponding unneutralized samples, lots 3 and 7. Thus the results of these experiments clearly show that high acid, high salt and oxidation produced by overworking, but more especially by the aid of catalyzers, are the important factors in the development of fishiness. The results also show that pasteurization and neutralization are effective means of checking fishiness. Let us compare these results and conclusions with those of previous in- vestigators, so that by means of this detailed consideration we may clearly establish the factors and practices that help and that hinder the development of fishiness. Acidity and Fishiness The conclusion that a high acidity in the cream favors the development of fishiness is borne out by the work of the following investigators : Rogers l8 , Davis 23 , 24 , Rogers et al 25 , Reakes, Cuddie and Reid 26 , Thomson 2 o» 2 i» Fleischmann 38 , Dyer 39 , Erickson 45 , Bouska and Washburn 47 , Supplee 50 , Hunziker 5l , McKay 52 , Brown , 53 , Cusick 55 , and others. The importance of acidity in the development of fishiness is clearly indi- cated in the following quotation from the work of Rogers, Thompson and Kiethley 25 : “In a tabulation of the examination of 259 samples of experimental butter from cream of known acidity, of 137 samples from cream having an 14 Wisconsin Research Bulletin 5 7 acidity below 0.3%, only two or 1.5% were marked ‘fishy/ while of 122 samples having an acidity of 0.3% or over, sixty or 49.2% were fishy.” A high degree of acidity in cream generally shows that the cream is also highly fermented. It might thus be thought that the acidity is merely incidental and that the real agency causing the development 6f the fishy flavor is biological, and not the acid per se. That this is not the case has been clearly demonstrated by Rogers and Gray lg by comparing the flavor of butter made from sweet cream with butter made from another portion of the same cream acidified artificially. Table III is quoted from Rogers and Gray. (Rogers & Gray [ 18 J Table X.) TABLE III.-THE EFFECT OF ARTIFICIALLY ADDED ACIDS ON THE FLAVOR OF BUTTER Acid added Acidity of cream Remarks after 15 days None Per Cent 0.144 Very greasy Lactic 0.216 Very greasy Lactic 0.432 Fishy and greasy None 0.126 Trifle Oily Acetic 0216 Very greasy, rancid and fishy odor Acetic 0.350 Very fishy and greasy None 0.126 Clean but greasy Hydrochloric 0.225 Trifle unclean and oily Hydrochloric 0.450 Very fishy and greasy Salt and Fishiness Some investigators believe that salt improves the general keeping quality of butter. Among these are: McKay and Larsen 60 , Fettig 62 , Rahn, Brown, and Smith 19 , Larsen, Lund and Miller g3 , Weigmann., 2 , Jensen 30 , Klein C4 . However, at lower temperatures other investigators have shown that salt lowers the keeping quality. Investigators offering convincing evi- dence on this point are Gray and McKay 14 , Hunziker, Mills and Spitzer 65 , Kildee 66 , Washburn and Dahlberg 42 . From a study of these references we can conclude that salt at the storage temperatures deteriorates rather than preserves the butter. However, it does not necessarily follow from this conclusion that this particular deterioration, the development of fishiness, is also accelerated by salt. There is some difference of opinion on this point. Some investigators have found fishiness only in salted butter, while others have found it in unsalted butter as well, as in salted. 0’Callaghan l5 states that many consignments of Australian unsalted butter have turned out fishy in London. Thomson 20 , 21 states that a fishy flavor will develop in both salted and unsalted butter, but more commonly in the salted butter. Rogers 18 in all his work on fishiness has never met fishiness in unsalted butter. Washburn and Dahlberg 42 have found that salted butter is more likely to turn fishy in storage than unsalted butter. The Fishy Flavor in Butter 15 Supplee 50 found that out of a total of 105 characteristic comments, 93 were found in samples containing salt. None of his unsalted inoculated samples were scored fishy. The following Table IV compiled from various sources indicates the importance of salt in the development of fishiness : TABLE IV.— EFFECT OF SALT ON FISHINESS IN BUTTER AS COMPILED FROM LITERATURE Kind of butter Amount of salt Acid St( >rage Flavor Source Temp. °F. Time Good butter None 18 12 5 2 mo. 2 mo. 2 mo. Stale Weak Good Thomson, 20>21 Choice butter Salted 18 12 5 2 mo. 2 mo. 2 mo. Very fishy Fishy Trace fishy Pasteurized — ripened None 20 7 mo. Good Rogers et al., 26 Pasteurized — ripened Salted 0 10 20 7 mo. 7 mo. 7 mo. Storage Fishy Fishy Unpasteurized — ripened Salted 0 10 20 7 mo. 7 mo. 7 mo. Fishy Fishy Very fishy Unpasteurized .... Pasteurized Salted Low salt High High 35 35 3 mo. 3 mo. Fishy Cheesy O’Callaghan, 27 Unpasteurized — ripened Unsalted 0.68 0 45 da. Metallic Supplee, ou Unpasteurized — ripened Salted 0.68 0 45 da. Fishy Neutralized and then ripened. . . . Unsalted —10 8 mo. Cusick, 66 Neutralized and then ripened. . . . Salted —10 8 mo. Fishy Unpasteurized — ripened, churn 1 0.0 0.58 —15 113 da. Metallic Washburn and Dahlberg, ^ Unpasteurized — ripened, churn 1 . 2.51 0.58 —15 113 da. Fishy From our experimental results and from the literature we can conclude that salt favors the development of fishiness, but that under certain condi- tions this flavor may develop without the aid of the salt. Overworking and Fishiness That overworking the butter promotes the development of fishiness in the butter during storage has been shown by Rogers lg , Weigmann 20 , Reakes, Cuddie and Reid 26 , Hunziker 5l , and McKay 52 . The following Table quoted from Rogers lg clearly demonstrates that fishiness is accelerated by overworking. 16 Wisconsin Research Bulletin 57 TABLE V.-THE INFLUENCE OF OVERWORKING ON STORAGE BUTTER No. Acidity Working Comments 1 Unripened starter added Not worked Aged Overworked Oily, trifle rancid 2 0.405% No't worked Suggestion of fishy, fruity Overworked Fishy 3 1 0.586% Not worked Aged and acid Overworked Badly fishy and aged Metals and Fishiness It has been known for some time that the presence of metallic salts has a deteriorating effect on the quality of the butter. This has been' shown by Weigmann 3 , 4 , Henzold 67 , Marcas and Huyge 68 , 69 , Hoft 70 , Kooper 71 , Rogers et al 29 , Rogers 31 , Hunziker and Hosman 72 , Palmer and Combs 7g McKay 54 , Hunziker 5l , McKay and Larsen 6Q , Ruehle 74 , and Molkerei Zeitung 75 . Rogers 31 states that the presence of iron in the cream in amounts as small as one part per million parts of cream has a marked effect on the flavor of the resulting butter. This amount of iron, he says, may easily come from rusty cream cans or some other utensils. Marcas and Huyge 68 , 69 found that cream with a normal iron content of 0.005 parts per thousand increased to 0.240 parts per thousand during twenty-two hours’ contact with a rusted can. The literature shows that the most common flavor that develops in storage butter as a result of the presence of metallic salts is fishiness: Rogers et al 29 Rogers 31 , Washburn 70 , Hunziker 5l , and Klein 44 . Rogers et al„ 9 found that a great many of the butters to which iron had been added became fishy, and that where the control also became fishy the control always was the last to show the flavor. The development of fishi- ness in butter containing added copper sulphate was even more marked. They found that after forty days’ storage most of the butter to which copper had been added showed a fishy flavor, and that after three months of storage a decided rank, fishy flavor had developed. Tin did not produce any fishiness. Rogers 31 found in some experiments with a pasteurizer in which the copper was badly exposed, that the score of the fresh butter was decreased three to four points below that from cream pasteurized in a well tinned pasteurizer, and that in thirty days the butter from the copper pasteurizer had become very fishy. Pasteurization and Fishiness Pasteurization is commonly regarded as an effective means of improving the keeping quality of butter. That it is also an important means of con- The Fishy Flavor in Butter 17 trolling the development of fishiness has been shown by a number of in- vestigators : 0’Callaghan 11 , Rogers 18 , Rogers et al 25 , Thomson 21 , Reakes, Cuddie and Reid 26 , Supplee 50 , Cusick 55 , McKay 52 & 54 , Hunziker 5l , McKay and Larsen 60 , Brown, Smith and Ruehle 56 , and Washburn and Dahlberg 42 . Hunziker 37 has found that extremely high pasteurizing temperatures, such as 185° F. and higher, may cause a very poor quality butter, the result- ing butter tending to have a disagreeable flavor, suggestive of fishiness. Rogers lg found that butter made from cream pasteurized at 180° F. still may develop fishiness during storage. Rogers et al 25 found similar results. Marker 77 claims that butter made from cream pasteurized at 170 to 175° F. and holding for 15 to 20 minutes, does not go fishy when placed in storage. The literature tends to show that pasteurization at lower temperatures for a longer time is more effective in checking the development of fishiness than pasteurization at higher temperatures for a shorter time. From a review of the literature we can conclude that pasteurization helps to check the development of fishiness, but that it is not an absolute preventive. General Summary of Experiments and Literature From the experiments conducted and the literature reviewed we can draw the following conclusions : 1. High acid, high salt, overworking and the presence of metals such as iron and copper favor the development of fishiness. 2. Pasteurization helps to check the development of fishiness, but it is not an absolute preventive. 3. Neutralization, by removing one of the favorable factors, high acid, helps to prevent fishiness. A Study of the Cause of Fishiness in Storage Butter The cause of fishiness in butter has been attributed to products of de- composition of some of the constituents of butter through biological agencies. Among the advocates of this theory are: Storch 1 , Kirchner 2 , O’Callaghan 6> n> i 5 » Piffard 9 , Harrison l0 , Sommerfeld l7 , Rahn Brown and Smith ig Orla Jensen 30 , Klein 33 , Fleischmann 38 , Supplee 50 , Brown, Smith and Ruehle 56 , Cusick 55 . Another school again has taught that the cause of fishiness is mainly chemical. Among the advocates of this theory are Rogers lg , Weigmann 3 , 4 , Thomson 20 , 21 , Reakes, Cuddie and Reid 26 , and Dyer 39 . Weigmann 3 , 4 attributed the cause of fishiness to iron which entered the cream through the use of poorly tinned vessels. He was able to produce fishy butter by adding iron lactate to ripened cream. Rogers l8 is of the opinion that fishiness is caused by a slow, spontaneous chemical change to which acid is essential and which is favored by the presence of small amounts of oxygen. 18 Wisconsin Research Bulletin 57 Thomson 20 , sums up his experimental work on fishiness in butter by saying that it is largely a chemical change and that the production of the flavor is favored by a high degree of acid and salt, and a temperature of storage of between 30 and 40° F. There is a third school that attributes the immediate cause of fishiness to the presence of trimethylamine. Rogers l8 says that the immediate cause of fishiness in butter is believed to be due to the presence of trimethylamine in butter, although Rogers himself does not subscribe to this theory. Som- merfeld l7 was of the opinion that the fishy taste was due to trimethylamine and that it resulted from cattle grazing on meadows which were frequently flooded. Supplee 50 with considerable experimental evidence, advanced the trimethylamine theory, holding that the cause of its formation was mainly biological. A similar conclusion was drown from the work of Cusick 55 . The data from Supplee 50 and also Cusick’s work, seem to indicate that the most logical source of the trimethylamine- in the butter is lecithin. If lecithin is to be regarded as the source of the fishy flavor, then two con- ditions must be fulfilled; first, there must be enough lecithin normally present in milk, cream and butter to give rise to sufficient trimethylamine to account for the flavor; second, the optimum conditions under which lecithin will yield trimethylamine must coincide with the conditions which are known to favor the development of fishiness in storage butter. The following pages are concerned with a consideration of these two conditions. The Amount of Lecithin in Dairy Products The lecithin content of milk and other dairy products has received the attention of a number of investigators. Table VI gives a summary of the lecithin content of milk, cream and butter as compiled from the literature. There can be no doubt about the fact that milk contains lecithin; the work of a large number of investigators verifies this : Tolmatschaff 93 , Wrampelmeyer 78 , Beilstein 79 , Stoklasa 80 , Burow gl , Schmidt-Muhlheim 8 ,, Bardas and de Raczkowski 83 - 87 , Koch and Woods g8 , Nerking and Efaensel 8g Lewkowitsch 90 , Gilkin gl , Brodrick-Pittard 92 , Supplee 50 , Winterstein g4 Koch 95 , Fetzer 96 , Osborn and Wakeman g7 , Arbenz 98 , Cusick 55 . Table VI shows that there is an appreciable difference in the amount of lecithin found by various investigators. Part of the difference is un- doubtedly due to imperfections in the analytical methods used, and part to an actual difference in the samples. Very little is known about the factors that determine the amount of lecithin in milk. The results of Bardas and de Raczkowski g3 - 87 and Brodrick-Pittard g2 indicate that the lecithin content is highest in the early part of the lactation period. Fetzer 96 , and Brodrick-Pittard 92 found that the lecithin content varied directly with the fat content. Brodrick-Pittard also state that it is dependent upon the individuality of the cow. Besides the factors mentioned above to account for differences in lecithin content, several investigators have shown that the lecithin content of butter The Fishy Flavor in Butter 19 TABLE VI.— THE LECITHIN CONTENT OF DAIRY PRODUCTS Kind of milk M ilk Cream Bu tter Source Low- est High- est Average Low- est High- est Per cent Per cent Per cent Per cent 0.007 Per cent 0.033 Wrampelmeyer, 78 0.15 0.17 Beilstein, 79 0.170 0.090 0.186 0.113 Stoklasa, so 0.049 0.058 Burow, 31 0.04 Schmidt-Muhl- heim, g* Cow 0.043 0.058 Bardas and Raczkowski, S3~S" Cow 0.036 0.049 Koch and Woods , 88 Cow (17 samples)... 0.036 0.116 Nerking and Haensel, 39 Human (10 samples) 0.024 0.080 Ass (6 samples) .... 0.006 0.039 Ewe (4 samples) .... 0.051 0.167 Goat (11 samples). . 0.036 0.075 Mare (8 samples). . . 0.007 0.017 Cow 0.017 0.170 Lewkowitsch, 90 Gilkin, 91 Cow 0.0515 0.050 Cow 0.0553 (15% cream) 0.090 Brodrick-Pit- tard, 92 Cow 0.0723 Supplee, 60 is influenced by the pasteurization of the cream. Thus Bordas and de Raczkowski g3 - 87 found, in comparing the lecithin content of raw and pasteurized milk, that heating at 60° C. for 30 minutes reduced it 14%, 95° C. for 30 minutes reduced it 28%, and autoclaving at 105 to 110° C. for 30 minutes reduced it 30%. Dornic and Daire 99 found that buttermilk from naturally ripened pasteurized cream contained more lecithin than milk. Supplee 50 found that sweet cream butter contained 0.0723% lecithin, while the corresponding pasteurized ripened cream butter contained only 0.0433%. A Study of Conditions Under Which Lecithin Yields Trimethylamine If lecithin is to be regarded as the mother substance of the fishy flavor, trimethylamine, then it must undergo decomposition under the conditions that produce fishy butter. Moreover, using conditions imitating those in butter, the conditions that give the maximum decomposition of lecithin 20 Wisconsin Research Bulletin 57 into trimethylamine must coincide with the conditions known to favor the development of fishiness in storage butter. With this in view the decom- position of lecithin under various conditions was studied. Trimethylamine Determination The trimethylamine determination used consisted of two steps : First, the quantitative aspiration of volatile bases from the alkaline solution and ab- sorption in standard acid; second, the conversion of the ammonia into hexamethylenetetramine by means of an excess of neutral formaldehyde with the liberation of an equivalent amount of acid which can be titrated to determine the amount of ammonia, and from that the amount of trime- thylamine by difference. The aspiration used was that of Folin and Farmer l00 and Folin and Macallum l0l , and also used by Supplee 50 for isolating volatile bases from samples of fishy butter. Fifteen cubic centimeters of the substance con- taining the volatile bases were transferred to the proper tube of the Folin apparatus, covered with a thin film of a thirty per cent rosin solution in turpentine to prevent foaming, ten grams of anhydrous potassium carbonate were added and the contents of the tube aspirated for 5 hours. The volatile bases set free were collected in 5 cc. of N/10 sulphuric acid diluted to 50 cc. and the excess of acid at the end of the aspiration titrated with N /100 sodium hydroxide using alizarin as indicator. The result of this titration thus gave the total amount of volatile bases. To determine the trimethytamine by difference the ammonia was con- verted into hexamethylene-tetramine by the addition of 20 cc. of 40% neutral formaldehyde after the above titration for the total volatile bases. The amount of acid liberated from the ammonium salt was titrated with N/100 sodium hydroxide, and the end point determined by comparing the color with a standard in a comparator block. This titration gave a meas- ure of the amount of ammonia, and thus by difference between the total bases and the ammonia the trimethylamine was determined. In all cases the volume of the solution to be titrated was kept as small and constant as possible. The amount of the indicator was the same for each titration. Blank determinations were run with each set of determi- nations. Before any determinations were made by this method it was studied to determine its efficiency. The conversion of ammonia into hexamethylene- tetramine by formaldehyde has been studied by a number of investigators : Delepine 102 , Cambier and Brochet l03 , Ronchese l04 , Malfatti Wilkie l06 , Parker lo7 , Thau l08 , Meurice l09 , Sanders ll0 , Budai m , and Supplee 50 . It has been shown that the reaction is complete and not reversible 6 CH 2 0+4 NH 4 Cl-MNaOH-*(CH 2 ) 6 N 4 +10 H 2 0+4NaCl That this conclusion is correct and the method based on it is sound has been substantiated by our experiments. In a series of ten determinations, starting with a known amount of neutral ammonium chloride, and adding The Fishy Flavor in Butter 21 neutral formaldehyde, the titrations of the acidity developed accounted for 99.93% of the ammonia known to be present, (extremes, — 99.86 to 99.95.) In a series of ten determinations, starting with a known amount of am- monium chloride, and taking it through the aspiration and the formaldehyde conversion processes, 99.275% of the ammonia was accounted for (extremes 98.63 to 100.00). In a series of four determinations, starting with a mixture of known amounts of ammonia and trimethylamine and taking them through aspira- tion and conversion processes, determining the trimethylamine by difference, 99.23% (extremes 98.63 to 99.74%) of the original ammonia, and 97.57% (extremes 97.45 to 97.68%) of the original trimethylamine were accounted for. These results justify the use of this method for the determination of trimethylamine in the presence of ammonia. Preparation of Lecithin The lecithin used in the experiments was prepared from egg yolks ac- cording to the method described by Long 42 . The yolks of six dozen fresh eggs were separated from the whites and then squeezed through cheese cloth into a large bottle containing 1888 cc. ether. The bottle was corked securely, put in a shaker and shaken continuously for ten hours ; 3000 cc. of alcohol were then added and the mixture again shaken for twelve hours. The mixture was allowed to settle and the alcohol-ether solution filtered and distilled to a pasty condition at a low temperature on a water bath and finally by the aid of a vacuum. The residue was taken up in pure ether, the solution filtered, concentrated (as much as possible out of contact with air) in a distilling flask and precipitated with pure neutral acetone in excess. The lecithin settled out as a white heavy precipitate. This operation of dis- solving in ether and reprecipitating by acetone was repeated three times, the last product being carefully dried in a vacuum desiccator over calcium chloride. In this way a product was obtained which, at first, was light yellow in color and plastic, but which with age and loss of water, became darker and horny. Experimental Using the lecithin prepared as outlined above emulsions of the lecithin were made and subjected to various conditions. Each of the samples pre- pared was exactly 15 cc. in volume. The concentration of lecithin and other conditions in these solutions were such that they imitate the conditions in butter. It was assumed that butter becoming fishy had 0.1% lecithin (see Table VI), 16% moisture and 2 to 4% salt. Assuming that all the salt, acid and lecithin in the butter were in solution in the brine of the butter, emulsions were prepared to imitate the concentrations of the above assumed butter brine. The following combinations were studied : 22 Wisconsin Research Bulletin 57 1. 0.1% lecithin. i irncalt^ 2. 0.1% lecithin, 0.1% ferrous lactate. nSa 6 3. 0.1% lecithin, hydrogen peroxide 1 cc. to 15 cc. emulsion. 2. Salt 2% (Same as above) 3. Salt 4% , “ “ 4. Unsalted, 0.25% lactic acid “ “ “ 5. Unsalted, 0.50% lactic acid “ “ “ 6. Salt 2%, lactic acid 0.25% “ “ “ 7. Salt 2%, lactic acid 0.50% “ “ 8. Salt 4%, lactic acid 0.25% “ “ “ 9. Salt 4%, lactic acid 0.50% “ “ “ (Note: The figures refer to the percentage of salt, lecithin, acid and ferrous lactate in the corresponding butter. The actual concentration in the brine or emulsions studied is therefore higher than the figures given. The concentration of the emulsions is actually 6J4 times as high as the figures given in the outline.) Thus in this experiment there were three series of samples. The first series consisted of lecithin emulsions under various combinations of salt and acid, the second series was similar to the first with the addition of 0.1% ferrous lactate to all samples in the series, the third series was similar to the first with the addition of 1 cc. of three per cent hydrogen peroxide to each sample in the series. Small amounts of ether were added to each tube to inhibit the development of microorganisms. The tubes were then tightly corked, sealed with paraffin wax and incubated at 35° C. for six weeks. At the end of that time the amount of trimethylamine in each tube was determined according to the method described. The trimethylamine found was then calculated and expressed as per cent of the total trimethyla- mine theoretically possible from the amount of lecithin in the sample. In this calculation lecithin was taken as the distearyl type and the nitrogen present was assumed to be all in the form of choline. Table VII gives the results of this set of experiments. The results of this experiment as given in Table VII show that there is not much decomposition of lecithin into trimethylamine where oxidation is not favored, but that where oxidization is favored there is a very marked decomposition. In all cases the best decomposition is each series was ob- tained where high acid and salt were present; in the hydrogen peroxide series, five times as high as in the samples where acid and salt were not pres- ent. Salt alone and acid alone gave only slightly greater decomposition than the control except in the hydrogen peroxide series where the presence of the acid caused a decided increase. In general the samples with the highest acidity gave the highest percentage decomposition. The Fishy Flavor in Butter 23 TABLE VII.— THE CHEMICAL DECOMPOSITION OF LECITHIN INTO TRIMETHYLAMINE 0 . 1% lecithin plus Nothing 0.1% ferrous lactate 1 cc. hydrogen peroxide to 14 cc. No salt, no acid . A (Per cent of tri 4.99 methylamine of 4.74 total possible) 4.31 2% salt, no acid 5.17 0.86 5.26 4% salt, no acid 5.17 0.86 5.17 No salt, 0.25% lactic acid 5.60 5.69 8.62 No salt, 0.50% lactic acid. . . , . . 5.69 5.69 9.48 2% salt, 0.25% lactic acid 6.20 5.86 11.20 4% salt, 0.25% lactic acid 6.20 5.77 11.20 2% salt, 0.50% lactic acid 6.89 7.32 22.40 4% salt, 0.50% lactic acid 7.07 7.75 22.83 Note: The amount of lecithin, ferrous lactate, salt and acid are expressed on the basis of (16% moisture) butter corresponding to these emulsions, i. e., the actual concentration of these substances in the emulsions is 6)4 times as great as stated in the table. TABLE VIII.— THE CHEMICAL DECOMPOSITION OF LECITHIN INTO TRIMETHYLAMINE 0.1% lecithin plus: Treatment of sample Nothing 0.1 % fer- rous lactate Hydrogen peroxide 1 cc. to 14 Oxygen 0.1 % fer- ' rous lac- tate plus oxygen No salt, no acid (Per cen 1.89 t of trimeth 1.46 ylamine of 2.49 total pos 2.16 sible) 2.50 2% salt, no acid 2.85 1.66 5.77 2.85 4.65 4% salt, no acid 2.85 2.50 5.77 2.85 4.99 No salt, 0.25% lactic acid. 4.22 3.32 8.79 6.55 6.60 No salt, 0.50% lactic acid. 4.48 4.32 10.34 10.52 7.93 2% salt, 0.25% lactic acid. 5.60 4.98 11.29 14.65 8.43 4% salt, 0.25% lactic acid. 5.60 5.77 11.20 14.82 8.13 2% salt, 0.50% lactic acid. 5.77 7.41 13.78 17.23 8.87 4% salt, 0.50% lactic acid . 6.29 7.41 14.04 17.75 9.13 Note: The amount of lecithin ferrous lactate, salt and acid are expressed on the basis of butter (16% moisture) corresponding to these emulsions, i. e., the actual concentration of these substances in the emulsions is 6J4 times as’ great as stated in the table. 24 Wisconsin Research Bulletin 57 Thus the results show that acidity, salt and oxidization play an important part in the decomposition of lecithin into trimethylamine just as these same factors do in the development of fishiness in butter during storage. This experiment was repeated in the same way except that mercuric chloride was used as a preservative and with two additional series, (a) Lecithin plus oxygen (b) Lecithin plus ferrous lactate plus oxygen. In these two series the air above the samples in the bottles was displaced by forcing oxygen through from an oxygen drum. Instead of the test tubes used in the first experiment half pint milk bottles were used for the sam- ples. The milk bottles were tightly corked and incubated for three weeks at 28° C. The amount of trimethylamine was then determined and calcu- lated as per cent of the total trimethylamine theoretically possible from the amount of lecithin used. The results are given in Table VIII. The results given in Table VIII show that lecithin will undergo decom- position and yield trimethylamine under condition's that exclude bacterial action. This decomposition takes place most readily under conditions that combine salt, high acid and oxidization. These results correlate very nicely with the conditions known to favor the development of fishiness, and thus lend support to the theory that the fishy flavor results from the chemical decomposition of lecithin yielding trimethylamine. Trimethylamine and Fishiness If trimethylamine is the cause of fishiness in butter, it should cause a characteristic fishy flavor when worked into normal butter. This was actually tried out by working 100 parts of trimethylamine lactate into a million parts of fresh butter ; this sample was then submitted to a number of judges, and laymen, and was readily recognized as fishy by all. The tri- methylamine lactate used in this test was obtained from the lecithin de- composition experiments described above. After the titrations for tri- methylamine had been made the solutions from the various titrations were concentrated, the trimethylamine liberated by the addition of potassium bi- carbonate and absorbed in lactic acid of known strength. The amount of trimethylamine in the solution was finally determined by the trimethylamine determination described above, and enough of the solution added to obtain 100 parts of trimethylamine lactate per million parts of butter. By working trimethylamine and its salts into butter Supplee 50 also showed that it readily caused the characteristic fishy flavor. Rogers 18 , however, worked large quantities of trimethylamine into butter without producing the characteristic fishy flavor. These negative results of Rogers are probably due to the very fact that large quantities were used, for in large amounts trimethylamine smells stringent like ammonia, and does not smell charac- teristically fishy. Thus, if large amounts of trimethylamine or its salts are worked into butter, it imparts to the butter a disagreeable, but not disinctly fishy, flavor. The amount of trimethylamine worked into butter in our experiment, and the amounts worked into butter by Supplee 50 are reasonable amounts for the The Fishy Flavor in Butter 25 purpose. In order to obtain 100 parts of trimethylamine per million parts of butter, the butter would have to have a lecithin content of 0.136%, assuming lecithin of the distearyl type and assuming that all of the nitrogen in the lecithin goes to trimethylamme. This amount of lecithin is possible in butter as shown by Table VI. Further Supplee 50 found 85 parts of trimethylamine salts per million of butter caused a distinct fishy smell and taste. Besides the objection that trimethylamine worked into butter did not pro- duce a fishy flavor, Rogers objects further to the trimethylamine theory on the grounds that he found a distinct fishy smell in the distillate from the fishy butter acidified with dilute sulphuric acid. In this objection Rogers makes the a priori assumption that volatile bases cannot be dis- tilled from acid solutions. That this is incorrect is demonstrated by the following trial. A solution of trimethylamine sulphate, 0.5% was made distinctly acid to litmus and alizarin, then carefully distilled and the dis- tillate gathered. In each of several such trials the distillate had a dis- tinct fishy odor and flavor. That Rogers’ assumption is incorrect is fur- ther demonstrated by the following experiment in which ammonia is dis- tilled from an acid solution: Twenty grams of ammonium sulphate were dissolved in a liter of dis- tilled water in a two liter distilling flask. The flask was connected with a condenser which was connected to a receiving flask by means of an adapter, making a closed system. The condenser and receiving flask had been rinsed with ammonia free water. In distilling a small flame was used and special care was taken to prevent superheating of the walls of the distilling flask above the level of the liquid. A series of distillations were made from the ammonium sulphate solutions acidified to various degrees by means of sulphuric acid. In each case 50 cc. of distillate were gathered. The amount of ammonia, in the distillate was determined by Nesslerizing and comparing with a standard, and the reaction of the solution from which it was distilled was determined colorimetrically. The results are given in Table IX. TABLE IX. — THE DISTILLATION OF AMMONIA FROM ACID SOLUTION Amount of dilute sulphuric Reaction Ammonia in distillate acid added pH parts per million 0.0 6.2 95.7 1.0 3.5 40.9 2.0 3.1 22.3 3.0 2.7 17.0 4.0 2.5 13.9 5.0 2.4 13.7 26 Wisconsin Research Bulletin 57 The results given in Table IX show that an appreciable amount of am- monia is distilled from solutions that are quite acid. These results can readily be explained as follows : The ammonium sulphate in aqueous solution dissociates into ammonium ions and sulphate ions. (NH 4 ) 2 S0 4 2NH+ + S0 4 In the aqueous solution there are also hydroxyl ions and hydrogen ions, and at any given temperature the product of their concentrations is a con- stant. At high temperatures this constant is higher than at low tem- peratures. Therefore at the boiling point of the solutions there will be an appreciable concentration of hydroxyl ions, the exact concentration in each case being dependent upon the hydrogen ion concentration. . h 2 o + oh- Since ammonia is a weak base there will always be some undissociated am- monium hydroxide formed in the presence of ammonium ions and hydroxyl ions. The concentration of ammonium ions in the series of solution studied was the same in all, but the hydroxyl ion concentration varied with the acidity as explained above. Therefore the concentration of the undissociated ammonium hydroxide is dependent upon the hydrogen ion concentration. NH + + OH- NH OH Ammonium hydroxide is formed according to the following reversible re- action : NH 3 -f h 2 o nh 4 oh Since NH 3 is volatile, it follows from this equation that there must always be a trace of ammonia vapors above any solution, that contains undis- sociated ammonium hydroxide. The amount may be infinitesimal, but when it is continually removed as by steam distillation an appreciable amount appears in the distillate. Trimethylamine in Butter If trimethylamine is the cause of fishiness in butter, it must be present in such butter in greater quantities than in normal butter, and in such quantities that it can be determined quantitatively. Therefore samples of fishy butter both commercial and experimental, were analyzed for trimethylamine. The following procedure was used: One hundred grams of butter to be analyzed were extracted in a separatory funnel five times with hot water acidulated with hydrochloric acid at the rate of 50 cc. normal acid to the liter. The washings were separated from the fat into a beaker and further acidified with hydrochloric acid and then evaporated to a volume of 40 cc. This concentrated mixture was made up to 50 cc. with distilled water, and aliquot portions of 15 cc. used for a trimethylamine determination according to the method previously described. The results are given in Table X. The Fishy Flavor in Butter 27 TABLE X.— THE TRIMETHYLAMINE CONTENT OF FISHY BUTTER Sample number Description of sample Trimethylamine con- tent in parts per million parts of butter 1 Fresh butter None 2 Fresh butter i None 3 Fishy butter 32.65 Commercial sample No. 1 32.45 4 Fishy butter 22.37 Commercial sample No. 2 22.37 Fishy butter 5 Sample of experimental butter 35.97 furnished by Hunziker 35.52 Fishy butter 6 Experimental butter containing 23.72 0.4% acid and 0.1% Fe lactate 23.80 Fishy butter 7 Experimental butter containing 25.2 0.6% acid and 0.1% Fe lactate 25.1 Fishiness in Lecithin-Added Butter If lecithin is the source of the trimethylamine which seems to be the cause of fishiness in butter, then butter made from cream to which lecithin had been added should become fishy more readily than butter made from the same cream without lecithin added, provided there are no other factors that limit the amount of -trimethylamine produced regardless of the amount of lecithin present. In order to try this out experimentally pasteurized sweet cream was divided into three lots ; to one no lecithin was added, to the second one 0.1353% lecithin was added, and to the third 0.2706% lecithin was added. Each of these three lots was in turn divided into two portions; one was left untreated, to the other enough lactic acid was artifically added to increase the acidity to 0.5%. These six batches of cream were then churned separately. The butter was divided into two portions ; one was left un- salted, the other was salted to 3.0%. These two portions of butter were in turn divided into three portions ; one was normally worked, the second was overworked, and to the third hydrogen peroxide was added at the rate of 3 cc to 200 grams of butter. These butter samples were placed into bottles and stored in a refrigerator at 35 to 40° F. From time to time the butter was scored for flavor. The results are given in Table XI. 28 Wisconsin Research Bulletin 57 TABLE XI.— FISHINESS IN LECITHIN-ADDED BUTTER Leci- Acid Salt Age, Normally Over- Hydrogen thin days worked worked peroxide g 31 0 0 74 102 T3 31 31 Oily Oily Oily o Oj; 03^ 3% 74 Very oily Very oily Oily •J JTJ ^ C3 102 Oily & si. fishy Oily & si. fishy Oily & si. fishy c 31 Oily Oily Oily c 03 0% 74 CO ggS 102 o or. So 31 Slightly oily Slightly oily Oiiy o’ £ 3% 74 Slightly oily Oily Oily 'O « C/3 102 Oily Oily Oily T3 31 Slightly oily Slightly oily Oily GJ •r*o 0% 74 Slightly oily Oily Oily c 43 2 10 . «o 102 Oily and sweet Oily and sweet Oily '3 o 31 Oily Very oily Very oily 073 03 -O 1-4 03 3% 74 Very oily Very oily Very oily ►J 102 Slightly fishy Slightly fishy Fishy The results in Table XI show that there was a fishy flavor in all the salted, artifically soured samples of butter made from lecithin-added cream. The amount of lecithin added to the cream did not impart a very noticeable flavor to the fresh butter, but as Table XI shows, even after 31 days storage, oily flavors had developed. The development of the oily flavor was favored by the same factors that are known to favor the development of fishiness, viz. salt and acid. Since fishiness developed only in the lecithin added butter samples, these results lend support to the theory that trimethylamine is the immediate cause and lecithin is the mother substance of the fishy flavor in butter. The Fishy Flavor in Butter 29 Biological Agencies and Fishiness The results thus far presented demonstrate that lecithin undergoes a chemical decomposition with the formation of the fishy substance, trimethylamine, under conditions' that are known to favor the development of fishiness in butter. These facts logically lead to the conclusion that the development of fishiness in butter is, at least in part, due to the above reaction; but they do not necessarily exclude the possibility that in ad- dition biological factors are involved. The development of fishiness in butter has been attributed to biological agencies by a number of investigators, among these are : Storch x , Kirchner 2 , O’Callaghan 6 , u , 13 , 45 , 27 , 4Q . Piffard 9 , Harrison l0 , Sommerfeld l7 , Rahn, Brown and Smith l9 , Orla-Jensen 30 , Klein 33 , Fleischmann s8 , Sup- plee 5o , Brown, Smith, and Ruehle 56 , and Cusick 55 . Attempts to produce a fishy flavor in butter by bacterial inoculation have lead to conflicting results. 0’Callaghan 27 describes experiments in which he produced fishiness in butter by inoculating sterile cream with B. acidi lactici and Oidium lactis. However, Harding, Rogers, and Smith g were unable to produce fishiness by inoculation with pure cultures of Oidium lactis; and Rogers lg states further that many lots of butter be- come fishy in which Oidium lactis is absent both in the butter or in the cream from which it was made. Harrison l0 in studying different species of bacteria found in milk was able to produce fishy butter by inoculating the cream. No detailed experi- ments are described, however. Kirchner 2 states that oily, fishy and tallowy butter is formed by acid producing bacteria and molds, but he gives no experimental data. Rogers lg in studying creameries where outbreaks of fishy butter occurred found no unusual varieties of bacteria, nor was he able to produce fishy cream by inoculating with bacteria isolated at such creameries. Thom- son 2 0 , 21 a ^ so was una ble to obtain colonies productive of fishiness by making cultivations from very fishy butter. Reakes, Cuddie and Reid 2Q found no significant differences in the bacterial flora of fishy and normal butter. He also found that plugs of fishy butter when inserted into good butter did not cause fishiness. Brown, Smith and Ruehle 56 were unable to identify any special bacterial flora with fishiness. Hammer 43 isolated the organism, Bad. ichthyosmius , from a can of fishy evaporated milk, and was able to reproduce the flavor in milk, cream and evaporated milk, but not in butter either by direct inoculation or inoculating the cream before churning. He concluded that this organism is not of direct importance as a cause of fishiness in butter under practical conditions. Supplee 50 was able to produce fishy butter from cream that had been ripened 24 hours with either Bad. ichthyosmius or another fishy organism isolated from fishy butter. Similarly Cusick 55 was able to produce fishy butter from cream ripened with Bad. ichthyosmius, especially in those samples ripened in addition by B. lactis acidi. The bacteria decreased rapidly during storage, but less rapidly in the salted samples than in the unsalted Wisconsin Research Bulletin 5 7 30 . samples. He isolated the organism from salted fishy butter produced in the above manner at the end of the period of storage. It is thus that many of the attempts to reproduce fishiness bacteriologi- cally in butter failed. A few investigators, however, apparently seemed successful. In view of these conflicting results, a study was made of the production of trimethylamine from lecithin and from skim milk by bacteria. Decomposition of Lecithin by Bacteria In order to study the decomposition of lecithin by pure culture of various bacteria it was first necessary to prepare sterile lecithin emulsions. This immediately raises the question of a possible decomposition of the lecithin in the sterilizing process, and a formation of trimethylamine from this source. This possibility was investigated experimentally in the following manner: Lecithin emulsions were prepared and sterilized under conditions which did not allow any trimethylamine that may have been formed to escape. This was accomplished by placing the lecithin emulsion in the first of three bottles, connecting this by means of glass tubing to a second bottle which acted as a safety bottle to take care of any foaming from the first bottle, etc., and the second bottle was finally connected to a third bottle containing 20 cc. N/l sulphuric acid, the glass tubing used ended in the acid. Thus any trimethylamine that was formed during sterilization could not escape. The lecithin emulsions were sterilized for 30 minnutes at 15 pounds steam pressure, and cooled slowly. The lecithin emulsion was then made alkaline with potassium carbonate and aspirated as usual. From the titrations of the acid the amount of trimethylamine was determined. The results are given in Table XII. The results given in Table XII show that there is a slight decomposition of the lecithin in the sterilizing proces, but the decomposition is so slight TABLE XII.— THE DECOMPOSITION OF LECITHIN DURING STERILIZATION Lecithin emulsions Sterilized in cc. N/100 acid to neutralize volatile bases Ammonia equivalent in cc. N/100 acid Tertiary amine in cc. N/100 acid Per cent of tertiary amine of total possible 1 gram lecithin 50 cc. water Closed bottles 0.7 0.15 0.55 0.4167 0.5 gram lecithin 25 cc. water 25 cc. skimmilk Closed bottles 0.3 0.05 0.25 0.3570 0.5 gram lecithin 25 cc. water 50 cc. skimmilk Closed bottles 0.25 0.05 0.20 0.2856 0.5 gram lecithin 25 cc. water Open bottles 0.35 0.05 0.30 0.4284 The Fishy Flavor in Butter 31 that it does not interfere with the experiments on the biological decomposi- tion of lecithin. In order to determine whether lecithin emulsions aspirated directly without sterilizing yield trimethylamine, 25 cc. of 2 per cent lecithin emulsion were aspirated and titrated for trimethylamine. It was found that 0.14% of the total trimethylamine possible was liberated. Thus only a trace of trimethylamine is formed as the result of the sterilization. An attempt was made to grow organisms in pure lecithin emulsions. Lecithin was weighed out, 2.573 grams, emulsified, and made up to 100 cc. with distilled water. From this stock emulsion a series of samples were prepared by placing 1 cc. of the stock in 20 cc. distilled water. These sam- ples were sterilized for 30 minutes at 15 pounds pressure. Several such tubes were inoculated with Bad. ichthyosmius, and others with an organism isolated from fishy cream. Both cultures were furnished by W. B. Ham- mer. These sample bottles were then tightly corked and incubated at 28° C. for 44 hours. The amount of trimethylamine was then determined in the usual manner, and it was found that there was no trimethylamine in any of the samples. In order to furnish a more favorable medium for the organisms, samples were prepared similar to the above, but in which skim milk was used instead of water. One cc. of the above stock lecithin emulsion was placed into 20 cc. of skimmilk made from milk powder. After sterilizing these samples were inoculated with the same organisms as above, and incubated at 28° C. for 72 hours. Both these organisms first coagulated the milk to a soft curd and then digested the curd leaving a brownish liquid. No fishy smell could be detected from any of these samples, but there was a repugnant odor in all the inoculated samples which probably masked the fishy smell. The samples were aspirated and analyzed in the usual manner. The results are given in Table XIII. TABLE XIII.— TRIMETHYLAMINE PRODUCED FROM SKIMMILK AND LECITHIN BY BACTERIA Organism Total volatile bases cc. N/100 Ammonia expressed in cc. N/100 acid Tertiary amine ex- pressed in cc. N/100 acid Per cent tertiary amine of total possi- ble from lecithin Bad. ichthyosmius 52.65 26.0 26.65 799.5 Fishy cream organism. . . 57.1 28.9 28.2 845.5 The results given in Table XIII show that about eight times as much tertiary amine was produced as is theoretically possible from the lecithin known to be present in the experiment. This would indicate that tertiary amines can be formed from proteins by bacteria, unless we can account for the results in some other manner. Supplee 50 also found trimethylamine in cultures of Bad. ichthyosmius in skimmilk, to the extent of 94.4 parts per million. 32 Wisconsin Research Bulletin 57 The most plausible sources of these large amounts of tertiary amines are the proteins. However, in the bacterial decomposition of proteins pri- mary and secondary amines are more likely to be formed than tertiary amines. This suggests the possibility that the volatile bases in the determi- nation, not destroyed by formaldehyde and assumed to be tertiary amine, are in part primary and secondary amines. In order to investigate this possibility, primary and secondary methyl amine were prepared and purified. Then starting with known amounts of each of these in separate determinations the usual aspiration and titrations were applied. It was found that both the primary and secondary methyl amine could readily be aspirated and absorbed completely in the standard acid, but when the formaldehyde was added it was found that they were not acted upon completely even after 24 hours. Thus the primary and sec- ondary amines were not eliminated by the method used by us in Table XIII, and were included under the heading “tertiary amines.” In order to determine quantitatively the amount of tertiary amines found under conditions which can also produce primary and secondary amines it is then necessary to eliminate the primary and secondary amines in some manner. These were eliminated in subsequent determinations by means of nitrous acid. The method then used is as follows : The volatile bases were aspirated into N /I acid. To the N /I acid 1 cc. of 30% sodium nitrate solu- tion was then added, shaken, and after two minutes another cubic centi- meter of nitrite solution was added. The nitric oxide fumes are removed from this acid solution by aspirating and finally by rapidly bringing the solution to a boil. This acid solution which now contains only the tertiary amines and some ammonia, can then be aspirated and titrated in the former manner for ammonia and tertiary amines. In applying this method to a mixture of known amounts of ammonia and primary, secondary and tertiary amines, it was possible to account for 97.57% of the tertiary amine known to be present. Therefore this method may be regarded as an efficient method for the purpose. Using this more accurate method of analysis the experiments reported in Table XIII were repeated with slight modifications. Two series of samples were prepared, consisting of 20 cc. portions of solution of skim- milk powder, which had been extracted with alcohol and ether. To one series no lecithin was added, and to the other 2 cc. of a 2.257% lecithin emulsion were added to each sample. After sterilizing for 30 minutes at 15 pounds pressure, four samples of each series were inoculated with Bad. ichthyosmins, and four of each series with the fishy cream organism. The samples were then incubated for 72 hours at 28° C. The amount of tri- methylamine was then determined, making use of the method described above. The results are given in Tables XIV and XV. The results given in Table XIV show that only negligible amounts of trimethylamine are produced from lecithin free skimmilk by the two organisms studied. This demonstrates that the large amounts of trimethy- The Fishy Flavor in Butter 33 TABLE XIV.— TRIMETHYLAMINE FORMED FROM LECITHIN FR MILK BY BACTERIA Organism Sample number Cc. N/100 acid neu- tralized by volatile bases Ammonia expressed" in cc. N/100 acid Tertiary amine ex- pressed in cc. N /100 acid Milligrams of tertiary amine found per gram of total solids Bad. ichthyosmius . . . . 1 12.21 12.10 0.11 0.0361 2 12.09 11 .99 0.10 0.0328 3 12.09 11.98 0.11 0.0361 4 (spoiled) Average. . 0.107 0.0350 Fishy cream organism 1 13.34 13.21 0.13 0.0427 2 13.41 13.30 0.11 0.0361 3 13.10 12.98 0.12 0.0394 4 13.12 13.01 0.11 0.0361 Average. . 0.118 0.0383 TABLE XV.— TRIMETHYLAMINE FORMED FROM LECITHIN ADDED MILK BY BACTERIA Organism No. Cc. N/100 acid neu- tralized by volatile bases Ammonia expressed in cc. N/100 acid Trimethyl amine ex- pressed in cc. N/100 acid Milligrams of tri- methyl- amine found Trimethyl amine ex- pressed as per cent of total pos- sible Bad. ich- thyosmius. . . 1 10.67 10.13 0.54 0.3190 9.66 2 10.63 10.12 0.51 0.3013 9.12 3 10.72 10.23 0.49 0.2895 8.75 4 10.70 10.18 0.52 0.3072 9.29 Aver- age.. 0.51 0.3043 9 21 Fishy cream organism . . . 1 12.56 12.01 ' 0.55 0.3250 9.84 2 12.66 12.12 0.54 0.3190 9.66 3 12.51 11.95 0.56 0.3309 10.02 4 12.66 12.11 0.55 0.3250 9.84 Aver- age.. 0.55 0.3250 9.84 34 Wisconsin Research Bulletin 5 7 lamine reported in Table XIII were incorrect and due to the method of analysis not eliminating the primary and secondary amines. The results reported in Table XV show that appreciable amounts of tri- methy lamine are produced by the two organisms from milk containing lecithin. We must therefore conclude that these organisms can directly or indirectly decompose lecithin into trimethylamine. In all probability choline is first split off through enzymatic action or through the acid formed during these fermentations. All the cultures were decidedly acid after incubation. Or it may be that enough choline was hydrolyzed off during the process of sterilization, and that this choline was then used as a pabulum by the organisms. The work of Bordas and de Raczkowski lends support to such an explanation, for they found that milk sterilized in an autoclave for 30 minutes lost about 30% of its lecithin content (probably through hydrolysis). A series of samples of a casein solution was prepared and inoculated with Bad. ichthyosmias, and the fishy cream organism, but analyses using the nitrous acid method showed that there was no trimethylamine formed. From these results it seems that even where bacteria are involved in the production of fishiness, the source of the trimethylamine is the lecithin. In order to determine whether these two organisms would produce tri- methylamine from lecithin emulsions in skimmilk under conditions that are known to favor the development of fishiness in butter, attempts were made to grow them in emulsions containing 10% salt and 0.30% lactic acid. However, under, these conditions the organisms showed no growth as judged by the physical appearance of the cultures and the absence of growth on agar slants. This would indicate that while these organisms can produce trimethylamine under favorable conditions, they cannot be regarded as im- portant agencies in the production of trimethylamine in storage butter, because the very factors that favor the development of fishiness in storage butter, high salt, and high acid prevent their growth entirely. The Production of Trimethylamine From Hydrolyzed and Unhydrolyzed Lecithin The fact that the development of fishiness in butter and the chemical de- composition of lecithin into’ trimethylamine are favored in acid conditions suggests that these changes require as their first step, the hydrolysis of the lecithin. Further, several investigators have found that choline is readily decomposed into trimethylamine, other amines and ammonia by several microorganisms. It was thus decided to study the bacterial and chemical production of trimethylamine from hydrolyzed and unhydrolyzed lecithin. Preparation of the lecithin emulsions Of the dried lecithin, 31.5165 grams were weighed out and emulsified with 3/20 N lactic acid solution to a volume of 1700 cc. This emulsion was The Fishy Flavor in Butter 35 divided into two portions of 800 and 900 cc. The 900 cc. portion was im- mediately made neutral to phenolphthalein with N/l sodium hydroxide. Its final volume was 1250 cc. The 800 cc. portion was hydrolyzed. This 800 cc. portion made up in 3/20 N lactic acid was hydrolyzed for 13 hours according to the method of Mahlengreau and Prigent 113 . A dark oil separated out at the surface and a creamy, fatty substance gathered on the sides of the flask as the hydrolysis proceeded, leaving the liquid ulti- mately clear with a yellow color. This clear yellow liquid was filtered off, the flask washed out with distilled water and the washings filtered. The solid matter left on the filter paper was washed with distilled water and the several washings added to the filtered yellow liquid. This liquid was then neutralized to phenolphthalein giving 1750 cc. of a golden yellow solu- tion. Choline determinations were then made on this hydrolized lecithin solution and also on the unhydrolyzed lecithin solution. Choline determinations Tosaka Kinoshita 124 has subjected all the methods of choline determina- tion to a critical test and found that the platinum chloride and the mercuric chloride methods were the best. Brieger ll5 also studied the mercuric chloride method of determining choline and found that six molecules of mercuric chloride combine with one molecule of choline chloride to form a double salt, thus : C 5 H ]4 NO Cl - 6Hg Cl 2 H 2 0 Gulewitsch llg also assigned this formula to the double salt. Moruzzi 117 , Morner llg , and Schulze 119 also studied and used this mercuric chloride method successfully. This method was therefore used in our work. Duplicate samples of 40 cc. of the unhydrolyzed emulsion and duplicate 100 cc. and 40 cc. samples of the hydrolyzed lecithin were acidified with TABLE XVI. — THE AMOUNT OF CHOLINE IN THE HYDROLYZED LECITHIN EMULSIONS Sample 1 No. j 1 Volume 1 of sample used 1 1 I Weight of i choline mercuric | chloride ob- tained I 1 Weight of choline deter- mined ! 1 Per cent of choline produced of total possible 1 1 ! 1 1 i 100 i I Grams 1.682 1 1 i Grams 1 0.1132 1 89.18 2 100 1.6760 1 1 0.1129 1 88.82 3 40 1 0.6705 1 0.0451 88.76 4 40 0.6714 0.0452 88.90 Average 70 1.1750 0.0891 88.92 36 Wisconsin Research Bulletin 57 dilute hydrochloric acid and rapidly evaporated to dryness over a water bath, dried in a vacum desiccator over sulphuric acid, extracted with absolute alcohol and filtered, the filter paper being washed with absolute alcohol to remove all the choline chloride. The filtered liquid was concentrated to about 30 cc. and a saturated solution of mercuric chloride in absolute alcohol was added to each solution. These solutions were then allowed to stand over night. The next day the short prismatic crystals of the double salt of choline and mercuric chlorides were filtered off and washed four times with absolute alcohol, and dried in a vacuum desiccator to constant weight. No choline could be determined from the unhydrolyzed lecithin emulsions but the hydrolyzed lecithin emulsions yielded large amounts. The results are given in Table XVI. The results in Table XVI show that 88.92% of the choline was liberated in the method of hydrolysis used. Heffter 12Q using lecithin extracted from liver « tained only 25% of the theoretical amount of choline, Moruzzi 117 obtains only 77% from egg lecithin; Erlandsen 121 42% from heart lecithin; and Osborne and Wakeman 97 39.5% from milk lecithin. In all these cases the alkaline method of hydrolysis was used. Maclean 122 was able to obtain 92% of the theoretical choline from lecithin. Mahlengreau and Pilgent 113 using the acid method of hydrolysis found as high as 96.4% choline from lecithin. Coriat 123 claims that he obtained the theoretical amount of choline from brain lecithin. Preparation of Samples for Study of Chemical and Bacteri- ological Decomposition of Hydrolyzed and Unhydrolyzed Lecithin To study the chemical decomposition of the lecithin into trimethylamine samples were prepared in which the following conditions and combinations of them were studied : acid, salt, copper sulphate, ferric lactate, and hydrogen peroxide. For the study of the bacterial decomposition two cultures were used: Bad. ichthyosmius, and the fishy cream organism. All of the samples were so measured out that each contained 0.25 gram of lecithin in a volume of 186 cc. The samples were all incubated for 14 days at a temperature of 28° C., and at the end of this time were analyzed for trimethylamine using the earlier method with the nitrous acid modifica- tion. Table XVII gives a description of the samples and the amount of trimethylamine produced. From Table XVII it is very evident that hydrolysis of the lecithin in- creases the trimethylamine yield markedly. In general the amount of tri- methlyamine produced from the hydrolyzed lecithin is twice as high as the amount produced from the unhydrolyzed lecithin. In general the presence of salt caused a marked increase in trimethylamine in the unhydrolyzed samples, but only a slight increase in the hydrolyzed The Fishy Flavor in Butter 37 TABLE XVII— PERCENTAGE DECOMPOSITION OF UNHYDROLYZED- AND HYDROLYZED LECITHIN INTO TRIMETHYLAMINE Sample No. A Description of sample: 0.25 gram lecithin in 186 cc. solu- tion containing: Uni hydrolyzed lecithin Hydrolyze lecithin d Sterile samples HgCh 0.1% Bad. ichthy- os mills Fishy cream organism Sterile samples HgCh 0.1 % Bad. ichthy- os mius Fishy cream organism 1 Nothing added-control 1.46* 4.23 4.39 3.04 12.27 12.91 2 0.5% lactic acid 1.66 3.78 3 0.5% lactic acid and pasteur- ized 140° F., 30 min 1.98 1.95 1.98 3.16 3.13 3.10 4 0.5% lactic acid and pasteur- ized 180° F., 1 min 1.66 1.66 1.63 3.07 3.11 3.09 5 15% salt 1.50 3.12 6 Copper sulphate 0.1%, hydro- gen peroxide 2 cc 3.19 4.84 7 Ferrous lactate 0.1%, hydro- gen peroxide 2 cc 2.34 4.36 8 Lactic acid 0.5%, salt 15%. . . . 4.20 3.86 9 Lactic acid 0.5%, copper sul- phate 0.1%, hydrogen perox- ide 2 cc 6.77 15.66 10 Lactic acid 0.5%, ferrous lac- tate 0.1 %, hydrogen peroxide 2 cc 6.64 14.95 11 Salt 15%, copper sulphate 0.1%, hydrogen peroxide 2 CC. . 3.06 5.04 12 Salt 15%, ferrous lactate 0.1%, hydrogen peroxide. 2 cc 3.00 4.58 13 Lactic acid 0.5%, salt 15%, copper sulphate 1 %, hydro- gen peroxide 2 cc 12.91 16.14 14 No. 13 pasteurized 140° F. 30 min 13.66 16.57 15 No. 13 pasteurized 180° F. 1 min. . . . 12.85 15.98 16 Lactic acid 0.5%, salt 15%, fer- rous lactate 0.1%, hydrogen peroxide 2 cc. . 12.13 15.15 17 Sucrose 1.0%. 11.80 12.48 19.67 21.17 ♦Per cent of trimethylamine of total theoretically possible. 38 Wisconsin Research Bulletin 57 samples. If we compare the salted samples with the corresponding unsalted samples (compare samples 5 and 1, 8 and 2, 11 and 6, 12 and 7, 13 and 9, 16 and 10), we find that in the case of the unhydrolyzed lecithin the presence of the salt caused on an average an increase of 2.45% in the trimethylamine yield as compared with an average increase of only 0.34% in the hydrolyzed lecithin solutions. This suggests the hypothesis that the stimulating effect of the salt in the trimethylamine production is due to its solvent action; in the unhydrolyzed lecithin samples the salt brings the lecithin more nearly in true solution so that it can be acted upon more readily, while in the hydrolyzed lecithin samples the choline which gives rise to the tri- methylamine is already in solution. If we compare the samples with acid with the corresponding samples with- out acid, we find that acid caused a big increase in the trimethylamine yield. (Compare samples 2 and 1, 8 and 5, 9 and 6, .10 and 7, 13 and 11, 16 and 12). In the unhydrolyzed lecithin samples the presence of acid caused an average increase of 4.96 in the per cent of trimethylamine, while in the hydrolyzed samples it caused an average increase of 7.43. The method of pasteurization did not seem to make an appreciable dif- ference in the amount of trimethylamine produced. It may be significant, however, that the pasteurization at 140° F. for 30 minutes gave slightly higher yields than the samples pasteurized at 180° F. for 1 minute, especially in the unhydrolyzed series. (Compare samples 2, 3 and 4, and 13, 14 and 15.) This would lend support to the hypothesis that in pasteurizing at 140 or 145° F. for 30 minutes there is more choline hydrolyzed off than in pasteurizing at 180° for 1 minute. It will be noted that even in the bacterial decompositions the largest yields of trimethylamine were obtained in the hydrolyzed series, indicating that here too choline is first formed by hydrolysis. Pasteurized samples 3 and 4 were not acted upon by the two organisms studied, undoubtedly due to the acidity of the medium. These samples showed no growth on agar slants. General Discussion It has been shown in the early part of this bulletin by means of experi- mental work and a review of the literature that the conditions that favor the development of fishiness in storage butter are high acid, high salt, and oxidation produced either by overworking or by the presence of iron or copper salts. It has further been shown in this work that lecithin will undergo a purely chemical decomposition yielding trimethylamine as one of the products. Further, the optimum conditions for the production of this fishy flavored substance, trimethylamine, from the lecithin emulsions are identically the same as the optimum conditions for the development of fishi- ness in storage butter. These results lend strong support to the theory that fishiness in butter is due to trimethylamine formed as a result of a purely chemical decomposition of lecithin. The Fishy Flavor in Butter 39 Analyses taken from the literature show that butter contains sufficiently large amounts of lecithin to produce the amounts of trimethylamine neces- sary to cause the fishy flavor. The analyses indicate that most butter sam- ples contain enough lecithin to produce from 50 to 100 parts of trimethyla- mine per million parts of butter. In this work 100 parts of trimethyla- mine lactate per million parts of butter caused a decided fishy flavor. Supplee found that 85 parts of trimethylamine or its salts caused a char- acteristic fishy flavor in most cases. Undoubtedly even somewhat smaller amounts will produce fishiness. The samples of fishy butter analyzed (Table X) showed from 23 to 36 parts of trimethylamine per million. Un- doubtedly most butter samples contain enough lecithin to produce fishiness, but there is also the possibility that some samples do not contain enough lecithin to become fishy. Such samples would not become fishy even when subjected to conditions which we would ordinarily expect to produce fishi- ness. It is also possible that the period of lactation and the ration may have some influence upon the lecithin content of milk and hence butter. This might serve as a basis for explaining the observation made by Rogers that fishiness is never produced under winter conditions. That lecithin is the source of fishiness is further indicated by the fact that fishiness developed more readily in the lecithin added butter samples (Table XI). While these observations lead to the conclusion that the main source of fishiness is lecithin and that fishiness is produced from it by a chemical decomposition yielding trimethylamine, they do not exclude the possibility that biological agencies may be partly responsible for the development of fishiness either from lecithin or possibly even from some other substances. The production of trimethylamine from proteins would involve a synthetic process. While this does not exclude the possibility that part of the tri- methylamine may be formed by bacterial action, from proteins, it makes it less likely. Experiments with Bad. ichthyosmius and the fishy cream organism showed that these two organisms produced only very small traces of the trimethylamine from skimmilk and casein solutions. The amounts produced were within the limits of experimental error of the analytical method, and we may say that these organisms produced no trimethylamine. This finding is contrary to that of Supplee 50 who reported that Bad. ichthyosmius produced 94.4 parts of trimethylamine per million parts of skimmilk. As shown in this work these results are incorrect on account of the fact that the analytical method did not eliminate the primary and sec- ondary amines. When these are eliminated no trimethylamine is found unless lecithin is present in the culture. It was found that both of these two organisms could produce trimethylamine from lecithin under the proper conditions. We must conclude that lecithin is the logical source, and thus far the only demonstrated source of trimethylamine in butter. It was found that the two organisms studied could produce trimethylamine from lecithin under favorable conditions, but they failed entirely to grow under conditions that are favorable for the development of fishness, i. e., 40 Wisconsin Research Bulletin 57 salt and high acid. We must concur with Rogers gl in the opinion that the high concentration of salt in the butter brine and the low temperatures of storage at which fishiness develops effectually exclude the possibility of fishiness being caused by microorganisms. The few investigators who claim they have produced fishiness in butter by means of micro-organisms either did not exclude the factors that favor a chemical decomposition of lecithin, or used conditions that are not normal. O’ Caliaghan in producing fishiness by means of Oidium lactis used B. lactis acidi with it and thus produced an acidity that would favor the purely chemical decomposition. Moreover, Rogers lg , and Reakes, Cuddie and Reid 26 failed to corroborate O’Callighan’s findings. Both Supplee 50 and Cusick 55 produced fishiness in butter by means of Bad. ichthyosmius, but in order to do so they developed the organisms in the cream before churning far in excess of what could be expected under normal conditions. In addi- tion they found that this organism produced fishiness in their experimental butter samples- most readily in the neutralized samples. This again is con- trary to the normal development of fishiness in butter. Opposed to the few successful reports we have a number of investigators reporting failure in attempts to produce fishiness by bacterial inoculation directly into the butter or into the cream, among these are Harding, Rogers and Smith 8 , Rogers lg , Reakes, Cuddie and Reid 26 , Hammer 43 , etc. Any enzymatic theory for the development of fishiness in butter seems untenable in view of the experience of Rogers and others that pasteurization at 180° F. does not necessarily prevent the fishy flavor. These considerations eliminate biological agencies as the direct cause of fishiness in storage butter. The conclusion is that the main cause is the chemical decomposition of the lecithin into trimethylamine. The role of the Various Factors Concerned in the Development of Fishiness From the experimental evidence presented in this bulletin and from the literature, we may advance hypotheses to explain the role of the various factors concerned in the development of fishiness. Why Is Fishiness More Common in Salted Than in Unsalted Butter? 1. Salt brings out flavors. In general salt in butter intensifies the flavors. This also holds true for fishiness. When trimethylamine was incorporated into unsalted butter and then part of this butter salted, it was found that the salted portion had a more intense fishy flavor. This is in accord with the findings and the opinions of 0’Callaghan 61 , McKay and Larsen 60 , Rogers lg , and Washburn and Dahlberg 42 . Supplee 50 found that the addition of sodium chloride to trimethylamine salts of fatty acids caused the precipitation of the sodium soap and the formation of trimethylamine hydrochloride. This reaction may . explain in part the intensifying effect of salt on the fishy flavor. The Fishy Flavor in Butter 41 However, besides this intensifying effect of the salt, it must also be in- volved in the actual production of the flavor, because fishiness is not pro- duced by simply incorporating salt into an unsalted sample of butter, the corresponding salted sample of which has become fishy in storage. 2. Salt dissolves- lecithin. Lecithin is soluble in common salt solutions. By bringing the lecithin of the butter into solution in the brine of the butter, the salt accelerates the hydrolysis and oxidation of the lecithin into trimethylamine. This ex- planation is fully supported by Table XVII where salt decidedly increased the trimethylamine yield over that of the corresponding unsalted emulsions in the unhydroylzed lecithin series, but caused only a negligible increase in the hydrolyzed series where the choline of the lecithin was already in solu- tion. Our conclusion is that the main role of salt in the development of fishiness is to bring the lecithin of the butter into solution in the brine. 3. Salt lowers the freezing point of the brine . Salt keeps the brine of the butter under storage conditions in an unfrozen state, in a physical condition that is more favorable for chemical reactions than the frozen condition would be. This undoubtedly is partly responsible for the stimulating effect of the salt on the development of fishiness. 4. Salting necessitates additional working of the butter. Salting, especially when dry salting is used, necessitates additional working to distribute the salt evenly in the butter. This additional working is favor- able to the development of fishness as will be explained under the discussion of the role of overworking. 5. Fishiness in unsalted butter. There is no evidence to show that salt takes part chemically in the pro- duction of trimethylamine. Its action seems to be entirely physical as ex- plained above; hence it may be expected that fishiness may occasionally occur in the absence of salt. In such cases, however, the corresponding salted samples will develop fishiness sooner and of a greater intensity. Why Does an Acid Condition Favor the Development of Fishi- ness in Storage Butter? 2. Acids favor the hydrolysis of lecithin. That lecithin is hydrolyzed by acids, strong and dilute, has been shown by a number of investigators: Brieger ll5 , Riedel 124 Moruzzi 117 , Molen- greau and Prient 113 , Diakonow 125 , Hefftner 126 , and Gilson 127 ; Mahlengreau and Prient obtained 96.03% of the nitrogen from lecithin by hydrolyzing for 72 hours on a water bath in a N /10 acetic acid solution. Diakonow, Gilson, 42 Wisconsin Research Bulletin 57 Hefftner, and Moruzzi showed that dilute acids hydrolyzed lecithin to some extent even in the cold. Hammersten and Heddin 12g state that lecithin is slowly decomposed by dilute acids. It has been shown in this work that hydrolysis (Table XVII) greatly accelerates the production of trimethylamine from lecithin. Our conclusion is that part of the role of the acid in the development of fishiness is that it hydrolyzes the lecithin. 2. Acidity aids oxidation in butter. It has been found by Dyer 39 and Rogers et al 29 , that the rate of oxidation in butter is proportional to the acidity. This was determined by measuring the free oxygen content of butters of different acidity. This is in harmony with our results in Table XVII, where we found that the presence of acid caused an increase of 4.96 in the per cent of trimethylamine in the unhy- drolyzed lecithin Emulsions, while in the hydrolyzed samples it caused an average increase of 7.43%. Our explanation of these results is that the acidity favors both the hydrolysis of lecithin and the oxidation of the choline into trimethylamine. The rate at which the choline becomes available for oxidation by the hydrolysis of the lecithin by the dilute acid is slower than the rate of oxidation of the choline to trimethylamine. For that reason, acid caused a greater increase in the trimethylamine yield in the hydrolyzed lecithin series where the choline is already free, than in the unhydrolyzed lecithin series where the amount of choline available for oxidation was limited by the rate of hydrolysis. 3. Acids in the cream cause the absorption of metals. The solubility of the metals with which the cream comes into contact is proportional to the acidity of the cream. It has been found by Weigmann 22 , Marcas and Hugge 69 , Kooper 7l , and Rogers et al 29 , that an appreciable amount of iron and copper may be dissolved by sour cream in contact with utensils. The manner in which these metals cause fishiness will be dis- cussed later. Is Trimethylamine Formed By Oxidization or Hydrolysis? The work of a number of investigators shows that choline yields traces of trimethylamine quite readily on boiling in neutral, acid and alkaline solutions. It is not certain by what process this trimethylamine is formed, but the work of Wurtz 129 indicates that it is hydrolytic. Wurtz found that concentrated choline solutions yielded trimethylamine and ethylene glycol on boiling. While we must consider it possible that the development of fishiness in butter, the production of trimethylamine from choline, is hydrolytic, the indications are that it is actually a process of oxidation. The observations of a number of investigators that fishiness in butter is favored by over- The .fishy Flavor in Butter 43 working indicates oxidation. Our results given in Tables VII, VIII and XVII show that in the chemical decomposition of lecithin into trimethyla- mine oxidizing agents such as hydrogen peroxide, pure oxygen, and catalytic agents increased the production of trimethylamine very decidedly. If we conclude that the production of trimethylamine is an oxidative process, we can then offer a rational hypothesis to explain the effect of overworking and the effect of iron and copper salts. Why Does Overworking of Butter Favor the Development of Fishiness? 1. Ovenvorking may increase the air content of butter. An increase in the oxygen content of butter suggests itself at once as the explanation for the effect of overworking. Rogers lg actually found an increase of 10% by overworking by means of a spatula. He also found a small but distinct increase in normally overworked butter. Later Rogers et al 2Q showed that overworked butter does not necessarily contain more air than normally worked butter. Hunziker 5l is of the same opinion. It is certain that the effect of overworking cannot be explained entirely on the basis of an increase in oxygen content. While we must still consider this factor as being of possible minor importance, we must look for other factors. 2. Overworking distributes the incorporated air more thoroughly. The most plausible explanation for the fact that overworking favors the development of fishiness in butter is that in overworking the grain of the butter is destroyed and the air more finely divided so that a larger area of the butter is in contact with the air. If the diameter of the air bubbles is reduced one-half the surface area of the bubbles is doubled. This should cause an increase in the rate of oxidation, and Rogers et al 29 has actually found that in an overworked sample during cold storage the decrease in the oxygen content of the sample was about 50% greater than in a normally worked sample. 3. Overworking makes the solution of the lecithin in the brine more complete. Overworking distributes the brine of the butter more thoroughly, as well as the air, and favors the solution of the lecithin in the brine. In that way overworking brings about a thorough mixture and contact of all the factors concerned, viz.: lecithin, acids, salt, oxygen, and catalytic agents such as iron and copper. This naturally leads to an acceleration in the production of trimethylamine from the lecithin. 44 Wisconsin Research Bulletin 57 Why Do Iron and Copper Salts Favor the Development of Fishiness? It is a well known fact that metals that can readily be changed from a higher to a lower valency and vice versa will act as oxygen carriers, i. e., they will catalyze oxidation reactions. Thus it may be expected that iron and copper will catalyze the production of trimethylamine from lecithin. That iron and copper salts actually catalyze the oxidation of organic substances has been demonstrated hy a number of investigators. Neuberg and Blumenthal 130 were able to isolate and identify isovaleric aldehyde as an oxidation product resulting from the action of ferrous sulphate and hydrogen peroxide on gelatine. Orgler 131 produced a strong fruity odor by oxidizing egg albumin with copper sulphate and hydrogen peroxide; he identified acetone in the distillate from such a mixture. Rogers et al 29 found a substance which gave a marked iodoform test in the distillate from skimmilk to which ferrous sulphate had been added and allowed to stand at room temperature for 12 days. Pitz 132 produced pungent dis- agreeable compounds by the action of iron sulphate, and hydrogen peroxide in salt solutions of casein and iact-albumin. Hunziker and Hosman 72 demonstrated that copper salts could cause oxidation in butter which would have a deleterious effect upon the flavor and keeping quality. Why Does Pasteurizing Tend to Eliminate Fishiness? It has been shown that pasteurization generally tends to eliminate fishiness in storage butter. In looking for an explanation the first thought that suggests itself is that this is due to the destruction of the bacteria. However, his idea is not in accord with the well supported conclusion that fishiness is not caused by biological agencies. Moreover, it has been shown that pasteurization at 180° F. for one minute, although just as efficient in the destruction of bacteria as pasteurization at 145° F._ for 30 minutes, is not as efficient in preventing fishiness. Our conclusion is that the effect of pasteurization in preventing fishiness is not due to the destruction of bacteria. 1. Pasteurisation May Affect the Solubility of Lecithin. Thunberg 133 , Warburg and Meyerhof 134 , Maclean 135 , and Mathews 136 have found that lecithin readily absorbs oxygen and becomes less soluble. Whether the solubility of lecithin in brine is affected in this way is not stated, but in all probability it is. Such an oxidation of lecithin can take place more readily at 145° F. in 30 minutes than at 180° F. in one minute, since at 145° F. the cream is continually agitated and exposed to the air for a comparatively longer period of time. Thus it may be that in this way pasteurization, making the lecithin less soluble and less available for de- composition, reduces the amount of trimethylamine formed to a minimum. The Fishy Flavor in Butter 45 2. Pasteurisation Reduces the Lecithin Content of the Butter. References cited earlier in the bulletin show that lecithin can be hydrolyzed by dilute and weak acids such as acetic and lactic acids. It is then to be expected that pasteurization at 145° F. for 30 minutes owing to the longer duration of heating will hydrolyze more of the lecithin than will pasteuriza- tion at 180° F. for 1 minute, and that the more acid the cream the more lecithin will be hydrolyzed. Bordas and Raczkowski g3 - 87 have shown that the lecithin content was reduced 14% of the original content in the raw milk by heating for 30 minutes at 60° C. If sour milk had been used undoubtedly the hydrolysis would have been even higher. Cusick 55 has shown that butter made from pasteurized high acid cream is lower in lecithin than the unpasteurized samples. Supplee 50 found that the lecithin content of butter made from pasteurized ripened cream was 0.0433% as compared with 0.0723% in the butter made from the same cream unpas- teurized. Dornic and Daire 99 found similar results. It is thus seen that pasteurization at 145° F. for 30 minutes actually decreases the lecithin content of the butter. In the pasteurization the lecithin is partly hydrolyzed liberating choline, the mother substance of the trimethylamine. Choline is very soluble in water and will therefore be lost in the buttermilk, thus reducing the amount of trimethylamine that can be made even under com- plete decomposition of the remaining lecithin to such a point that it will not become evident as a fishy flavor. Thus our conclusion is that pasteurization tends to eliminate fishiness first because it reduces the lecithin content of the butter, and second, because the lecithin that remains unhydrolyzed in the butter is less soluble and will therefore decompose less readily. Summary 1. Fishiness is one of the most common of the storage butter defects, and on account of its offensive nature, it causes severe losses due to the resulting cut in the price. 2. The conditions that favor the development of fishiness in butter are : — (1) A high acidity in the cream. (2) High salt content in the butter. (3) Overworking of the butter. (4) The presence of iron or copper salts. 3. The theory has been advanced that the fishy substance is trimethylamine produced from lecithin in the butter. This theory has not been generally accepted, nor is there any agreement on the agency which is supposed to cause the decomposition of the lecithin into trimethylamine. 4. The experiments here described have demonstrated that lecithin will undergo a purely chemical decomposition at room or incubator tempera- tures. The conditions that caused the greatest yield of trimethylamine from the lecithin coincide with the conditions that are known to favor the development of the fishy flavor in butter. 46 Wisconsin Research Bulletin 57 5. The addition of trimethylamine lactate to butter at the rate of 100 parts per million of butter caused a distinct fishy flavor. 6. Trimethylamine was isolated from the samples of fishy butter examined. 7u Butter made from cream to which lecithin had been artificially added became fishy more readily than the untreated butter made from the same lot of cream. 8. Bad. ichthyosmius and another organism that had been isolated from fishy cream produced trimethylamine from lecithin under favorable con- ditions, but failed entirely to grow in the presence of salt and acid in concentrations such as would be found in the brine of butter. With the additional inhibiting effect of low temperatures in storage, it is very unlikely that bacteria are the cause of fishiness. 9. The above fwo organisms failed to produce trimethylamine from skim- milk and casein solutions. 10. The production of trimethylamine was more rapid from hydrolyzed lecithin emulsions than from the unhydrolyzed lecithin emulsions, both by the chemical decomposition and by means of the above two organisms. 11. The conclusion is that the development of fishiness in storage butter is due to the chemical decomposition of the lecithin normally present in butter. 12. Salt favors the development of fishiness because: (1) Salt intensifies flavors in general. (2) Salt brine is a good solvent for lecithin. (3) Salt lowers the freezing point of the brine. (4) Additional working is required to incorporate the salt. 13. Acid favors the development of fishiness because : (1) Acids favor the hydrolysis of the lecithin. (2) Acids favor the oxidative processes in butter. (3) Acids cause the cream to dissolve iron and copper from the utensils. 14. Overworking favors the development of fishiness because: (1) Overworking may increase the air content of the butter. (2) Overworking distributes the incorporated air more thoroughly.. (3) Overworking makes the solution of the lecithin in the brine more complete. lo. Iron and copper favor the development of fishiness in butter because they act as catalysts in the oxidation of organic substances. 16. Pasteurization tends to eliminate fishiness because: (1) During pasteurization the lecithin readily absorbs oxygen and becomes lqss soluble. (2) Pasteurization causes the hydrolysis of the lecithin to some extent; the products of hydrolysis are lost in the butter- milk, so that the resulting butter cannot possibly yield as much trimethylamine as the unpasteurized butter. The Fishy Flavor in Butter 47 BIBLIOGRAPHY 1. Storch, W. — Milch Zeitung, p. 304, April 16, 1890. 2. Kirchner, W. — Handbuch der Milchwirtschaft, (Paul Parey, Berlin). 3rd edition, p. 352. 3. Weigmann, H. — Milch Zeitung 20, p. 1019, 1891. 4. Weigmann, H. — Molkerei Zeitung (Berlin) 1892, p. 442. 5. Oliver, John — “Milk* Cheese and Butter” (Crosby Lockwood & Son, London) p. 304, 1894. 6. O’Callaghan, M. A. — “The Cause of Fishiness” (Pamphlet issued by Dep’t of Agriculture, New South Wales) 1899. 7. duRoi (Prenzlau) — Milch Zeitung, 9, p. 134, 1900. 8. Harding, H. A., Rogers, L. A., and Smith, G. A. — N. Y. (Geneva) Agr. Exp. Sta. Bui. 183, pp. 179-181, 1900. 9. PifTard, H. E. — N. Y. Produce Review and American Creamery, Nov. 13, 1901, p. 20. 10. Harrison, F. C. — Ontario Agr. College, 27th annual report, p. 78, 1901. 11. O’Callaghan, M. A. — Agr. Gazette, New South Wales, 12, pp. 341-346, 1901 12. Willoughby, Edw. — “Milk, Its Production and Uses” (Charles Griffin & Co., Ltd., London) p. 51, 1903. 13. O’Callaghan, M. A. — Agr. Gazette, New South Wales, 15, p. 257, 1904. t 14. Gray, C. E., and McKay, G. L. — U. S. D. A., Bureau of Animal In- dustry Bui. 84, 1906. 15. O’Callaghan, M. A. — Agr. Gazette, New South Wales, 18, p. 223, 1907. 16. Kirchner, W. — “Handbuch der Milchwirtschaft” (Paul Parey, Berlin) 5th edition, p. 394. 17. Sommerfeld, Paul — “Handbuch der Milchkunde” (J. F. Bergmann, Weisbaden) p. 641, 1909. 18. Rogers, L. A. — U. S. D. A., Bureau of An. Ind., Circular 146, 1909. 18a. Rogers, L. A., and Gray, C. E. — U. S. D. A., Bureau of Animal In- dustry Bui. 114. 19. Rahn, O., Brown, C. W., and Smith, L. M. — Mich. Agr. Exp. Sta. Tech. Bui. 2. 20. Thomson, G. S. — “Milk and Cream Testing and Grading Dairy Prod- ucts” (Crosby Lockwood & Son, London) pp. 192-199. 21. Thomson, G. S. — “British and Colonial Dairying” (Crosby Lockwood & Son, London), p. 136, 1913. 22. Weigmann, H. — “Mykologie der Milch” (M. Heinsius, Leipzig), p. 124, 1911. 23. Davis, L. M. — California Agr. Exp. Sta. Circular 60, 1911. 24. Davis, L. M.— U. S. D. A., Farmers’ Bui. 499, 1912. 25. Rogers, L. A., Thompson, S. C., and Kiethley, J. R. — U. S. D. A., Bureau of An. Ind. Bui. 148, 1912. 48 Wisconsin Research Bulletin 57 26. Reakes, C. J., Cuddie, D., and Reid, H. A. — The Journal of the Dep’t of Agr., New Zealand, Vol. 4, No. 1, pp. 1-6, 1912. 27. O’Callaghan, M. A. — “Dairying in Australasia” (Angus & Robertson, Ltd., Sydney), pp. 586-591. 28. Steinhoff, J. W. — Ontario Dep’t. of Agr., Annual Report, Vol. 1, 1913. 29. Rogers, L. A., Berg, W. N., Potteiger, C. R., and Davis, B. J. — U. S. D. A., Bureau of An. Ind. Bui. 162. 30. Orla-Jensen, S — “Die Bakteriologie in der Milchwirtschaft,” (G. Fischer, Jena), p. 123. 31. Rogers, L. A. — Proc. 13th Annual Meeting, Wis. Buttermakers’ Association, p. 70, 1914. 32. Snyder — (Cited from Rogers 31 ). 33. Klein, J. — “Milchwirtschaft,” (Paul Parey, Berlin), p. 238, 1914. 34. Ernst, W. — “Milk Hygiene,” (translated by Mohler and Eichhorn; Publishers, Alex Eger, Chicago), p. 137 and 177. 35. Rogers, L. A.— Milk Dealer, 310: 10-12, 1914. 36. Lewkowitsch, J. — “Chemical Technology and Analysis of Oils, Fats and Waxes,” (MacMillan & Co., New York), Vol. 2, p. 798, 1914. 37. Hunziker, O. F. — 28th Annual report, Purdue Agr. Exp. Sta. p. 39. 38. Fleischmann, W. — “Lehrbuch der Milchwirtschaft” (Paul Parey, Berlin), p. 322, 1915. 39. Dyer, D. C. — U. S. D. A., Jour, of Agr. Research, 6, pp. 927-952, 1916. 40. O’ Callaghan, M. A. — Chicago Dairy Produce, Apr. 25, 1916, p. 8. 41. O’Callaghan, M. A. — Agr. Gazette, New South Wales, 27, p. 860, 1916. 42. Washburn, R. M., and Dahlberg, A. C. — Jour, of Dairy Science, 1, p. 114, 1917. 43. Hammer, B. W. — Iowa Agr. Exp. Sta., Res. Bui. 38. 44. Klein, L. A. — “Principles and Practice of Milk Hygiene,” (J. B. Lip- pincott, Philadelphia), p. 27 and 67, 1917. 45. Ericson, Elov — N. Y. Produce Review and American Creamery, 48, p. 594, 1918. 46. Washburn, R. M., and Holmes, E. J. — (Cited from Ericson 45 ). 47. Bouska, F. W., and Washburn, R. M. — (Cited from Ericson 4 _). 48. Rogers, L. A. — N. Y. Produce Review and American Creamery, Vol. 47„ p. 804, March 26, 1918. 49. Washburn, R. M. — N. Y. Produce Review and American Creamery, Vol. 48, p. 1004, 1919. 50. Supplee, G. S. — Cornell Agr. Exp. Sta. Memoir 26, 1919. 51. Hunziker, O. F. — “The Butter Industry,” p. 209, p. 487, pp. 489. 52. McKay, G. L. — The Breeders’ Gazette, March 4, 1920, p. 600. 53. Brown, A. M. — Agr. Gazette, New South Wales, 31, pp. 490-94, 1920. 54. McKay, G. L. — The Breeders’ Gazette, July 29, 1920, p. 162. 55. Cusick, J. T. — Cornell Agr. Exp. Sta. Memoir 30, 1920. The Fishy Flavor in Butter 49 56. Brown, C. W., Smith, Lulu M., and Ruehle, G. L. A. — Journal of Dairy Science, 3, pp. 374-407. 57. Cusick, J. T.— Jour, of Dairy Science, 3, pp. 195-205, 1920. 58. Hamilton, T. — Queensland Agr. Jour., Jan., 1921, p. 17. 59. Hardy, G— Jour, of the Dep’t. of Agr., South Africa, Vol. 4, p. 263, March, 1922. 60. McKay, G. L., and Larsen, C. — “Principles and Practices of Butter- making,” 3rd edition, (John Wiley and Sons, N. Y.), p. 336. 61. Fryhofter, C. W.— U. S. D. A., Circular 236, 1922. 62. Fettig — Centralblatt fur Bakteriologie, II, Band 22, No. 32. 63. Larsen, C., Lund, T. H., and Miller, L. F. — S. D. State College of Agriculture, Agr. Exp. Sta. Bui. 122. 64. Klein, J. — “Milchwirtschaft,” (Paul Parey, Berlin) ,p. 218. 65. Hunziker, O. F., Mills, H. C., and Spitzer, G. — Indiana Agr, Exp. Sta. Bui. 160. 66. Kildee, H. H. — Minnesota Agr. Exp. Sta., 25th Annual Report, p. 48. 67. Henzold, O. — Milch Zeitung, Vol. 31, No. 52, pp. 822-23, Leipzig, Dec. 27, 1902. 68. Marcas, L., and Huyge, C. — L’ Industrie Laitiere, Vol. 30, No. 16, pp. 187-188, Paris, Apr. 16, 1905. 69. Marcas, L., and Huyge, C. — Exp. Sta. Record, XVI, p. 1017, 1905. 70. Hoft — Milchwirtschaftliches Centralblatt, Vol. 5, No. 6, pp. 250-252, Leipzig, June, 1909. 71. Kooper, W. D. — Milch Zeitung, Vol. 40, No. 29, pp. 285-287, Leipzig, July 22, 1911. 72. Hunziker, O. F., and Hosman, D. F. — Journal of Dairy Science, Vol. 1, No. 4, pp. 320-346, 1917. 73. Palmer, L. S., and Combs, W. B. — Journal of Dairy Science, Vol. 2, No. 6, pp. 444-452, 1918. 74. Ruehle, G. L. A. — Michigan Agr. Exp. Sta. Quart. Bui. 3, No. 3, p. 103, 1919. 75. Editorial — Molkerei Zeitung (Hildesheim), Vol. 25, No. 58, pp. 1095-96, July 28, 1911. 76. Washburn, R. M. — N. Y. Produce Review and American Creamery, Vol. 48, p. 1004, 1919. 77. Marker — (Cited from McKay and Larsen “Principles and Practice of Buttermaking,” p. 202.) 78. Wrampelmeyer, E. — Landw. Versuchs., 42, No. 6, pp. 437-38. 79. Beilstein, 2nd edition, Vol. 1, p. 394. (Exp. Sta. Record, Vol. 5, p. 342.) 80. Stoklasa, Jul. — Zeitschrift fur Physiol. Chem.- XXIII, 1897, p. 343. 81. Burow, R. — Zeitshrift fur Physiol. Chem., XXX, p. 495, 1900. 82. Schmidt-Muhlheim — Arch. f. d. ges. Physiol. 30. 83. Bordas, F., and Raczkowski, Sig. de — Jour, de Pharmacie et de Chemie, Vol. 16, No. 6, p. 292, 1902. 84. Bordas, F., and Raczkowski, Sig. de — Comptes Rendu de l’Academie des Science, (Paris) 135, 302-303, 1902. 50 • Wisconsin Research Bulletin 57 85. Bordas, F., and Raczkowski, Sig. de — Comptes Rendu de l’Academie des Science, 134, 1592, 1902. 86. Bordas, F., and Raczkowski, Sig de — Comptes Rendu de l’Academie des Science, 136, pp. 56-57, 1903. 87. Bordas, F., and Raczkowski, Sig de — Archiv fur Kinderh., 40, p. 18. 88. Koch, W., and Woods, H. S. — Jour, of Bio. Chem. 1, p. 211, 1905. 89. Nerking, J., and Hsensel, E.— Biochem. Zeitschrift, 13, 348-353, 1908. 90. Lewkowitsch, J. — “Chemical Technology and Analysis of Oils, Fats and Waxes,” 5th edition, (MacMillan & Co., N. Y.) Vol. 2, 776-861, 1914. 91. Gilkin, W. — Biochem Zeitschrift, 21, 348, 1909. 92. Brodrick-Pittard, N. A. — Biochem Zeitschrift, 67, 382, 1914. 93. Tolmatscheff — Med. Chem. Untersuchungen von Hoppe-Seyler, Heft 2, p. 272, 1867. 94. Winterstein, E. — Zeits. f. Physiol Chem. 41, 486-504. 95. Koch, Waldemar — Zeits, f. Physical Chem., 37, 327-330, 1906. 96. Fetzer, L. W. — Abstracted in Science n. s., 33, 339, March 3, 1911. 97. Osborn, T. B., and Wakeman, A. J. — Jour, of Bio. Chem., 21, 539, 1915. 98. Arbenz, E. — Chem. Abs., 14, p. 81, 1920. 99. Dornic and Daire — Ecole lait surgeres, Ann. fals., 3, 533-38. (Chem. Abs., 5, p. 727, 1911.) 100. Folin, Otto, and Farmer, C. J. — Jour, of Bio. Chem., 11, 493, 1912. 101. Folin, Otto, and Macallum, A. B. — Jour, of Bio. Chem., 11, 523, 1912. 102. Delepine — Bui. Soc. Chem. (Paris), Series 3, Vol. 13, p. 163, 1895. 103. Cambier et Borchet — Bui. Soc. Chim. (Paris), series 3, Vol. 13, p. 396, 1895. 104. Ronchese — Jour. Pharm. Chim., 6, 25, 611-17; Bui. Soc. Chim., (Paris) Series 4, Vol. 1, p. 900, 1907. 105. Malfatti — Zeitschrift f. Anal. Chem., 47, 273, 1908. 106. Wilkie, J. M. — Jour. Soc. Chem. Ind., Vol. 29, p. 6, Chemical Abs. 4, 1005. 107. Parker— Gas World, 62, No. 16, 6, p. 11. 108. Thau — Das Gas und Wasserfach, 64, 770-2, 1921 ; Gas World, May 1, 1915, p. 11. 109. Meurice, R. — Chemical Trade Journal, 70, 103, 1922. 110. Sanders, A. — Das Gas und Wasserfach, 64, 770-2, 1921. 111. Budai, Koloman — Zeitschrift f. Physiol. Chem., 86, 106-121, 1913. 112. Long — Jour, of the Amer. Chem. Soc., 30, 881-895, 1908. 113. Mahlengreau, F., and Prigent, G. — Zeits. f. Physiol. Chem., *77, 107, 1912. 114. Tosaka Kinoshita — Zentralblatt Physiol., 24, 776-9,1911. 115. Brieger, L. — “Weitere Untersuchungen uber Ptomaine,” Berlin, p. 54, 1885. 116. Gulerwitsch, Wl. — Zeits. f. Physiol Chem., 24, p. 513, 1898. 117. Moruzzi, G.— Zeits. f. Physiol. Chem., 55, 352-359, 1908. 118. Morner, C. Th.— Zeits. f. Physiol Chem., 22, 514, 1892. The Fishy Flavor in Butter 51 119. Schulze, E. — Zeits. f. Physiol. Chem., 60, 155-79, 1909. 120. Heffter, A. — Chem Centralblatt, 1, 459, 1891. 121. Erlandsen, A. — Zeits. f. Physiol. Chem., 51, 71, 1907. 122. MacLean, H. — Biochem. Journal, 9, 351, 1915. 123. Coriat, L. H. — Am. Jour. Physiol., 12, 353, 1904. 124. Riedel— Berichte, p. 26, 1907. 125. Diakonow — Centralblatt f. d. Medic. Wissenschaften, 1868, p. 438, (Cited by Gilson in Zeits. f. Physiol Chem., 12, 585, 1888. 126. HefTtner — Archiv f. exp. Pathol, und Pharm. Bd. 28, S. 97; Zeits. f. Physiol Chem., 55, p. 353. 127. Gilson, E. — Zeits. f. Physiol Chem., 12, 585, 1888. 128. Hammersten, Olof, and Hedin, S. G. — “Lehrbuch der Physiologischen Chemie,” 8th edition (J. F. Bergmann, Wiesbaden) p. 227. 129. Wurtz, Ad. — Compt. Rend. Vol. 66, p. 772. 130. Neuberg, C., and Blumenthal, F. — Beutrage zur Chemischen Physi- ologic und Pathologie, Vol. 2, No. 5-6, p. 238, Braunschweiz, June, 1902. 131. Orgler, Arnold — “Beitrage zur Chemischen Physiologie und Path- ologic,” Vol. 1, No. 10-12, pp. 583, Braunschweiz, May, 1902. 132. Pitz, W. — Wisconsin Ag. Exp. Sta. Bui. 302, pp. 49-50, August, 1919. 133. Thunberg, T. — Centralblatt f. Physiologie, 23, 625, 1909. 134. Warburg, O., and Meyerhofer, O. — Zeits. f. Physiol. Chem., 85, 412. 135. MacLean, Hugh — “Lecithin and Allied Substances, The Lipins,” p. 68, 1918. 136. Mathews, A. P. — “Physiological Chemistry,” 3rd edition, (Wm. Woods & Co., N. Y.) p. 98, 1921. $ 30 -/ bV/5 Jui, Research Bulletin 58 December, 1923 i Service Relations of Town and Country J. H. KOLB AGRICULTURAL EXPERIMENT STATION OF THE UNIVERSITY OF WISCONSIN AND UNITED STATES DEPARTMENT OF AGRICULTURE CO-OPERATING MADISON 1923 THE SIX-FOLD SERVICE RELATION OF TOWN AND COUNTRY The Economic service includes merchandising, marketing, and financing. The merchant by selling his wares to the farmer contributes directly to the latter’s standard of living. Marketing is the reverse side of the relationship; it is the local assembling of the farmers’ product for world consump- tion. Financing works both ways, the farmer buying and selling in town and the merchant buying and selling from the farmer. Both processes require financing. The Educational service holds the key to the future with reference to social attitudes affecting town and country re- lations. This service is represented locally by schools, libraries and lyceums of various sorts for lectures, music, drama and art. The Religious service has to do with those idealizing and motivating forces in rural life. Its local representative is the church and its various organizations of benevolence and re- ligious education. It has at its command music, drama, beautiful imagery, commanding architecture, reverent rituals and inspiring personalities. The Social service is concerned with those functions which make for sociability, contentment and happiness and quite as much, for those of social welfare represented by pro- grams of health, charity or corrective work. The Communication service has within its grasp powerful means for unifying or dividing community life. The channels for understanding and communion are here. They are such as the telephone, telegraph and the radio; the mail service; the newspaper and magazines ; transportation in all its forms ; for the roads themselves become means of communication. The Organization service. No service relations can be effective or permanent until they are appropriated, incorporat- ed and incarnated within the life of local groups of people. This organization service, then, whether of cooperative en- deavor, of community organization which is nothing more than the local harmonizing and harnessing of all interests or that of leadership from within, become matters of prime im- portance demanding more than a haphazard or casual at- tention on the part of the town and country community. MUTUAL IMPLICATIONS For the Town. For the townsman it means that his town must become the specialized service station for the larger community. Each town may well specialize in those services which it can render, most efficiently. This means, of course that no one town and its community can live unto itself but must work out interrelations with other towns and their communities. For the Country. The implication for the farmer involves the assumption of a keener responsibility for the larger com- munity life rather than attempting to hold to a family or a neighborhood economy. The farmer is in real need of this larger organization relationship where his own interests may be safe-guarded and at the same time united with those of his town or city in order to effect an efficiency commensurate with the greatly expanded needs of his day. Service Relations of Town and Country THE SERVICE ORGANIZATION OF TOWN AND COUNTRY A BRIEF OF FINDINGS AND PRINCIPLES A FEW HEADLINES selected at random from the day’s local press, such as “Farmers and Townsmen Join Forces,” “Farm and City Groups Combine,” “Working It Out Together”, and “Big Town and Country Meeting Tonight”, indicate an awakening public consciousness regarding the mutual dependence of groups in society. The experiences of the past years, the uncertainty of the present, and the rapid economic and social changes overtaking old institutions are slowly but surely ^focusing general attention upon the social arrangements in rural life. The present study analyzes a very small segment or part of the whole in an attempt to make a little clearer this great mesh of relationship. The farmer and his family are continually dependent for all manner of services upon agencies lying beyond the gate of their farmstead. They join with neighbors for some of these services in neighborhood groups and just beyond these they find themselves in contact with the small town or village. Changing Relations of Town and Country. This study becomes an examination of the tendencies and forces which are reconstruct- ing the relations between the town and the country. Some of the factors for change which have been creeping in are shiftings and shrinkings in rural population, greatly increased facilities for com- munication and transportation and the readjustments in local social and economic institutions and organizations made necessarjr by the greatly expended needs and the increasing complexity of rural life. An indication of the direction which these tendencies are taking in this changing relationship may be seen from a study of open country neighborhood or primary groups made in the same area . 1 Two general conclusions were drawn from this study. First, that the neighborhood groups are becoming fewer in number although they have a tendency to become somewhat larger in size and second, that those neighbor - ^Kolb, J. H. Rural Primary Groups, Res. Bui. 51, Agr. Exp. Sta. Uni. Wis. Madison, 1921 2 Wisconsin Research Bulletin 58 hood groups in close proximity to a town or village tend to give way first. Considered from the basis of services rendered such as economic, educational, social and the like, of the 121 neighborhood groups found in the county, 60 were performing for their people one or more services out of a possible six; 40 were rendering two services but only five groups were discovered giving as many as four services. For the balance of the services these farming people were dependent up- on nearby towns. Those who had worked out no local arrangements of their own, were completely dependent upon the town. This' points to a linkage of town and country in respect to many of the larger community relations. But elements of change are also in evidence in the town and village itself. Some of the underlying forces are the same as those influencing the neighborhood or primary groups although the local manifestations may be somewhat different. For example, the farmer is no longer bound to any single town. His area of service relations has been greatR expanded. This becomes a challenge to the former channels of relationships which have been more or less taken for granted. Here are two groups therefore, the town and the country struggling to make their adjustments to modern demands and in the process they are finding themselves held together by common tasks and by problems of an ever changing character. Many of these problems to be sure, are common to any such areas of reorganization or of dis- organization in whatever section of society they may occur. The Areas Studied. All of Dane County was considered for certain parts of the study. Four towns in the eastern part were then selected for intensive work. The towns themselves were first studied and then farm families living in their service areas were visited. Dane County is fairly typical of the dairy section of the southern part of the state and the region studied is also one of the two largest tobacco producing areas in the state. Elkhorn, in Walworth County and Waupaca, in Waupaca County were studied in the same manner to serve as checks upon tendencies discovered in the first county. Waupaca boasts of being the “Potato Capital” of the state; and Elk- horn is proud of its established dairy business lying within the Chicago whole-milk area as well as of its stable and prosperous population. The Elkhorn area represents one in which the old limitation of rural isolation is reduced to a minimum. Cement highways radiate in every direction, telephone wires line every roadside and practically every farm has its car or truck or both. The project was formulated in such a way as to examine in close detail a limited area but in a comparatively wide range of relationships. The purpose was to bring into relief this particular phase of social or- ganization and to clear the way for other detailed studies. Service Relations of Town and Country 3 TOWN AND COUNTY INTERDEPENDENCIES “Servicetown is your city, Mr. Farmer. Without you it would not be here today. It is for you and through you that it lives,” dramatically declares a country editor in trying to arouse the various elements in his community to the importance of their common weal. He then suggests that it has taken years td develop some of the attitudes and antagon- isms between town and country, and that probably it will take years of hard work to repair the seams. In the background for such philosophiz- ing, of course, lie the historic facts surrounding the origin of the American small town. A recent visitor from Denmark spoke casually of the great inconvenience of gathering up the milk each day from the many widely separated farms and said that the common village milk depot used in his country was much handier. This is typical of the many differences arising from the fact that many American country towns, in contrast to those of the Old World, came after the settle- ment of the separate farms, and did not become the residence centers for the farm families. These centers are not so likely to possess the primary group characteristics so evident in the country neighborhood settlements of an early day. The enterprisers, speculators, merchants of the town “make their living off the the farmer”, the farmers used to say. The day of interpendencies, however, has arrived. The Service Relationship Made Definite. What are these services ? All classifications tend to become arbitrary but for the sake of being definite, it is proposed that no rural community is living up to its op- portunity unless attention is being given to at least six fundamental services : The economic, including merchandising, marketing and financing; the educational; the religious; the social; communication and transportation ; and that of organization activity itself. In the county of Dane there are 29 towns and villages exclusive of the city of Madison, and 23 of these perform at least five of the listed services for their surrounding rural territory. By services is not meant any idea of charity or benevolent gifts, it is rather a successive social re- lationship of useful offices where advantages are conferred and paid for if necessary. It is that which promotes the interest, happiness and welfare of a group concerned. Services Affect Directly Standards of Rural Living. Since the farmer and his neighbors are thus so closely bound to the town, the type of services which are available to him and the character of those agencies which make these services their specialty are matters of first importance. In no small way this town becomes the farmers’ gateway into the world. The gate swings both ways, however, toward the farmer as goods and services are passed to him by the town from the world, and away from him as he passes his goods and services on 4 Wisconsin Research Bulletin 58 through the town. If the rural community is to progressively raise its standards, goods and services must be increased to meet these needs. This means a distribution by local centers and agencies capable and determined to augment their qualities and quantities. This has specific reference not only to those generally recognized elements in a physical standard of life, but also to those making up the sociability, the aesthetic, the religious or the educational. The correlation between such standards and the store window display, the advertising circular or the spirit in the school room have become very evident during the course of study. In such a program there can be little opportunity for an epidemic of booster clubs or noisy campaigns for “shoe string” factories for the town but rather a scientific, progressive development of the agricultural landed area lying around about each town. The one principle in the situation clearly revealed during the course of in- vestigation is that of the necessity for mutual understanding and action. THE TOWN MEETS AN OPPORTUNITY In any scheme of relationships where interdependencies are involved there are at least two parties to the contract. In this case it is the town and its farmers and the farmer and his town. The findings re- garding the first of these will next be considered. Some Characteristics of the Small Town. As an aggregation of service agencies of various sorts, the town presents interesting characteristics. There are close correlations between the size of the town and the kind and number of agencies which tend to collect there. In the very small type of about one hundred population, there is found a couple of general stores, a garage and some sort of a produce or shipping concern for the marketing of the farm or dairy products. As the three hundred population class is passed the number and kinds of agencies increase rapidly. Among the commercial concerns the tendency after the one thousand class is reached, is to have a decrease in number of the general types and a sharp increase in the specialized store such as a grocery, shoe or hardware establishment. When all types of service agencies are plotted in relation to town populations, the merchandising, personal and professional services show a definite tendency to increase greatly with the size of the town. The other agencies as communication, financial or educational maintain an even tenor in the average per town comparison. The ratio of town to general trade area population presents an interesting characteristic. Some of the small centers rendering less than five services have difficulty in maintaining a 30 to 70 per cent ratio in favor of the trade area. The next group of towns near or a little above the five hundred class and performing most of the services but Service Relations of Town and Country 5 in no specialized form, has an average ratio of about 30 to 70 per cent ip favor of the trade area. From this point the larger per- centages for the country side of the equation are slowly reduced as the size of the town increases. When the town and country population are compared on the basis of the number of people required for service agencies, it is found that there are less people per church, for example in and about the small towns than for those of larger size. Expressed in the other way, there are more churches per hundred population in and about the small than the larger towns. This does not mean more institutions per town necessarily, but small volume per agencies and in many cases a con- sequently less efficient rendering of the service. This general tendency holds for most of the types of agencies with the decided exception in the cases of the personal and professional service and that of the social and organizational activities. In these cases it appears that it is actually inadequacy in the service due to lack of a sufficient number of agencies. The volume of business measured in dollars for the smaller towns in the vicinity of five hundred population but below one thousand, when divided between country and town was in a ratio of about 72 to 28 per cent in favor of the country. When the thousand mark is passed the percentage tends to drop to about 65 per cent from the coun- try. This relationship, especially for the merchandising services, is fully developed in part two of the study. Types of Service Centers Classified by Services Rendered. Any at- tempt at a classification by types must be considered temporary pend- ing further study of more towns and in more widely distributed areas but for the Dane County towns the following types are in evidence. 1. The Single Service Type. This is usually an open country or cross roads stand where there is a single service performed as by a general store, a church or school. It also may be represented in the very small hamlet where the population falls below one hundred. Centers of this type were not considered in this study since in the strict sense of the word they are not towns at all but usually stand at the service center of a neighborhood or primary group. These centers were dealt with in the previous investigation referred to as the primary group study. 2. The Limited, Simple Service Type. In the county were twelve centers of this type, Cottage Grove being the one fully examined. The average population for these twelve small villages was 250, the range being from about 100 to about 400. These centers fall short in the “Six Service” standard. Their general trade areas are relatively small and the specialized service areas are usually completely lacking. Open country neighborhoods do not appear within the trade zones. 6 , Wisconsin Research Bulletin 58 3. The Semi-Complete or Intermediate Type. Fifteen of this type were found in the county or on its immediate border. Their average population was 800 with a range of from about 400 to just a little over 1,200. This type is found most frequently in the area studied and stands in a sort of intermediate relation to the larger centers which may be county seats or have certain characteristics often associated with county seats towns. Cambridge, Deerfield, Mazomanie and Black Earth, the towns studied most closely, come easily in this classification. Most of the towns in this class render the six services but are frequent- ly lacking in some essential as for example, Cambridge is without its railroad, and Deerfield pays small attention to social and organization activity. The general trade areas of this type of center are relatively large and the amount of business from the farm source is nearly 75 per cent of the total. The open country neighborhoods are not found close by and the specialized ares are small or entirely absent. 4. The Complete and Partially Specialized Type. In the county or its immediate vicinity and extending their service influence into the county, were seven towns of this type. Their average population was 2,750, ranging from about 1,200 to just a little over 5,000. The type in the county is clearly illustrated by Mt. Horeb in the western part and Stoughton in the eastern part. Elkhorn and Waupaca, both county seats, are also good examples. The Stoughton area presented a situa- tion entirely different from Cottage Grove, Cambridge and Deerfield. It is sufficiently large to render all the services represented in the questionnaire. The farmers either trade at Stoughton or they don’t. Their choices are not so much in the matter of towns but in the matter of stores in the town. Since the town is larger the farmers take more for granted, they feel no particular sense of responsibilhy for its succcess, their attitude becomes less personal and they depend more largely upon their own small clubs and organizations for their more intimate and social relationships. Therefore, open country neighbor- hood groups are to be found in the vicinity. The “general trade” areas are relatively smaller than in third type, but the specialized areas are much larger. 5. The Urban and Highly Specialized Type. This type, of course, is represented by the urban city center. Its relation to the rural community and to agriculture is indirect. Its interests be- come more divided with manufacturing and wholesaling interests. Madison, Milwaukee and Janesville represent this type for the area under study. There are doubtless other types which studies in other sections could reveal between the city just above the 5,000 classification and Madison a center of nearly 40,000 population. This type is characterized by highly specialized service agencies. It is to these centers that farmers and their wives come for purchases where quality, variety and chance for a discriminating selection are the de- Service Relations oj Town and Country 7 termining factors. Their “general trade” areas are relatively small and the specialized areas are very large. Country neighborhood! centers tend to cluster about them a comparatively few miles beyond the city limits. These centers are incomplete in the sense that they cannot cater to the general trade needs with emphasis upon quick conveniences to the buyer, as can the general store in the small town where the farmer can leave his list of goods while he goes away to have the horses shod and returns to find all his requirements met by this same store and the same salesman. Or again, a farmer does not expect to find spools of barbed wire fencing decorating the show windows of the hardware store “on the square” in Madison. The Interrelation of the Types. A schematic representation of the distribution of these various types is shown in a conventionalized curve Chart I. — A Theoretical Graph Indicating the Distance Relation of Centers of the Five Types. in Chart I. If this is the correct theoretical picture, a farmer in an area with a comparable density of population and transportation facility may expect to find an open country stand or small hamlet within a distance of about two and a half to three miles from his farm. He may find a center of the second type about four miles away and if he travels to the fourth type about fifteen or sixteen miles distance, he must pass through another of the second type. As he pushes on toward the city center, the fifth type, between thirty-five and forty miles away, he must again pass through two other towns of the 8 Wisconsin Research Bulletin 58 second type and two of the third. There are seven of this city type of center within a radius approximating forty miles from the city of Madison. 2 In a private study made in Kansas the findings indicated that towns of the fourth type, usually county seat towns, were twenty- five to forty miles apart, that the third or intermediate type were twelve to twenty miles apart and the small villages of the second type were separated six to ten miles. This comparison would seem to be indicative of a correlation between the density of wealth and popu- lation and the size and location of the various types of service centers. In Dane County, Wisconsin, for example, the volume and density of these service relations are such as to make possible the accumulation of service agencies in centers of the third type about every eight miles and so on through the series. Chart II is drawn to show the in- terrelation of these types respecting their size and service areas. Chart II. — A Theoretical Graph Indicating the Interrelation of the Service Areas for the Five Types of Centers. The Trade Area Idea. Undoubtedly there is what may be termed a “trade area” round about each town where families tend to go for general trading and which they often designate as their “home town”. This study indicates that this “general trade” usually means such econom.c services as merchandising in groceries, work clothes or farm machinery together with banking and marketing. These areas are not coextensive to be sure, but as can be noted from the maps, Figures 4 to 10, they do follow along about the same lines. The moment “trade area” is made to mean a particular kind of trade or is -Hy courtesy of ilie advertising department cf Curtis Publishing Company. Service Relations of Town and Country 9 used in a technical sense, trouble is encountered, for as is clearly seen from the maps, Figures 6 and 8, trading in furniture and in good clothing as suits and women’s ready-to-wear, do not follow the “general trade” lines nor any other. As is pointed out in the text, the contrast when Cottage Grove and Waupaca are compared is most strik- ing. Nor do the social, religious or organization service areas tend to follow the “general trade” or any other of the economic service areas. There is more of a correspondence in the cases of the communication services and the high school areas but even here it is not convincing. This would seem to suggest that as this town launches out to ex- ploit its opportunities, it encounters natural, social cleavages and ar- rangements and that this mesh of resulting service relationship tends to align itself in zones following rather definite principles. This trade zone in its relationship to the other service zones will be outlined in the last section of this first part of the report. BUSINESS AND LIVING EMPHASIS FOR AGRICULTURE At the other end of this picture of service relationships stands the farm home. It is the initial production unit but equally important, it is the final consumption unit. Success, measured by even the “greatest net return” depends quite as much upon the household consumption manager as upon the farmstead production manager. After all, in the economic scheme of things, consumption becomes the reason and the end for production. This “wealth spending” emphasis for agriculture is of vital importance at the present time. Standards of Consumption in Rural Life. A close reciprocal rela- tion exists between standards of living and that of the farming process. Standards cannot be elevated indefinitely apart from an economically successful program of production and marketing yet on the other hand, big or assured profits are no final guarantee of good standards for health, thrift, justice, religion, sociability or even of permanent residence on the land itself. Therefore, there must be a community concern in those factors or conditions which go to influence such ideals and standards, for sQciety has a very large stake in these home units where life has its origins. “Why” Co to this Town to Spend Money? Since SO much is in- volved in this matter of spending as it relates to rural living the question of “why” the mrm family goes to this or that town to satisfy its needs and wants becomes important. About 787 families answered the question as best they could. Many times it appealed to them as a foolish question but many times also they really did not know. 10 Wisconsin Research Bulletin 58 The element of convenience or “nearest” as it was stated, figured very largely in that set of economic services which can be assembled under the title, “general trade”. This includes groceries and work clothes to a large extent and to a less extent farm machinery and lumber and still less furniture. Marketing and banking followed closely this general set of services. When the question of good clothes came up, an entirely different situation was involved. This became a matter of variety, selection, and quality without much respect to distance since this spending was not a matter for weekly attention but arose per- haps two or three times a year. Answers regarding the influence of the six trade towns, the competing centers and the mail order house was secured and are fully analyzed in the last part of the report. The high percentages of “not specified” in a number of the classi- fications are startling. After listening to hundreds of these answers one is much impressed with the seeming lack of attention and thought given this matter of spending. Family budgets are things unknown. The importance of this cannot be overlooked and its correction may re- quire as much “extension” educational effort as has been given to per- fecting the catechisms of production. The agencies in the town dis- pensing these services so closely related to his living standards cannot longer remain a matter of indifference to the farmer. He must assume his share of responsibility for this town of his and arouse himself to the fact that he has outgrown the old household and neighborhood economy. “Why” Go to This Town to Spend Time? When time is spent, money is usually spent also but for different reasons. As the “whys” were called for in that realm of life including education, religion, social life and recreation, an entirely new collection came back. Here tradition, personal friendships, blood ties and club, fraternal or denominational preference made their claims. In this set of factors, however, are to be found influences more direct and more powerful in determining ideals and standards of consumption and living than those associated only with the commercial considerations. Despite this, the farmer up to the present has not shared equally in the leadership and management of those agencies located in the town which control this collection of services However, many of these services find expression through local neighborhood institutions. These services are more personal in character and have been more cherished locally; in fact they become distinguishing characteristics of primary groups. This leads directly again to the natural social arrangements and organization dis- played by the mapping of the service areas. These factors and the zones of service relationships must now be related. The Zones of Service Relationships. Various attempts have been made to show the mesh of service relationships which gather people into Service Relations of Town and Country 11 groups and bind together town and country. There are principles and processes of selection at work all the time and these can, with sufficient study, be understood. In the first place, reference must again be made to the neighborhood group study published as Research Bulletin No. 51. Even casual observation of the maps shows a uniform absence of Lodi Prairie du Sac Sa.uA Ci/y Be/I evil/e FIG. 1.— COMPOSITE SERVICE AREAS OF TOWNS AND ACTIVE NEIGHBOR- HOOD GROUPS IN WESTERN DANE COUNTY 12 Wisconsin Research Bulletin 58 such open country neighborhood groups in the immediate vicinity of the small towns. This indicates that this rural population naturally looks to the town for practically all of its services ; or in one sense here is another group, small town and immediate country, which has primary group characteristics. This might be called the village neighborhood. Sometimes the farmers and villagers mingle freely in all their service relationships and sometimes the farmers simply use the village center because of its convenience of location for their club or personal group gatherings. To measure the extent of this group relationship one more map was constructed, Figure 1. This represents western Dane County as it is divided from the eastern part of the county for school administration purposes. There was a schedule in hand from each farm family and the vote or tabulation was made by sections, that is, if the majority of the families depended on the town in question for four out of the six services, the section was included in that town’s composite service area. The average radius for such areas for all towns was 4.3 miles. When more than this average dis- tance was traversed an increasing number of services passed over into the hands of the active open country neighborhood groups or if sufficiently far away, to agencies of competing towns and to the larger centers. It would appear, therefore, that there are likely to be what might be called concentric zones of organization influence or drainage service basins around each town center. In the zones nearest the center the services are entirely discharged or it might be said that the drainage is complete in the direction of the town. In the areas farther out secondary systems are set up and social water sheds ap- pear on an ever widening scale. A correlary of the foregoing observation is equally important, namely, that many times a zone or basin for one service may lie partially or entirely within the zone for another service. For example, the farm families in the immediate zone of Cottage Grove are also in the Deerfield furniture zone for the simple reason that there is no furniture store in Cottage Grove The same is true in the matter of a masonic lodge. These same families are again by preference in the zone of the city of Madison for women’s good winter coats This is to say that the farmer is not to be bound by any scheme of hierarchical group arrangements through which he must climb to get out into the vcorld. He is at once in contact both directly and indirectly with this wider relationship of world affairs. But what is the character of this group life within these various zones? If every on’e of these service centers, whether open country or town of whatever type, could operate perfectly freely and without ob- struction the picture would be a nucleus surrounded by spherical zones of constantly increasing size but of decreasing intensity until finally they would fade out like the ripples in a still lake when a stone is thrown. Service Relations of Town and Country 13 Fortunately for society, this artificial condition does not exist. The effect of many stones thrown together, some large some small, is con- stantly evident. In the intensive zones close to the center the bonds are of the primary group kind, personal and intimate. The economic are there, but often they are secondary in influence, or are just taken for granted. The farmer has the characteristics of small groupishness but then he is not different from others, for even the university people have to have their self-electing fraternities and their exclusive clubs. In another zone and overlapping the first, if the center is a town, is a secondary zone of “general trade”, banking and often of high school. The majority of the centers here studied were of the type designated as number 3, the semi-complete or intermediate, and Chart I shows them about four miles from those of the second and a smaller type. This zone tends to increase in size with the size of the town until the fourth type of center, the complete and partially specialized, is reached. Then its ratio begins to decrease when town and country comparisons are made on basis of population or of the volume of busi- ness. As these zones are followed out very soon the eddies of influence from the larger centers are recognized in addition to those influences exerted by the open country groups. As these outer waves of the sma.ller town type and those of the larger type meet and battle one another back and forth, one finds himself in a third zone where specialization is the element pulling this way or that. This zone is the furniture and good clothing area, for example, which Figure 18 showing the zones of Madison’s influence, clearly illustrates. It is the area for occasional and specialized personal or professional service as the larger center’s clinic, or the recreational and cultural service of a big motion picture exhibition or a musical concert. These relations are indicated in schematic fashion in Chart II described above. The Principle of Specialization. With the amount of detail given to the various service relationships, the implications should be apparent. First of all. no service center, large or small, can live unto itself. The mesh gathers all with its cords. That the town needs the farmer and the farmer needs the town is a popular expression frequently in evi- dence. But in this scheme of things the small center cannot hope to make a success of specializing in dress suits for example but it can rout the city when it comes to selling overalls. Principles of merchan- dising need to be worked out in relation to the social organization situation of the town. This will doubtless mean as time goes on and the changes now in progress take more definite effect that certain agencies or even whole centers will have to go out of business or greatly modify their present service functions. It means also that these principles of rendering services cannot be worked out successfully by one center, selfishly and quite apart from all other centers. If specialization means anything it means that. Some of the larger city concerns have only recently learned this lesson at a high cost. Others 14 Wisconsin Research Bulletin 58 long since have been operating on the principle which one concern in St. Paul, Minnesota, uses in its direct mail advertising: “What you can’t purchase in your home town buy in St. Paul and at the Golden Rule.” These same implications hold true for all the other agencies. What is more, the converse of this specialization principle is also commanding. One group or agency cannot successfully evade responsibility for services which by all indications of these social relationships belong to it. Many local communities have evidently been attempting this of late in the matter of organized recreation and the result in many cases has been disastrous. They have been turning their young people into the territory of other groups where social control is at its minimum instead of adjusting themselves to changed needs and rendering the service themselves where social control is at its best. Principles of Town and Country Organization. The task of com- munity organization becomes that of harmonizing and putting to work the various service groups in a program of local betterment. At least three principles need consideration in such a program of town and country relations. First, the medium for such organization is confidence. The turning point is always motive, this implies the great social and emotional basis which has been stressed. Second, the stimulation of a community consciousness can come through recognition of common problems but also through a recognition of the special or group interests, namely that of farmer and of townsmen separately. Cooperation is built upon equalities. The farmer may organize about his interests and the townsman about his, then come together as a union of equals ready to fight common battles. Third, confidence and fellow feeling are conditioned pretty largely on understanding. This principle is being recognized and urged by many kinds of rural organizations and societies. This result can come about only by a sane but constant education on the part of each group regarding the work, the services, the difficulties, and the importance of the other group. Service Relations of Town and Country 15 PART II THE TOWN HAS ITS FARMERS How Important Is t-he Farmer In the Life of the Town? T O THE SMALL town standing midway between city and farm, many characterizations have been given. Some of these would ally it with the one and some with the other. So much emphasis has been given to the antagonisms between town and country which' this situation has produced that it may be well to study and to stress the interdependencies. The town is at once recognized as a place where folks live, a residence cluster, but it is much more ; it is a service center. It is an accumulation of service agencies whose existence and, therefore, the existence of the town itself, depend upon selling these services to the people of the town but also to those of the farming community round about. Then the town has an identity and con- sciousness all its own, differentiated from its supply depot, the city on the one hand and its clientele, the outlying community on the other. It is the purpose of this part of the study to first examine the town as an aggregation of these service agencies, to discover the kind and number of agencies which tend to come together in a center of given size or character. These service agencies will then be studied in more detail, both those which are commercial in purpose and those which are not. Finally the question will be raised as to what really makes a town, a town and service center. All the time the point of view suggested by the question at the head of the section is to be kept in mind: of what importance is the farmer in the life of the town? With this purpose in view, all the towns in Dane county, were given some direct and personal study. Some of the towns were ex- amined in greater detail and finally three of them, Cottage Grove, Deer- field, and Cambridge were completely studied in every detail. Wau- paca in Waupaca County and Elkhorn in Walworth County were studied to the extent that they might be used as checks upon tendencies discovered in the other towns. THE TOWN, AN AGGREGATION OF SERVICE AGENCIES The town will be considered only as a service center with its agencies of various kinds. The residence feature will be kept in the background. The town and its people are for the moment intent upon this central purpose of selling and rendering services. Other activities must not cloud the picture. 16 Wisconsin Research Bulletin 58 All the Towns and Ten Agencies. All the towns in Dane county with Elkhorn and Waupaca added, were examined for ten kinds of commercial services. The question to be answered is what kinds and what numbers of these commercial service agencies tend to gather in Table 1. — Average Number of Ten Kinds of Commercial Concerns per Town in All Towns in Dane County, Elkhorn and Waupaca Towns by size in population Num- ber of towns in class Average Number of Business Concerns per Town Gen- eral store Gro- cery Hard- ware Fur- niture Lum- ber and coal Gar- age Farm and dairy prod- ucts Bank Com- mer- cial amuse- ment Manu- factur- ing 100 to 300 11 2.1 .3 .7 .3 .7 1.4 1.3 .5 .3 .1 301 to 500 7 2.8 .7 1.3 .7 1.1 1.7 2.7 1.1 1.4 1.6 501 to 1000.... 6 3.6 1.1 1.0 1.1 1.5 3.1 2.3 1.8 1.6 1.0 1001 to 2000 3 3.3 2.6 2.0 1.6 2.3 5.0 3.3 2.0 3.0 3.0 2001 to 6000 2 1.5 10.0 6.0 3.5 3.5 7.5 9.5 2.5 5.0 7.5 towns of varying sizes? Table I gives the answer. In towns of less than 300 population on the average, you may expect to find only two general stores, one garage and one concern handling farm and dairy products. In towns up to and including 1,000 people, at least one agency for each of the ten kinds of services may be expected. Chart III shows Table II. — All Agencies in Eleven Towns Classified by Type of Service and Size of Town Population of towns Towns in class Agen< ues by Types of Service Rendered All agen- cies Mer- chan- dising Trades and repair Commu- nication and trans- porta- tion Finan- cial Per- sonal and profes- sional Reli- gious (church) Edu- cation (school and library) Social and organ- ization All towns... Av. for all... 97.5 41.7 10.4 6.1 3.8 14.1 4.4 2.3 11.7 100 to 300.. Av. in class 31.0 12.0 5.0 3.5 1.5 2.0 2.5 1.0 2.0 McFarland 28 13 2 3 ' 2 2 2 1 2 Cottage Grove 34 11 8 4 1 2 3 1 2 301-500 Av. in class 66.5 22.5 10.5 5.0 3.5 7.0 3.5 2.5 8.5 Black Earth 49 14 11 5 2 4 3 2 6 Cambridge 84 31 10 5 5 10 4 3 11 501-1000... Av. in class 62.3 24.6 7.6 4.0 3.0 10.3 5.5 2.3 5.3 Middleton 56 23 8 3 2 9 4 2 5 Deerfield 61 7 8 4 4 9 2 2 4 Mazomanie 70 24 7 5 3 13 5 3 7 1001-2000.. Av. in class 127.5 54.5 11.0 8.5 3.5 19.0 6.0 2.5 19.5 Mt. Horeb 212 50 12 9 3 16 4 2 22 Elkhorn 134 59 10 8 4 22 8 3 16 2001-6000.. Av. in class 218.0 103.5 19.5 10.5 8.0 34.0 7.0 3.5 27.0 Waupaca 191 94 18 10 5 28 4 3 26 Stoughton 242 113 21 11 11 40 10 4 28 Service Relations of Town and Country 17 Chart III. — Relation of Size of Town to Number of Merchandising Concerns per Town. 18 Wisconsin Research Bulletin 58 Chart IV. — Relation of Size of Town to Number of Service Agencies per Town Compared by Types. Service Relations of Town and Country 19 the relationship in a graphic way. The number of general stores rise until the 1,000 population class is reached then they tend to drop rather sharply. Grocery stores in contrast show a slow rise until the 2,000 population class is reached then they shoot upwards. All the Agencies and Eleven Towns. All manner of service agencies are now classified into eight different types and related to eleven towns scaled according to size of populations. Under the merchandising type are included buying and selling concerns, mostly of a retailing or a marketing character. With trades and repairs are classified such as painters, contractors, blacksmiths, cobblers or general repair men. In tiie class of communication and transportation comes telephone, post office, newspaper, railroad or motor transportation service. Personal and professional service is made up of the lawyer, doctor, dentist, barber, veterinarian and the like. For the financial service is gathered those of banking, insurance or investments and loans. The religious service brings together the information regarding the church as such with its various organizations. Religious educational services were studied separately but are not included in the figures unless it is so specified. Within the educational services, high schools, grade schools and libraries are included. The social and organization classification Table III. — Comparison of Town Population and Trade Area Population for Towns Whose Trade Areas Lie Completely in Dane County with Elkhorn and Waupaca Included Population of Towns and of Trade Areas Compared in Per Cent and in Number and Size of Trade Area Names of towns IN PER CENT IN NUMBER Size of a trade aree in squar miles Total Town Trade area Total Town Trade area All towns 100 63.5 36.5 90,717 57,657 33,060 1,138 Paoli 100 57.1 42.9 175 100 75 8 Mt. Vernon 100 29.9 70.1 408 122 286 17 Rockdale 100 30.3 69.7 459 139 320 8 Windsor 100 26.7 73.3 655 175 480 12 Cottage Grove 100 16.6 83.4 1,205 200 1,005 34 Cross Plains 100 22.8 77.2 1,320 300 1,020 41 McFarland 100 30.6 69.4 980 300 680 17 Dane 100 44.3 55.7 713 316 397 31 Verona 100 26.4 73.6 1,327 350 977 39 Black Earth 100 32.3 67.7 1,438 464 974 49 Cambridge 100 21.8 78.2 2,250 490 1,760 44 De Forest 100 32.5 67.5 1,519 493 1,026 27 Marshall 100 31.0 69.0 1,601 497 1,104 48 Waunakee. 100. 30.1 69.9 1,860 560 1,300 48 Deerfield 100 v 34.7 65.3 1,712 594 1,118 43 Middleton 100 35.2 64.8 2,249 791 1,458 54 Oregon 100 38.8 61.2 2,246 871 1,375 51 Sun Prairie 100 38.8 61.2 3,186 1,236 1,950 65 Mt. Horeb 100 38.0 62.0 3,556 1,350 2,206 80 Elkhorn 100 49.5 50.5 4,019 1,991 2,028 78 Waupaca 100 41.1 58.9 6,903 2,839 4,064 159 Stoughton 100 62.6 37.4 8,163 5,101 3,062 91 Madison 100 89.7 10.3 42,773 38,378 4,395 94 20 Wisconsin Research Bulletin 58 unites various formal or informal sociab.lity clubs, fraternal or wel- fare bodies and other organizational activity. How all these different kinds of agencies tend 10 codec., in towns of varying sizes ^s shown in Table II and graphically illustrated in Chart IV. The merchandising with the personal and professional services have the greatest proportional increase as the size of the town increases. The class above 300 population must be reached before a high school may be expected and the curves for churches and financial agencies as well as those for communication and transportation show a rather slow rise. Ratios of Town and Country Population. With Table III comes a comparison of the town population with that of its trade area .for all those towns whose trade areas lie entirely within Dane County and also for Elkhorn and Waupaca. The towns are arranged in order of size and with few exceptions there is a striking uniformity of per- centages of town and trade area population when they are grouped. In the class of 300 and below there is an unsteadiness although 30 per cent town and 70 per cent trade area represents about the average. In the two classes from 301 to 1,000 the ratio with two exceptions climbs from the 30 to 70 slowly toward the 40 and 60. Of the exceptions, the village of Dane is offset by Cambridge, the latter being especially successful in pushing its trade well out. Stoughton represents the turn- ing point when the ratio is in favor of the city and Madison shows the extreme. It is the inclusion of Madison which upsets the “All Towns” ratio. The Service Agencies by Town and Country Populations. How many people are needed to keep a service agency going is a question often asked. Table IV was drafted to give an answer for the ten commercial concerns when the towns whose trade areas lie in Dane County, exclusive of Madison, were classified by population groups and the population per concern contrasted as between town and trade area. The general store and the grocery store present the sharp-est con- trast when the same story is told as in Table I, for in towns of 100 to 300 population there is one general store for every 78 town people and one grocery store for every 1,336 town people which means, of course, that every town in this class does not have a grocery store. Most of the other concerns show a steady increase in population of town and trade area together per concern as the larger centers are reached. There is a tendency for a slight reduction for the hardware, produce and manufacturing concerns in the 2,001 to 6,000 class. On the whole this apparently means that for every hundred people in and about the towns of smaller population there are more of the general commercial agencies than there are in towns of larger population. Table V gives the result when all the agencies, classified by types as in Table II for the eleven towns, are examined in the same fashion. The table shows that there are more agencies per total population Table LV. Population for Each of Five Kinds of Commercial Concerns by Towns and Trade Areas for Dane County, Elkhorn and Waupaca Service Relations of Town and Country 21 Both O CO CM CO CO OCOlOOJH M WHNO t-H.t— 1 HHCO 1 Bank Trade areas 1C H H CO (M cO 1 -H O O lO O O CO (N Town o oo cq oo o ^(NKNCqoO CO t"- O CO 00 CO co ^ l>» tO 1 Both 472.9 804.3 376.1 717.4 1004.4 Garage Trade areas to o o co o < t'- tO CM »0 iOCON^hN CO to CO ^ ^ Town 121.4 237.2 201.1 305.1 529.3 K H O Both 1737.3 2212.0 1613.4 2142.2 2152.2 o O H Oh fc Furniture Trade areas CO to CM 00 O OOOOCOOO O 0 to to CO CM to O CM O O E* Ph O Ph Town 448.7 652.5 563.2 915.4 1134.2 Both 5202.0 1769.6 2016.5 1345.1 753.3 Groceries Trade areas 3866.0 1247.6 1312.5 773.0 356.3 Town 1336.0 522.0 704.0 572.1 397.0 m W Both 306 491.5 448.1 1076.1 5022.0 o m K & Trade areas 227.5 346.5 291.7 618.4 2375.3 H O Town 78.5 145.0 156.4 457.7 2646.6 Number of towns in class co Tf< co cm Population of towns 100-300 301-500 501-1000 1001-2000 2001-6000 22 Wisconsin Research Bulletin 58 of the trades and repair, communication and transportation, finance, religion and educational type for the towns of smaller size. Personal and professional agencies as well as social and organization activity shows the opposite tendency. The significance of this can readily be sensed in the problem of securing adequate medicial care for the small communities. The sample here in the number of towns is rather small for some classes and should be checked with other towns of other sizes. THE COMMERCIAL AGENCIES How important the farmer is to the business life of the town was found by spending a considerable amount of time in eight towns inter- viewing the business men by use of a detailed schedule regarding their business and its distribution as between farmer and other. Cambridge, Cottage Grove and Deerfield were studied completely. A representative of every agency was visited and the needed information obtained. Repre- sentative samples were secured from the other five towns, London, McFarland, Mazomanie, Middleton, and Waupaca. The figures secured represent business or other activity for the calendar year 1920. The study was made in the fall and winter of 1921 and since a full year of business was needed, the 1920 figures had to be used. It will be re- membered that this was the year when the break in prices was very sharp and so the absolute amounts may not have as much significance as the percentages and proportions, but it is the latter which are the most important in answering the leading question of town and country relations. The Total Business with Farmer and Other Compared. The business for the 227 agencies studied shows according to Table VI Table VI. — The Total Business and the Farm Business for Eight Towns Compared by Types of Service for the Year 1920* Number of agen- cies Total and Farm Business i Compared in Per Cent and in Number Types of service i] V PER CEN T IN NUMBER Total Farmer Other Total Farmer Other Total 227 100 72.2 27.8 $6,497,489 $4,694,560 $1,802,929 Merchandising 129 100 75.6 24.4 5,494,887 4,155,856 1,339,031 Trades and repairs 39 100 77.2 22.8 381,584 294,468 87,116 Personal and professional Transportation and com- 35 100 55.0 45.0 140,712 77,239 63,473 munication 24 100 34.8 65.2 480,306 166,997 313,309 *This table includes the business of six tobacco companies and the business of the thirty-two agencies studied in the town of Waupaca. These will not be included in the following tables of this section. that for the year 1920, 72.2 per cent came from the farmer and 27.8 per cent from other sources, chief of which, of course, was the town itself. When analysized by types, the merchandising shows 75 per cent farmer and 25 per cent other. Cottage Grove, Cambridge and Deerfield when Service Relations of Town and Country 23 Chart V. — Volume of Business for all Kinds of Agencies Compared by Tyres ard D'v’ded between Farmers and Others. Total Business $6,495,639 Parmer 4,692,710 Other 1,802,929 Me rchandising 5,493,037 Farmer 4,154,006 ether 1,339,031 Transportation and communication 480,306 Farmer 166,997 Other 313,309 Trades and Repairs 381 , 584 Farmer 294,468 Othdr 87,116 Personal and Professional Service 140,712 Parmer Other 77,239 I 63,473 1,000,000 3,000,000 DOLLARS 5,000,000 separated have practically the same proportion while Waupaca has a lower percentage from the farmer source. Chart V presents the dis- tribution in graphic form. Whenever farmer business is called into question, the town business man frequently speaks of the problem of extending credit. Table VII Table VII— The Business Agencies of the Seven Dane County Towns Distributed by Types and Credit Service Agencies Giving Credit in Per Cent and Number Types of service IN PER CENT IN NUMBER Total Number giving credit Number not giving credit Total Number giving credit Number not giving credit Total 100 61.8 38.2 189 119 70 Merchandising 100 80.6 19.4 103 83 20 Trades and repairs 100 48.6 51.4 35 17 18 Personal and professional 100 41.4 58.6 29 12 17 Transportation and communication 100 31.8 68.2 22 • 7 15 presents the number and percentages of the various types of business agencies which extend credit to customers whether farmers or others. The merchandising agencies, as might be anticipated, are the ones giving this credit service most frequently. By bringing into comparison for the Dane County towns, the total business and the credit business as divided between farmer and other, Table VIII is formed. In the merchandising type the farmer furnishes 78 per cent of the business and does 80 per cent of the credit business. Only in the personal and professional service classification does the farmer appear to be far 24 Wisconsin Research Bulletin 58 over-reaching in the ratio between cash and credit business, but here his credit amounts to 73 per cent of all the credit given while his total business is only 59 per cent. The Family Customers, Farmers and Others Compared. The customers are compared by families since this is a unit more readily comparable than the individual. With Table IX, the family customer. Table VIII— Comparison of Total and Credit Business for Farmuxs and Oriais Disrxisurjjo by Types of Service for the Year 1920 Total Credit and Cash Business Divided Between Farmers and Others in Per Cent and in Amount Types of service by cash and credit ; IN PER CENT IN AMOUNT Total faim and others Farmer Other Total farm and others Farmer Other All types total 100 73.6 26.4 84,802,458 $3,537,011 $1,265,447 Cash 100 72. S 27.2 3,764,606 2,703,433 1,061,173 Credit 100 80.9 19.1 1,037,852 833,578 20 ,274 Merchandising 100 78.1 21.9 3,871,291 3,023,190 1 848,101 Cash 100 77.2 22.8 2,892,062 2,232,607 659,455 Credit 100 80.7 19.3 979,229 799,583 188,646 Trades and repairs 100 77.9 22.1 373,892 291,383 82,419 Cash 100 77.7 22.3 345,010 268,154 76,856 Credit 100 80.7 19.3 28,792 23,229 5,563 Personal and professional 100 59.4 40.6 104,059 61,441 42,618 Cash 100 54.6 45.4 79,737 43,569 36,177 Credit 100 73.5 26.5 24,322 17.881 6,441 Transpjrtation and commu- nication 100 35.5 64.5 453,306 160,997 292,309 Cash 100 35.6 64.1 447.797 159,112 288,685 Credit 100 34.2 65.8 5,509 1,885 3,624 compared whh Table VT the total business, it is evident that with the merchandising and the trades and repairs services, the higher pro- portion of business is furnished bv the smaller proportion of farm families as compared to other families. This is equivalent to saying that the farm family is a heavier buyer than is the town family. Cambridge, Cottage Grove, and Deerfield when singled out show much the same tendency. The number of family customers per agency divided between farmers and others is next thrown into comparison with the number of farm families living within the town’s general trade area and the number of families in the town itself. This comparison indicates that for all the types of service each agency reaches 60 per cent of the farm families in the area and 84 per cent of the town families while for the merchandising service separately, each agency touches 68 per cent of the farm families, and 92 per cent of the town families. This indicates clearly the effect of competitive agencies and concerns reaching the same customers for small portions of their business. The individual business concern determined to increase its number of customers must ei.her enlarge its county constituency or eliminate its competitors. Service Relations of Town and Country 25 Table IX. — Farmer and Other Customers by Families Distributed by Types of Service* Type of service Farmer and Other Family C in Nu Customers, in Per Cent and mber IN PER CENT IN NUMBER Total Farmer Other Total Farmer Other Total 100 64.4 35.6 35,210 22,682 12,528 Merchandising 100 64.9 35.1 21,682 14,113 7,569 Trade and repairs 100 72.9 27.1 1,980 1,444 536 Personal and professional 100 68.0 32.0 6,570 4,468 2,102 Transportation and communication 100 53.4 46.6 4,978 2,657 2,321 ♦Customers are frequently duplicated as between the different services as well as between agencies in the same sendee. The Merchandizing Service Analyzed by the Volume of Business. The 103 merchandising concerns in the towns in Dane County are separated and analyzed by the volume of their business in Table X, Table X. — The Merchandising Service Analyzed by the Volume of Business Volume of business Number of Farm and Other Busini ess Compared in Per Cent and in Amount ii ST PER CENT in amount concerns Total To farmer To ether Total To farmer To other Total 103 100 78.1 21.9 $3,871,291 $3,023,190 848,101 0-5,000 12 100 62.7 37.3 32,959 20,675 12,284 5,001-10,000 10 100 45.2 54.8 79,477 35,921 43,556 10,001-15,000 14 100 46.8 53.2 183,136 85,698 97,438 15,001-20,000 12 100 70.1 29.9 207,408 145,387 62,021 20,001-25,000 9 100 73.5 26.5 211,621 155,566 56,055 25,001-30,000 9 100 68.6 31.4 261,492 179,400 82,092 30,001-35,000 5 100 64.9 35.1 124,799 81,032 43,747 35,001-40,000 5 100 89.6 10.4 196,000 175,600 20,400 40,001-45,000 3 100 71.6 28.4 128,465 92,000 36,465 45,001-50,000 4 100 87.3 12.7 193,009 168,409 24,600 50,091-over 21 100 83.6 16.4 2,252,945 1,883,502 369,443 still keeping the comparison of farm and other sources in the fore- ground. There appears to be a correlation between the greater volume of total business and the higher percentage of farmer trade. Table XT brings this total business classified by volume in comparison with the credit business divided into the same classes and Table XII carries the analysis into the comparison of farm and other family customers as well as the customers per concern in each classification. This com- parison simply details the one pointed out earlier when Tables VI and IX were brought together, namely that a greater proportion of business is furnished by a smaller proportion of farm families. Although some of the classes are represented by a rather small number of concerns in Table XII, there does seem to be an inclination for a greater number of farm family customers per concern to appear as soon as the $25,000 volume class is passed. 26 Wisconsin Research Bulletin 58 Table XI. — Merchandising Business Classified by Volume of Business and Distributed Between Cash and Credit and Farmers and Others Volume by cash and credit Number of concerns Total , Cash and Credit in Per Ci Business for Farmers and Others ENT AND IN AMOUNT in PER CENT IN AMOUNT Total Farmer Other Total Farmer Other Total 103 100 78.1 21.9 $3,871,291 $3,023,190 $848,101 Cash 100 77.2 22.8 2,892,062 2,232,607 790,583 659,455 Credit 100 80.7 19.3 979,229 188,646 0-5,000 12 100 62.7 37.3 32,959 20,675 12,284 Cash 100 63.6 36.4 25,909 16,480 9,429 Credit 100 59.6 40.4 7,050 4,195 2,855 ,001- 0.000 10 100 45.2 54.8 79.477 r 5.P21 43.556 Cash 100 40.8 59.2 67,727 11,750 27,541 8,380 40,186 3,370 Credit 100 71.3 28.7 10,001-15,000 14 100 46.8 53.2 183,136 85,698 97,438 Cash 100 43.2 56.8 157,408 25,728 68,082 89,326 Credit 100 68.5 31.5 17,616 8,112 1 5„001— 20,000 12 100 70.1 29.9 207,408 145,387 62,021 Cash 100 72.7 27.3 176,272 128,197 17,190 48,075 Credit 100 55.2 44.8 31,136 13,946 20001-25,000 9 100 73.5 26.5 211,621 155,566 56,055 Cash 100 69.5 30.5 149,745 104,028 45,717 Credit 100 83.3 16.7 61,876 51,538 10,388 25,001-30,000 9 100 68.6 31.4 261,492 179,400 82,092 Cash 100 43.9 56.1 108,752 47,750 61,002 Credit 100 86.2 13.8 152,740 131,650 21,090 30,001-35,000 5 100 64.9 35.1 124,779 81,032 43,747 Cash 100 64.7 35.3 90,117 34,662 58,315 31,802 Credit 100 65.5 34.5 22,717 1 , 45 35,001-40,000 5 100 89.6 10.4 196,000 175,600 20,400 Cash 100 81.6 18.4 105,130 90,870 85,816 19,314 Credit 100 98.9 1.1 89,784 1,086 40,001-45,000 3 100 71.6 28.4 128,465 92,000 36,465 Cash. ... 100 70.9 29.1 86,815 61,500 23,315 Credit 100 73.3 26.7 41,650 30,500 11,150 45,001-50,000 4 100 87.3 12.7 193,009 168,409 24,600 Cash 100 86.5 13.5 163,231 141,131 22,100 Credit 100 91.6 8.4 29,778 27,278 2,500 50,000-over 21 100 83.6 16.4 2,252,945 1,883,502 369,443 Cash 100 84.8 15.2 1,760,956 1,493,767 267,189 Credit 100 79.2 20.8 491,989 389,735 102,254 Extending the Business by Advertising. To some considerable ex- tent at least the advertising policy of a business may be considered as a measure of its ambition to extend and expand its service. This is more true to be sure of the merchandising types than the others as Table XIII fully demonstrates. This merchandising group is therefore more fully detailed in Table XIV. With some exceptions, doubtless due to the small number of concerns in the classes, there is correspondence be- tween the large volume of business and the amount spent for advertis- ing both when considered per concern and per family customer. The Service Relations of Town and Country 27 Table XII.— Family Customers of Farmers and Others Compared by the Volume of Business for the Merchandising Agencies Volume of business Number of Average Number of Cui Per Cent ani 3TOMERS BY FAMILIES IN ) in Number concerns u V PER CEN T i N NUMBED Total Farmer Other Total Farmer Other Total 103 100 65.0 35.0 272 177 95 0-5,000 12 100 50.2 49.8 223 112 111 5,001-115,000 10 100 71.0 29.0 200 142 58 10,001-15,000 14 100 59.4 40.6 244 145 99 15,001-20,000 12 100 52.1 47.9 309 161 148 20,001-25,000 9 100 58.8 41.2 277 163 114 25,001-30,000 9 100 74.7 25.3 304 227 77 30,001-35,000 5 100 56.2 43.8 450 253 197 35,001-40,000 5 100 80.4 19.6 230 185 45 40,001-45,000 3 100 79.4 20.6 408 324 84 45,001-50,000 4 100 86.8 13.2 190 165 25 50,001-over 21 100 73.6 26.4 273 201 72 Table XIII.— Agencies and Expenditures for Advertising by Types of Sbivice fdi tie Yevr 1923 Advertising by Agencies and Number Amount per Agency Types of service of agencies agencies advertising Amount spent examined annually per In In agencies by per cent number those which advertise Total 189 39.1 73 §82.63 Merchandising: 103 61.1 63 90.35 Trades and repairs 35 17.2 6 30.00 Personal and professional 29 6.9 2 5.00 Transportation and communication 22 9.1 2 75.00 Table XIV. — The Advertising Practice of the Merchandising Concerns .Arranged by Volume of Business, Average Amount Spent per Concern and per Family Customer Volume of business Number of concerns Concerns which Advertise and the Amount Spent per Concern and per Family Customer concern ADVEB S WHICH ITISE Average amount spent per concern Average amount spent per family customer in cents In per cent In number Total 103 61.1 63 $90.35 30.5 0-5,000 12 41.6 5 30.00 12.5 5,001-10,000 10 40.0 4 52.50 24.7 10,001-15,000 14 57.1 8 46.00 19.2 15,001-20,000 12 50.0 6 102.00 32.3 20,001-25,000 9 89.0 8 81.87 29.5 25,001-30,000 9 89.0 8 127.75 38.4 30,001-35,000 4 100.0 4 57.00 12.6 35,001-40,000 5 80.0 4 58.75 34.0 40,001-45,000 3 66.6 2 265.00 50.4 45,001-50,000 4 50.0 2 50.00 10.0 50,001-over 21 | 57.1 12 131.91 41.0 28 Wisconsin Research Bulletin 58 three towns completely studied, Cottage Grove, Cambridge and Deer- field, have a lower rate of $79.28 per concern and 24.2 cents per family customer for the merchandising business. Waupaca on the other hand has decidedly higher rates and a much higher percentage of concerns which do advertising. Figured on the basis of gross sales, the Dane County merchandising concerns spent a little over one tenth of one per cent in their advertising. In contrast with this very low figure three’ concerns in Waupaca were spending three per cent of their gross sales. The Gross Operating Margins or Costs in Merchandising. By operating margin or cost is meant the difference between the money taken in for goods sold and the money paid out for the purchase of goods with due allowance being given to any difference of inventory at close of the year. The formula is total sales minus cost of goods plus or minus the difference of inventories. Table XV presents this operat- Table XV. — The Gross Operating Margins or Costs for the Merchandising Service and the Annual Turnover by Volume of Business Volume of business Number of concerns Gross Operating Margins Compared in Per Cent Annual turn- over (total business divided by average inventory) Gross margin received in per cent of cost price Gross margin received in per cent of selling price Gross margin expected in per cent of selling price 0-5,000 6 32.8 24.6 25.8 3.1 5,001-10,000 3 42.7 31.5 27.6 5.2 10,001-15,000 10 20.6 17.8 25.6 4.9 15,001-20,000 8 28.1 22.5 25.6 4.6 20,001-25,000 4 25.9 21.2 . 22.5 3.2 25,001-30,000 5* 12.5 11.2 18.0 7.0 30,001-35,000 3 17.7 , 15.2 18.3 7.8 35,001-40,000 3 16.4 ' 12.9 20.0 5.5 40,001-45,000 3 21.3 18.3 21.6 5.0 45,001-50,000 3* 1.8 1.6 10.0 3.3 50,001-70,000 5 20.7 18.4 18.6 4.8 70,001-95,000 6 21.0 17.5 25.3 5.5 *One concern in each of these classes, a feed store in each case, operated at an actual ing margin for 69 merchandising concerns expressed in per cent. The percentages are calculated in two different ways. The first is on the basis of cost price of the goods which is the way the consumer usually thinks of it for then it really means the number of cents, out of each dollar which he gives to the merchant which he, the merchant, does not give out again in purchasing the goods. The second way is to use the selling price as a base for arriving at the percentage and this is the methods which the retailer uses. In the table, per cent “expected” figured on the selling price is compared with the margin actually re- ceived computed after the retailers practice. This per cent expected is what might be termed the mediant’s principle of merchandising, his method of quoting prices or his “mark up” above the wholesale price, as the language of the trade expresses it. Out of this margin, of course, Service Relations of Town and Country 29 must be paid all operating expenses as salaries, rent or interest, maintenance and compensation for the merchant’s own services. It is quickly evident that considerable part of the consumers’ dollar goes to pay for this retailing service. Although the number of concerns in some of the classes are rather small the evidence seems to indicate that there is a relationship between high margins and low volume of business or stated in the reverse manner, efficient merchandising in terms of rendering the service at low costs is conditioned upon having a sufficient volume of business. In Table XV is also included a caption of “annual turn over” which means the number of times the money investment in goods on hand had to be used in order to bring in the amount of the total sales. The formula is total business divided by the average inventory figured at selling or retailing prices. No. distinctive tendencies are observeable here, due probably to the variety in kinds of concerns compared, although the suggestion with an exception or two is that low turn over is associated with low volume of sales. The Financing Services of the Bank. Although materials were gathered for other forms of the financing service than, banking, the figures were hardly comparable due to differences in the kinds of agencies and their methods of keeping records. The local deposits and loans for the banks in the towns of Cambridge, Cottage Grove and Deerfield are given in Table XVI for the first call day in January of 1920 and of 1921, where the farmer’s and town man’s business are com- pared. The 1921 figures include those of one bank which was not in business the previous January but exclusive of this business which might be termed new business, the deposits increased over $21,000 during the year. The farmer was slowly losing ground in the propor- tion which he was furnishing when the two dates representing the year’s business are compared. It appears that he was transferring his savings over into his checking account to meet the inequality in the deflation of which he realized himself the victim in this period. Despite the fact that these towns represent a tobacco producing area and that the crop was moving slowly and at a much lower figure, the farmers were still more than financing themselves if the comparison of their bank deposits and the loans which they had drawn out, is any measure. This reserve or margin between farmers’ deposits and loans was greater in January 1921 than it was the year before. The volume of loans had been reduced by a greater amount rather than the de- posits increased. This may reflect an effort on the part of the farmer to liquidate his obligations with the surpluses secured before the crash in prices came and the calendar year, 1920, did not give enough time for the opposite movement to get under way. Table XVI.— The Local Deposits and Loans for the Banks of Cambridge, Cottage Grove and Deerfield Compared for Farmer and Other Business for the Years 1920 and 1921 Wisconsin Research Bulletin 58 codecs 00(NO GO CO O o o ^ TJH CO r 1-1 co CO 888 CO l>- HCON "The Jan. 1921 figures include a new bank in Deerfield which did not do business through the year 920. Service Relations of Town and Country 31 THE NON COMMERCIAL AGENCIES With the non-commercial service agencies are classed the school, the church and the various social and fraternal organizations or clubs. With the school, only the high school is considered since this is the educational agency which brings the farmer and the town’s people to- gether most definitely. It is with these non-commercial agencies that the associations between town and country are more personal and in- formal. Since the motive of selling the services and the necessity of looking to the town is absent, the relationship is on a rather different basis. In these agencies there is opportunity for the farmer to share in the responsibility of management or leadership. It is in this realm of services also where the farmer has provided more of his own agencies, especially for the religious and the social services, than he has for the merchandising service for example. This will be fully demonstrated in Part III when the service area maps of the towns show restricted limits for these services as well as a larger number of open country centers. The Educational Service. No consolidated schools are to be found in the area under study and only a comparatively few of the rural children living close in attend the town graded schools, therefore the high school is the agency where this educational service relationship between town and country must be observed. First of all it must be said that in none of the to.wns studied is there a high school in which the farmer has a joint share in the management that is, the legal limits of the district extend but little beyond the towns’ corporate limits. This means that the farmer sends his children to the town school by its per- mission and he or his township pays by tuition for whatever kind of education is there provided. That the area for the study of this service might be extended, the 15 high schools of Dane County, exclusive of the Madison schools, and the Waupaca high school were studied. Of all this number only four teach agriculture and only one has its ad- ministration on a community basis, that is one which includes the farming community in its legal area. These high schools have 1,459 pupils from Dane County of which number 52.1 per cent are from the towns and 47.9 per cent from the country. Of the pupils from farm homes 42.3 per cent were boys and 57.7 per cent were girls. The Wuapaca school considered separately, draws 45 per cent of its pupils from the country. The question of the farmer’s participation in the management of these schools is of prime importance. A very small percentage of farmers on the school board could be expected since the legal limits in all but the one case, extend so little beyond the town. By count of the 49 school board members, 13 or 26.5 per cent were farmers, some may have been retired farmers living in the towns but these were not counted. But what of the costs? Put on the basis of total cost per pupil, the average was $179 for the year 1921. Waupaca had a cost of 32 Wisconsin Research Bulletin 58 $168 per pupil. The cost to the farmer on this tuition plan is most de- cidedly a factor in the situation. On this basis, he hires the town school at a flat rate of $72 per year per child. This would seem to constitute a saving of a little over $100 per year. One illustration is enough to demonstrate the complete fallacy of such an economy regime. One county superintendent recites the story of a painful task to which he was recently called, that of writing a recommendation for one of the most promising high school graduates in his county to a clerical position in Milwaukee. This candidate was a farm boy full of promise for leadership. He liked the farm and wanted to take agriculture in the high school. It was not given. He became interested in the popular commercial department. This rural community has lost a leader. The Religious Service. Two agencies devoted to the religious service were studied in only the seven Dane County towns with Wau- paca as a check, namely the church itself and the agencies for religious education. In the ten schools for religious education it was found that only 46.2 per cent of the children were from farm homes with the balance of 53.8 per cent from the towns. This proportion is lower than the division for church membership. Of the lay leaders or teachers a very small per cent, 12.4, were from the country, with the 87.6 per cent from town. Here is also evidence that the farm family for some reason or other is delegating a very important function. The church membership for 15 institutions in- these same towns has a division of 51.5 per cent and 48.5 per cent in favor of the town. Strange as it may seem the leadership as represented by members of official boards of the laity, 55.9 per cent were farm people. The mem- bership for 1921 averaging about 180 per church is an increase of about 30 for each church over an estimated previous five year average. The attendance records were such that too much dependence cannot be placed in them. If the estimates given are used, and they are almost sure to be sufficiently liberal, almost exactly one-half of the members are in the habit of attendance. The seating capacities of the churches on the other hand, are such as to accommodate about two and one-half times as many people as are accustomed to appear for the regular services. The simple economy of this situation will sooner or later call for readjustments. The contributions for local expenses and for out of the community benevolences amounted to $10.60 per member for the year 1921. This is somewhat less than the per capita contribution for Price and Sheboy- gan Counties, Wisconsin, as revealed in studies made by the Committee on Social and Religious Surveys. 8 For three Waupaca churches the total contributions were $12.50 per member. Of the total contributions 70 per cent was used for the local work and 30 per cent sent out for benevolences. The average salary paid the clergy over and above the use of a dwelling was $1,215 annually. 8 Fry, C. Luther, The New and Old Immigrant on the Land, 1922, p. 110. Table XVII.— Service Relations of Town and Country 33 34 Wisconsin Research Bulletin 58 The Sociability Service. The agencies engaged in what may be termed the sociability service including various k.nds of social, recrea- tional or fraternal activity, are such as clubs, lodges and informal or- ganizations. In the Dane County towns, 30 of the agencies were studied. The membership canvass showed that 40 per cent were drawn from the country and 60 per cent from the town. The leadership as represented by the officers gives only 15.3 per cent to the farmer and 84.7 per cent to the townsmen. The record of contributions or mem- bership fees showed $11 given per member. There appeared to be' no way to check the distribution of these receipts between town and country. The participation and the leadership in these non-commercial activities for town and country comparison are assembled in Table XVII where the banking service is also included by the records of de- posits. and numbers on the boards of directors. The conclusion is difficult of escape that although the farmer is participating about equally he, is not assuming his proportionate share of the leadership. THE TOWN AS A TOWN AND SERVICE STATION How is a town different from a mere aggregation of agencies? Up- on the answer to this question depends much regarding the future of town and country relations. The ideal answer probably would be to the effect that a small town is made by the spirit of unity and a common sense of loyalty to its fundamental service purpose. These elements are' difficult of measurement and comparison. Certain objective characteristics are, however, quite easily distinguishable. Town and Community Enterprises. To carry to success certain enterprises involving the whole town or community depends upon a certain amount of cooperative effort irrespective of the immediate cash return to individual concern or agency. Probably the easiest contrasts appear in such enterprises as the parks, the approaches to the town, the streets, the utilities such as adequate lights or telephone service. Judged by such standards the small towns of Dane County present striking comparisons. Further study is needed before any complete classification can be made on this basis. Other evidences of town unity are such undertakings as athletic teams, bands, town loyalty to schools, community festivals or fairs. Cambridge has been an example of success with a large community festival bringing town and country together. One small yet significant feature in this program is a long established custom of exchange of hospitality. One year the merchants entertain the farmers and their families and the next, the farmers are hosts to the merchants and their families. Waupaca by an organization known as the “Hustler Plan” has by means of contests and cooperative programs succeeded in tying to- gether and to the town all the rural schools of the community. Other Service Relations of Town and Country 35 successful enterprises could be narrated but this would become a study in itself. Perhaps experience with the uniform closing plan typifies the opposite condition in which some towns find themselves. One illustra- tion will be sufficient. After much bickering back and forth it was finally decided in one town to close all day on holidays. Two merchants of greater faith than others regarding the success of the plan went fish- ing. Toward noon, since the fish were not biting well enough to com- pletely engage their attention, a feeling of suspicion that perhaps com- petitors had not kept faith and had opened their stores, began to creep over them. They abandoned hook and line and went home and sure enough, the store across was open and one customer was buying a paper of pins. Thus ended cooperative effort within that town as well as its community enterprises. Some Details for the Three Towns. Close study of all the service agencies in the three towns of Cambridge, Cottage Grove and Deerfield, show certain characteristics with reference to this element of com- munity organization. Cottage Grove because of its rather restricted size, its closeness to the city of Madison and the fact that it is not in- corporated, is not directly comparable with the others. Both of the other towns report well over a million dollars worth of business for the year 1920. Deerfield has a larger volume and a larger percentage from the farmer. This is largely due to the fact that Cambridge is not on a railroad and secures during the summer a considerable trade from summer campers and tourists. In contrast with amounts, Cambridge has more agencies for every kind of service. The difference in the number of social agencies as seen in Table II is the most prominent. The cond.tion of the merchandising service as a whole for these three towns is pictured in a graph, Chart VI. The total yearly business comparing the amounts from farm and town are shown for 62 concerns. The horizontal bars are plotted on the basis of average business per concern and on the vertical is the number of stores in the volume class. It is quickly seen that a few stores do most of the business. To be ex- act, the first five classified on the basis of highest sales for the- year 1920, do 23 per cent of all the business. The first ten concerns do 40 per cent of the business. The result is a large number of people en- gaged in the retailing service in a large number of small stores of about the same standard of inefficiency and the same grade of ordinary merchandise. Since the prices tend to gravitate to the level of the operating margins of the least efficient or marginal concerns, a burden rather than a service benefit is saddled on to the community. The concrete application of this is recognized by the farmer customer. One story will illustrate the point. A farmer’s wife went to the nearby town to buy a pair of shoes. There were five stores in town which carried shoes. She visited them all. She wanted a particular type of nationally advertised, dress shoe. Most of the shoes shown her were very ordinary although the salesmen argued that they were the best on the 36 Wisconsin Research Bulletin 58 Chart VI. — The Total Merchandising Business for Cottage Grove, Cam- bridge and Deerfield Analyzed by Volume and the Average Amount Done per Store in Each Class. market. She knew better. The town lost a sale and a custodier lost her patience. “They seem to think that anything is good enough for a farmer’’, she said. Service Relations of Town and Country 37 Influences of Outside Agencies. No town is a local town any more. Each unit represents a link in a long chain of state and national re- lationships. There are state, national or international conclaves of all kinds for the social and fraternal organizations. The local members make up the majority of the listeners rather than the exorters but they are constantly being exposed to outside influence, to stimulation through competitive feats and to an increasing demand for standarized pro- cedure in the local unit." The churches through their state or national, denominational boards of various kinds are under much the same in- fluences. The business man himself is by no means exempt. There is the traveling salesman from the wholesale house in Milwaukee. One such firm has had the same salesman in the territory studied for twenty-five years. The policy ‘of merchandising which this house be- lieves in can be recognized at once in the stores in these local towns. Then there is the whole system known in the trade as “Service.” Not only does the local merchant buy goods from over- head agencies but he buys their “service” such as standardized window display, circular letters with the name of his, own store included, ad- vertising materials with highly illustrated color plates entirely beyond uie printing capacity of the local print shop, the store newspaper written in the Chicago distributor’s office for a whole area and finally the “service literature” of all kinds and descriptions. Again the local "progressive” merchant at least, is a member of a trades organization and is on its mailing list for literature and the magazine. The influence of these associations were clearly seen in the merchandising practices, theories, and attitudes of the men in the towns studied. Three of these associations which were most evident in their influence were the Wis- consin Retail Hardware Dealers’ Association, the National Association of Retail Clothiers and the Retail Lumber Dealers’ Association. One concrete instance of this influence is the accounting system urged up- on the local hardware dealer by his state association. Where this system was in use the information for this study could be secured in fifteen minutes. All of this then indicates a tendency for bringing the small town into a scheme of relationships where its function is that of a distributor of services of all kinds ranging from groceries to religion, to the ultimate consumer after the fashion worked out by experts in larger overhead agencies. Responding to Community Needs. In order to discharge this function of distributing services to the community, the local town and its agencies can not become wooden and attempt to pass on any pre- digested product suggested by the outside producer, but it must be- come acquainted with what the local community needs. The retailer of services cannot stand behind his counter, his desk or his pulpit and wait for his clients to ask for and explain the things which they need. He must anticipate these requirements. He must associate with the busi- ness, civic or social organizations of the community and extend his in- terests beyond the corporate limits of his town for his is a role of 38 Wisconsin Research Bulletin 58 leadership. This leads directly to the third part of the study, an in- vestigation regarding what kind of a town the farmer really needs and wants. In fact the project was delayed a year in order that this service area of the town might be studied and correlated with the study of the town itself. Service Relations of Town and Country 39 PART III THE FARMER HAS HIS TOWN What Kind of a Town Does the Farmer Really Want? T HAT THE farmer wants his own organizations for certain kinds of services is evidenced on every hand, nevertheless it is equally evident that for certain other kinds of services he realizes his dependence upon nearby villages., towns, or cities. It has often been suggested that the farmer needs his town, but what kind of a town does he really want and what are the kinds of services for which he must turn to the town? A summary classification of six services has been made for the purpose of studying this town and country relationship. They are the economic, including merchandising, marketing and financ- ing; the educational, the religious, the social, the communication and that of organization activity itself. In seeking an answer to this question as to what kind of a town the farmer really wants, or stated in a Lttle different way, what are the factors which to a more or less degree, determine the trading and the social habits of the farmer with reference to the nearby towns, 787 farm families were personally visit- ed and questioned. The families were distributed in three counties and over the general trade areas of six villages, towns or cities. In Dane County four trade centers were selected which border upon one another and which lie within motoring distance of the larger city of # Madison. To act as a check for bringing out certain contrasts or comparison the Elkhorn trade area in Walworth County and the Waupaca area in Waupaca County were selected. Obviously not all the farm families within the various areas of town influence could be visit- ed, therefore the plan followed was to interview those families living near the margins, or away from the towns in question a sufficient distance to have a real choice as the result of being comparatively near another town. When studying a particular town the investigator followed out each road beyond these margins until the influence of the town respecting any or all of the six services was completely lost. It is recognized that the answers to some of the questions are a registration of opinion and that opinion is subject to suggestions of change but it is a truthful report of the reasons existing in the minds of the people regarding their relations to their towns. To be sure many did not really know why they went to this or that town for such and such a service, but the report of this lack of conscious decision re- lating to many important economic and social affairs of family and com- munity life are quite if not more significant and important for study. 40 Wisconsin Research Bulletin 58 Maps and Tables Tell the Story. The purpose is to tell the story of the replies received largely by maps and tables. The maps are in- tended to answer the question of “how far” the town extends its various services. The tables give answer to “the why” the farm family goes to the particular town or to any town for the various services The first three large maps, Figures 2, 3 and 4, show the three areas studied FIG. 2.— THE DANE COUNTY AREA— SOUTHEASTERN SECTION 307 Farm families were visited in this area. The red lines indicate state and county trunk highways. Service Relations of Town and Country 41 GENERAL TRADE AREA ® SMALL CENTER MAXIMUM SERVICE AREA ° FARM FAMILIES VISITED TOWNSHIP BOUNDARIES FIG. 3.— THE ELKHORN AREA IN WALWORTH COUNTY 212 Farm families were visited in this area. The red lines indicate state and county trunk highways 42 Wisconsin Research Bulletin 58 GENERAL TRADE AREA MAXIMUM SERVICE AREA TOWNSHIP " BOUNDARIES • SMALL CENTER FARM FAMILIES 5 VISITED i] 11 1- X F •/ \ / /. <• Inti > vT / ''' 3!XZ FIG. 4.— THE WAUPACA AREA IN WAUPACA COUNTY 268 Farm families were visited in this area. The red lines indicate state and county trunk highways Service Relations of Town and Country 43 indicating in each case the location of the families visited, the general trade areas and the maximum service of each of the six towns under study. The smaller maps are grouped according to the service under consideration and the six towns brought together for comparison. The large swinging curve on each map includes what shall be termed the maximum service area. This was formed by connecting the outer points to which any of the specialized services extended. This area is made a constant for all the service maps and the areas of the different services are platted within this for purposes of easy comparison. The tables in general have been arranged in series of three for each of the services under observation, for example, in the matter of groceries, first, reasons for selecting any town as a trading center for groceries, second, reasons for selecting the particular trade towns as places to buy groceries, and third, reasons for buying groceries in centers other than the six trade towns under surveillance. Not all of the maps nor all of the tables could be presented in printed form, therefore where comparisons were regular or where one service or set of reasons follow another very closely, the matter is handled with only a brief statement. THE ECONOMIC SERVICE General Trade. The first and leading question was in regard to general trade and was put in such a way as to virtually amount to the question of what the family regarded as its “home town”. Table XVIII indicates that the factor most prominent in determining the farmers’ Table XVIII. — The Farmer’s Reasons for General Trade or “Most Trading” in Any Town “Most Trading” Distributed by the County Areas in Per Cent and in Number Reasons for general trading in a town IN PER CENT IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth county Wau- paca county Total 100 100 100 100 787 307 212 268 Nearest 48.7 53.6 27.6 59.7 383 16434* 5834 160 Better price 10.2 2.4 19.8 11.4 80 734 42 3034 Best service 8.3 4.9 15.6 6.5 6534 15 33 1734 Friends or relatives 8.0 10.4 10.1 3.5 63 32 2134 934 ‘Always traded there” 6.7 4.7 9.7 6.7 53 1434 2034 18 Best roads 4.2 6.8 4.0 1.6 33 2034 . 8 34 4 Larger town 2.2 3.7 .4 1.7 17 1134 1 434 Most convenient 2.0 3.2 1.9 .7 16 10 4 2 Unspecified 9.7 10.3 10.9 8.2 7634 3134 23 22 *The fraction of 34 indicates a half coant for each of two reasons, in other words some people gave two answers to the same question so each was given its due weight. relation to the various centers was “nearest”. This was comparatively uniform for Dane and Waupaca Counties while in Walworth County it was apparently only about half as strong. This was due to the greater facility of travel in this latter area resulting from better roads which 44 Wisconsin Research Bulletin 58 FIG. 5.— THE GENERAL TRADE AREAS AND THE MAXIMUM SERVICE AREAS The solid black line indicates the general trade area while the broken swinging curve represents the maximum area to which any service extends. The small circles indicate location of open country stands. The arrows show the encroachments from near-by towns. Service Relations of Town and Country 45 meant that “better prices” and “best service” ranked higher. When the six towns were compared the same tendencies were revealed. “Nearest” predominated as the reason, though Elkhorn followed closely the characteristic shown by the larger group ;n its county, Walworth. The area comparisons are shown on both the large maps and the smaller service maps. The maximum service areas, the general trade and the over lap between the areas in Dane County are drawn. By planimeter measuremeilts the general trade areas of the towns fill the respective maximum areas in the following percentages; Cambridge. 60.0 per cent ; Cottage Grove, 47.9 per cent ; Deerfield, 57.0 per cent ; Stoughton, 65.8; Elkhorn, 52.5 per cent; and Waupaca, 49.1 per cent. Comparison by similar means of the general trade lines with those discovered by Dr. Galpin for Elkhorn in 1915 shows that the area has shrunk a little over 12 square miles or about the equivalent of one-third of a township. Differences in methods of measurement would account for some of this since in the earlier study, extreme points were con- nected while in this study the farm was the unit. The general shape has remained about the same excepting the outer limits have been some- what contracted and the area in the direction of Williams Bay ex- panded. The tendency has been one of making the area more compact in the immediate vicinity of the town. 4 Merchandising. The general merchandising service was sampled by studying closely four specialized services, namely groceries, furniture good clothing, work clothes and farm machinery trade. Groceries. The service involved in grocery merchandising usually included the marketing of eggs and in a very few cases butter or poultry. With the comparison by counties in Table XIX “Nearest” again headed the classification. In Dane County this reason clearly predominated. In Walworth and Waupaca Counties it also received Table XIX. — Reasons for Selecting any Town as a Trading Center for Groceries Grocery Trading Distributed by the County Areas in Per Cent and in Number grocery town IN PER cent ‘ IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane . county Wal- worth county Wau- paca county Total 100 100 100 100 787 307 212 268 Nearest 26.2 33.6 25.5 18.5 20634 103 54 4934~ Most convenient 11.8 25.4 3.5 2.6 9234 . 78 734 7 Best goods 9.2 10.3 .5 14.9 7234 3134 1 49 Better price 7.2 4.7 7.5 9.7 5634 1434 16 26 Friends or relatives 3.9 7.0 .3 3.4 31 2134 34 9 Market for produce 3.4 4.7 1.4 3.4 2634 1434 3 9 Best service 2.7 1.6 6.6 .9 2134 5 14 234 Groceries by truck 2.2 .5.5 17 17 Unspecified 33.4 7.2 54.7 46.6 263 22 116 125 major consideration in the answers given but scarcely 50 per cent of those interviewed were ready to specify a definite reason. When the 4 Galpin C. .T. The Social Anatomy of an Agricultural Community, Res. Bui. 34. Agr. Exp. Sta. Uni. Wis., Madison, 1915. Table XX.— Reasons for Selecting the Particular Trade Towns as Places to Buy Groceries Wisconsin Research Bulletin 58 The Six Trade Towns Compared in Per Cent and in Number In Number WAU- PACA COUNTY Wau- paca 00 j ^ -t 1 — i co j WAL- WORTH COUNTY Elk- horn § 5 DANE COUNTY Stough- ton 74 j S CO Deer- field 35 HOOThHWCONHH Cot- tage Grove 33 XJR : 1 XS 1^0^ : :hhh Cam- bridge CM 3SR SK j j to GO O (M ^ All trade towns £ co X IS X OONiO^NOONOO tDCO^NHHH a> In Per Cent WAU- PACA COUNTY Wau- paca 100 OO GO *© OO OO CO Tf< T— ( CO to ^ I CM CM to to WAL- WORTH COUNTY Elk- horn 100 to : t>- »-« : OO cm : ^ t>» to 00 H S3 Stough- ton 100 5rt° lH H HHCD H Deer- field 100 OOCO^O^-HCOt^COOi CM^T-nCMt^OOiO^CM M(NrH o o H fc Q Cot- tage Grove 100 irtCicO : O : O to to cm r-H co : co 'CCOh Cam- bridge 100 o *-4 cm oo co CONN^HCOIO t-h CM CO All trade towns } > 100 t^-GOCOt^CDCMCOCMCM fH rH Tf CD to CO CO (M rH CM CO Reasons for Selecting the particular town Total Nearest Most convenient Best goods Best price Friends or relatives Market for produce Best service Groceries by truck Unspecified Table XXI. — Reasons for Buying Groceries in Centers Other Than the Six Trade Towns Service Relations of Town and Country 48 Wisconsin Research Bulletin 58 six trade towns were seperated, “nearest” continued to be the chief reason for Cottage Grove, Deerfield and Stoughton areas while “best goods” ranked highest in the Cambridge and Waupaca areas. “Price” received consideration in the Elkhorn area but the percentage of replies was comparatively small. The six trade towns were contrasted with other centers including open country stands, nearby competing towns of similar size, large city centers or mail order houses, when the reasons for such successful competition appear to be “Convenience” and “nearest” of location. The area comparisions intended to show the extent to which the trade towns are able to extend this grocery service, reveal that Cam- bridge fills 43.1 per cent of its maximum service area, Cottage Grove 40.0 per cent, Deerfield 44.5 per cent, Stoughton 41.4 per cent, Elkhorn 63.6 per cent and Waupaca 44.2 per cent. With the exception of Elk- horn the uniformity is striking. Furniture. The reasons determining this trade service in Dane County continue to be the matter of nearness of location though it is in a much less decided proportion than in the case of groceries. In the Walworth County area it is very clearly a matter of “price”. In the Waupaca area the answers given were very small due to the fact that little furniture had been purchased for some time previous to the making of the filed study. Table XXII gives the complete figures. Table XXII— Reasons for Selecting any Town as a Trading Center for Furniture Furniture Trading Distributed by County Areas in Per Cent and in Number Reasons for selecting any town IN PER CENT IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth county Wau- paca county Total 100 100 100 100 787 307 212 268 Best goods 10.8 16.1 11.3 4.3 85 4934 24 1134 Nearest , 10.1 22.6 1.4 2.6 79^ 693/2 3 7 Better price 9.8 6.0 23.4 3.5 77M 183/2 4934 934 Friends or relatives 7.1 12.9 7.5 55^ 3934 16 Most convenient 4.3 9.8 1.7 3334 30 334 Larger town 2.6 6.7 20 Vi 2034 Unspecified 55.3 25.9 54.7 89.6 43534 7934 116 240 The areas of the furniture trade for the small towns of Cambridge and Deerfield were well restricted. Elkhorn and Stoughton were able to distribute nearly to the borders of their maximum services areas. Waupaca extends its service well over the southern portion of the area but meets difficulties of encroaching competition to the north. Clothing. The investigation of the service of merchandising in clothing included both good clothing and work clothes and for both Service Relations of Town and Country 49 FIG. 6.— THE ECONOMIC SERVICE-GROCERIES 50 Wisconsin Research Bulletin 58 Table XXIII. — Reasons for Selecting Ant Town as a Trading Center for Clothing of All Kinds* Clothing Trade Distributed by the County Areas in Per Cent and in Number neasuus xur selecting clothing town IN PER CENT IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth county Wau- paca county l Total 100 100 100 100 3148 1228 848 1072 Best goods 17.5 27.6 1.0 18.9 550 33834 834 203 Variety and selection 9.9 9.3 11.3 9.7 312)4 11334 9534 10334 Best price 9.7 7.8 13.4 9.0 306 96 11334 9634 Nearest 7.6 10.8 .5 9.8 240 13034 434 105 Most convenient 5.2 10.4 1.9 1.8 164 my 16 1934 Friends or relatives 2.4 4.2 1.5 1.0 75y 2 52 1234 11 Unspecified 47.7 30.1 70.4 49.8 1500 369 59734 53334 *This table as well as the two following are summaries of the answers to four questions regarding the clothing trade, namely, good clothes and work clothes for both men and women. This makes the totals just four times larger than the total number of schedules in each area. Table XXIV. — Reasons for Selecting the Particular Trade Towns as Clothing Centers Clothing Trade of the Six Trade Towns Compared in Per Cent and in Number Reasons for selecting the particular trade town A IN PER CENT All trade towns Dane County Wal- worth county Wau- paca county Cam- bridge Cot- tage Grove Deer- field Stough- ton Elk- horn Wau- paca Total 100 100 100 100 100 100 100 Best goods 22.9 9.3 4.3 11.6 5.3 1.8 44.8 67.3 61.9 45.3 4.4 19.3 3.1 1.7 16.8 6.8 5.1 47.2 .3 7.2 10.7 .3 3.2 .6 77.7 27.0 16.3 3.4 16.0 1.3 .2 35.8 Variety and selection Best price Nearest Most convenient Friends or relatives 1 nsi. ecified 4.5 16.3 5.5 6.4 14.3 12.4 1.9 9.5 10.1 25.1 3.1 12.0 B IN N UMBER Total 139034 55 5234 7934 30334 332 568 Best goods 31834 12934 5934 161 74 25 623 37 3234 36 5834 934 5 51 2034 1534 14334 1 15334 9234 19 91 734 1 20334 Variety and selection 334 8 20 234 934 24 3534 1 1034 2 258 Best price Nearest 234 9 3 334 734 634 1 5 Most convenient Friends or relatives Unspecified men and women. In “good clothing” is included womens’ "ready to wear” and suits and coats for men and women alike. The reasons given for going to any center for any kind of cloth- ing are shown in Table XXIII. The figures are a summary of four questions relating to good clothes and work clothes for both men and women. “Best goods” is the leading answer for the Dane and Waupaca Service Relations of Town and Country 51 FIG. 7.— THE ECONOMIC SERVICE— FURNITURE. 52 Wisconsin Research Bulletin 58 Table XXV. — Reasons for Trading in Clothing of all Kinds in Centers Other Than the Particu- lar Trade Towns The Trade Towns and Other Centers Compared in Per Cent and in Number Reasons for selecting a particular center A IN PER CENT 1 All trading centers The six trade towns Com- peting towns Small open country centers Large - city center Mail ojder houses Un- specified Total 100 100 100 100 100 100 100 Best goods 17.8 22.9 14.9 28.4 20.8 4.1 Variety and selection 10.2 9.3 6.7 29.3 2.2 Best price 9.8 4.3 9.2 12.6 5.9 58.9 Nearer 7.9 11.6 5.4 13.4 7.4 Most convenient 5.2 5.3 3.5 22.0 6.8 7.6 Friends or relatives 2.5 1.8 3.2 1.6 5.7 .4 Unspecified 46.6 44.8 57.1 22.0 24.1 26.8 100 B IN N UMBER Total 3148 139034 838 34 63 43934 231 185 Best goods . 562 318*4 12434 18 9134 9 34 Variety and selection 31934 129 34 5634 12834 5 Best price 307 59}4 7734 8 26 136 Nearest 247 161 45 8 34 3234 Most convenient 165 74 2934 ■ 14 30 “ 1734 Friends or relatives 7834 25 2634 1 ' 25 1 Unspecified 1469 623 479 14 106 62 185 County areas while “price” is the chief consideration in the Walworth County area. Both the good clothing and the work clothes followed this ranking when analyzed separately. When the six trade towns were compared by themselves “best goods” continued to lead in five of the town areas while the Elkhorn area stayed in line with its count}’ answers with “price” the determiner. The summary of competing centers shows “best goods” also in the lead for the cases of other near- by towns and even for the open country stands. The reasons given for large city trading in clothing is “better varieties and more selection” and with the mail order trade, “price” is the overwhelming considera- tion. The widest contrast in area distribution of services appears in the item of clothing. The extremes are presented by the good clothing areas for Cottage Grove and Waupaca. Cottage Grove is able to extend this service but little beyond its own village limits. Waupaca on the other hand spreads the service until it almost completely fills the maximum service area. In fact it is this one service which has determined more than any other, the extended points of this maximum service area, particularly to the north of the town. Between these two extremes the other towns arrange themselves in a definite order. The smaller centers of Cambridge and Deerfield are not able to extend this area as widely as those of work clothes and groceries, while Elkhorn and Stoughton seem to represent half way points where good clothing, work clothes and grocery areas are approximate^ coincident. There Service Relations of Town and Country 53 FIG. 8.— THE ECONOMIC SERVICE— WORK CLOTHES 54 Wisconsin Research Bulletin 58 FIG. 9.— THE ECONOMIC SERVICE— GOOD CLOTHES Service Relations of Town and Country 55 would seem to be a rather direct relation between the size of the town and its good clothing service area. Stoughton becomes an ex- ception due to the fact that it is surrounded by more active competition than is Waupaca. The area for work clothes in contrast to good clothing coincides almost exactly with the grocery areas for each of the six towns quite independent of their size or location. The detailed measurements of the good clothing service in com- parison with the maximum areas for the various towns is as follows : Cambridge, 19.3 per cent; Cottage Grove, 3.5 per cent; Deerfield, 21.7 per cent ; Stoughton, 65.0 per cent ; Elkhorn, 66.1 per cent ; and Wau- paca, 93.4 per cent. Machinery. “Good agent and service” are the leading reasons given for going to any center for buying farm machinery. Table XXVI in- Tabi/e XXVI. — Reasons for Selecting Ant Center for Trading in Farm Machinery Machinery Trading Distributed by Areas in Per Cent and in Number Reasons for trading in machinery IN PER CENT IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth county Wau- paca county Total 100 100 100 100 787 307 212 268 Good agent and service 19.6 34.4 3.1 15.7 154 10534 634 42 Better price 17.3 8.1 35.3 13.4 136 25 75 36 Nearest 11.7 16.8 2.1 13.6 92^ 5134 434 3634 Friend or relative 4.9 9.9 .5 2.6 3834 3014 1 7 Most convenient 3.7 4.1 .5 5.8 29 1234 1 1534 Unspecified 42.8 26.7 58.5 48.9 337 82 124 131 dicates that Walworth County, however, insists upon the exception and “price” is given the first place. The area of distribution for this kind of merchandise appears in every case to be somewhat larger than the grocery and more nearly equal to the general trade areas. The smaller Table XXVII. — Reasons for Selecting Any Town as a Banking Center Banking Distributed by the County Areas in Per Cent and in Number Reasons for selecting a town IN PER CENT IN NUMBER All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth county Wau- paca county Total 100 100 100 100 787 307 212 268 Trade center 30.9 22.1 52.8 23.5 243 68 112 63 Nearest 20.0 41.1 .7 11.0 157 126 134 15 2934 Most convenient.. . . 7.2 10.1 7.1 3.9 5634 3634 34 31 1034 Friends or relatives 4.6 3.1 8.5 3.4 9H 1334 534 234 51 18 9 “Alwavs have” 4.3 4.4 4.5 9.4 4.1 9M 20 11 Best service 4.1 1.8 2.6 1.1 3234 8'A 219 7 Stockholder , . 1.1 .8 1.4 3 3 Unspecified 27.8 16.6 15.6 50.4 33 135 56 Wisconsin Research Bulletin 58 FIG. 10.— THE FINANCIAL SERVICE— BANKING Service Relations of Town and Country 57 FIG. 11.— THE MARKETING SERVICE— PRODUCE OR SHIPPING CONCERN 58 Wisconsin Research Bulletin 58 towns as Cottage Grove for example, seem to be especially successful in pushing out this service beyond the limits of some of the other services because they are apparently able to compete with the larger cities such as Madison. This is borne out by the answers where “nearness and convenience” figure strongly. Financing. A number of services offered by towns to surrounding country come within this caption but only banking will -be reported. Banking. The generalization regarding banking as one of the financial services offered to farmers by a town is to the effect that it follows general trade in both considerations of extent and of reasons. Table XXVII shows “trade center” as the chief reason for banking in any town and this for all areas except Dane County where “nearest” is slightly in advance. This answer is not so different, however, than “trade center” since it was “nearest” which was the chief factor in in determining the “general trade” service. When the six town areas were compared singly, each followed the tendency of answers for the county of which it was a part. The “trade center” reason continued to lead also when “other centers” were summarized such as the nearby towns and even the large city centers. In the area comparisons the banking service continued to follow the trade areas rather closely. The Stoughton area inclines to become somewhat smaller due partially at least to the competition offered by the several surrounding towns. Marketing. The marketing service was examined with reference to such general products as milk, butter and eggs in all areas and such particular products as tobacco in Dane County and potatoes in Waupaca County. Judged by the area comparisons this service tends to follow the trade areas although it is uniformly somewhat smaller. The grocery area and the marketing area more nearly coincide. The grocery service seems to feel the competitive influence of small centers while marketing areas are more often encroached upon by the larger nearby towns as is illustrated by the Lake Mills encroachments into the Deerfield area. Cambridge is under a decided disadvantage in this service due to its lack of railroad transportation. THE COMMUNICATION SERVICE Roads. The large maps showing the areas studied indicate the distribution of state and county trunk highways. In Dane County all roads including the railroads lead to Madison. This city stands at the hub of a radiating communication system. Stoughton is located on two trunk lines, one state and one county. Cambridge has a state highway which becomes a decided .asset in its service relations. Cottage Grove although having a county road lies between two important arteries and is somewhat at a disadvantage especially in its service areas in the direction of Madison. Service Relations of Town and Country 59 FIG. 12.— THE COMMUNICATION SERVICE— MAIL DELIVERY 60 Wisconsin Research Bulletin 58 Service Relations of Town and Country 61 Waupaca has four state and two county highways entering its corporate limits. The direct influence of these roads in bulging out the maximum service area at various points especially to the north is very clearly in evidence. In the Elkhorn area facility of transportation has reached its highest development. Not only do three state and one county highway enter the town but much of the surface of its high- ways are concrete and the influence was evidenced by the greater distance which people in this area were willing and accustomed to travel as compared to the other counties. Rural Mail Delivery. The service area maps for governmental free delivery of mail do not seem to possess any direct relationship to any other of the service areas. The total extent of the areas are not so different from the general trade areas for example, but their dis- tributions are completely different. Cottage Grove has a narrow elongat- ed area while Deerfield and Elkhorn are restricted on one side almost to their corporate limits. The consequence of this arrangement means that often the town cannot maintain direct communication with the farm families in its service area. Such families claiming a town as their center must accept another town as their post office address and must expect the mail from their own town to be delayed at least a day. Table XXVIII is a percentage table representing the proportion of families who received their mail from the center which they claim as their general trade center. The families interviewed, it will be recalled, are those living near the border of the trade areas. Those families located nearer the center, of course, usually get their mail direct but after all, it is those outlying families which are really important to the town because of possible competition from other centers. The percentage of people claiming the town as trade center, who also have it as their post office, ranges from 71 to 87 per cent. The counterpart of this table would be that the towns in question supply mail to families claiming membership in trade areas of border- ing towns. Telephone Service. The telephone exchange areas generally follow more nearly the general trade lines than does the mail service. Elk- horn has been especially successful in extending this service well toward the limits of the maximum service area. Stoughton, on the other hand, is much hemmed in on the north by a strong independent company with an exchange at Kegonsa. The Local Newspaper. No maps or tables are presented for the newspaper service because of the difficulty of showing clearly the wide areas covered and the many overlappings. In every case with the possible- exception of Elkhorn where the newspaper area tends to follow the “Rural Free Delivery” area, the laboratory maps show that the local press service of each of the six towns extends well beyond the limits accepted as the maximum service boundaries. Cottage Grove does not have a local paper. The two Waupaca news sheets have a 62 Wisconsin Research Bulletin 58 FIG. 13 — THE COMMUNICATION SERVICE— TELEPHONE Service Relations of Town and Country 63 FIG. 14.— THE EDUCATIONAL SERVICE— HIGH SCHOOL v o 3 64 Wisconsin Research Bulletin 58 FIG. 15.— THE RELIGIOUS SERVICE -CHURCH Service Relations of Town and Country 65 circulation covering virtually the whole county. Dane County is also pretty well covered by Madison as well as Milwaukee dailies. THE ORGANIZATION SERVICE The High School. The educational service of the town only through its high school will be considered. The elementary schooling for the farm children in the area studied is given by the district schools and by the graded schools for town children. There are no consolidated schools in the areas. The service areas mapped, it should be observed, do not represent legal areas but rather the areas from which farm children are drawn to the town’s high schools. The local townships pay the tuition charge for this service and the farmers have no voice in the management of the schools. A schedule canvass of all the high school pupils in Dane County as well as those in the Waupaca schools shows that the predominating reason for attending this or that town high school was because it was the “nearest” high school. In Dane County 53 per cent of the farm children travel less than four miles to high school. The great majority of these are at home night and morning to help with the work. The distance from the school appears to vary directly with the size of the town and the school. THE RELIGIOUS SERVICE The Church. The service areas of the towns respecting church attendance are all comparatively smaller than the other areas studied up to this point. Elkhorn appears to be an exception but farm families having church connections in Elkhorn are sparcely distributed over the occupied area. One reason for this more limited service of the Table XXIX. — Reasons for Selecting Ant Particular Center for Church Attendance Church Attendance Distributed by the County Areas in Per Cent and in Number reasons ior selecting a particular center IN PER CENT IN NUMBER Total Dane Wal- worth Wau- paca Total Dane Wal- worth Wau- paca Total 100 100 100 100 787 307 212 268 Preferred denomination lo- cated theri - 18.2 17.9 2.8 30.6 143 55 6 82 Nearest 16.4 17.3 4.7 24.6 129 53 10 66 Always attended there 16.2 28.0 5.6 11.2 128 86 12 30 Prefer country church 7.1 .7 22.1 2.6 56 2 47 7 Friends or relatives 5.9 10.4 4.2 2.2 47 32 9 6 Good pastor 2.5 5.9 .4 .4 20 18 1 1 Prefer city church 1.0 3 3 Unspecified 21.0 12.0 40.0 15.0 161 37 84 40 Don’t go 12.7 6.8 20.2 13.4 100 21 43 36 66 Wisconsin Research Bulletin 58 town to the country is the presence of a considerable number of open country churches. The Deerfield and the Stoughton areas are even more striking examples of this influence. The big Liberty Prairie open country church draws heavily from the Deerfield maximum area and the Koshkonong churches, four in number and known as “the East” and “the West”, cut down the Stoughton area. In Walworth County about Elkhorn the several open country churches are often known as “Community” churches where the local neighborhood con- sciousness is the bond rather than nationality linked with a particular type of denomination as is the case in many places in Dane County. The tabulations of the reasons for selecting any town or center for church attendance, Table XXIX, shows “preference for denomina- tion located there”, at the head of the list. This answer really means that the farmer goes to this or that town to church because the congregation of his faith gathers there. The next reason in order of frequency is “nearest” irrespective of denomination qualifications. There can be small doubt that tradition also plays a large part in the service arrangement since- the reason “always attended there”, bulks rather large. It is significant to note the variations in the percentages of “don’t go” and “unspecified”. Walforth County heads the list with high percentages in both counts. Waupaca County comes next and is followed by Dane County which has a rather low figure of non-at- tendance. Another phase of the general service which should be studied is that of religious education. THE SOCIAL SERVICE Within the social activity service is included a number of activities with the sociability idea uppermost, such as parties, socials, movies and other informal recreational affairs. From the mapping it can be seen that this service area for the town is rather uniformly more re- stricted than the general trade areas and more nearly approximates the grocery or the high school areas in extent though of different dis- tribution. In the Elkhorn and Waupaca areas the open country com- munity centers, clubs, or houses are noticeable in their influence. In Waupaca County also the significance of good roads in determining the shape of the social area is clearly seen. A comparison of the social and church relations of the families of the high school children in Dane County tells an interesting story. It is evident from Table XXX that people both of the village and the farm are able to distribute some of their social activity widely. Al- though the high school towns command much of the interest yet other towns come in for their share. In contrast with church activity the social activity is divided more widely between towns. O'f the reasons given by the farm families for se^cting any center for social activity, Table XXXI, shows that “friends 01 relatives” comes first and “preference for local” activities is second. This does not imply Service Relations of Town and Country 67 FIG. 16.— THE SOCIAL SERVICE— SOCIABILITY, FORMAL AND INFORMAL 68 Wisconsin Research Bulletin 58 that much of the social life is not spread over wider territory and that recreational trips are not taken or that social affairs far away from home are not engaged in, but it apparently does mean that for a regular social program the local, personal, face to face contacts count most. The percentage of answers of “don’t go” is rather large and it is very uniform in all areas. Here again it does not indicate that people giving such negative answers have no leisure time' activity at all but rather that they do not count upon it regularly and do not con- sider themselves a part of such a program either in town or at the local center. Table XXX. — Church and Social Relations of Families of All High School Children in Dane County Church and Social Activity and Where Families Live in Per Cent and in Number Place of attendance IN PER CENT IN NUMBER Church Soc dal Church Social In On In On In On In Od tovn farm town farm town farm town farm Total 100 100 100 100 387 430 387 430 Hgh school town 86.4 59.7 58.4 53.7 334 257 226 231 Other towns 10.8 37.0 13.4 6.3 42 159 52 70 Divided between school town and other towns 1.0 2.7 25.6 27.0 4 11 99 116 Not specif! ed 1.8 .6 2.6 3.0 7 3 10 13 Table XXXI.— Reasons for Selecting Any Center for Social Acrivirr Social Activities Distributed by the County Areas in Per Cent and in Number Reasons for selecting any center IN PER CENT IN NUMBER Total Dane Elk- horn Wau- paca Total Dane Elk- horn Wau- paca Total 100 100 100 100 787 307 212 268 Friends or relatives 14.6 32.5 1.9 4.1 115 100 4 11 Prefer local activities 10.8 6.2 12.3 14.9 85 19 26 40 Church center 9.3 18.2 4.7 2.6 73 56 10 7 Nearest 8.9 6.1 21.3 70 13 57 Good club or organization 4.6 6.8 1.4 4.5 36 .... ^ 3 12 Movies 2.4 1.9 2.4 3.0 19 6 5 8 Prefer larger town 1.1 1.6 1.5 9 5 4 Unspecified 33.6 17.6 57.1 33.2 264 54 121 89 Don’t go 14.7 15.0 14.1 14.9 116 46 30 40 THE EDUCATIONAL SERVICE Within the classification of services, rendered by a town as hav- ing to do with farmers’ organizations, is included cooperation and as- sistance in farmer or community clubs, or movements such as Grange, Service Relations of Town and Country 69 FIG. 17.— THE ORGANIZATION SERVICE— FARMERS’ CLUBS AND COM MUNITY ORGANIZATIONS 70 Wisconsin Research Bulletin 58 Equity, Farm Bureau or Breeders’ Association. Such service from the standpoint of the town may be still considered small yet if the town is to be of wide service this form of activity can be included in its program if it is nothing more than serving as a headquarters and meeting place. The area maps of this service relationship for every town under study is very much restricted. Only Elkhorn and Deerfield even come near approximating their grocery areas. This condition is probably due to at least three causes, namely centering of the activity of these clubs Table XXXII. — Reasons for Selecting Any Center for Farmer Club Activity Farmers’ Organization Distributed by the County Areas in Per Cent and in Number xveitauua iui any center in per cent IN NUMBER Total Dane Wal- worth Wau- paca Total Dane Wal- worth Wau- paca Total 100 100 100 100 787 307 212 268 Organization meets there 16.0 13.0 32.1 6.7 126 40 68 18 Nearest 6.8 1.0 .9 18.3 54 3 2 49 Prefer local meetings 2.2 6.0 16 16 Trade town 1.4 3.3 0.4 11 10 1 Unspecified 39.6 73.0 4.3 29.4 312 224 9 79 Don’t go 34.0 9.7 62.7 39.2 268 30 133 105 and organizations at open country points, failure of the town to recognize in this relationship an opportunity for service and finally the very low percentage of farmers vitally interested in this line of en- deavor. In Table XXXII this last point is clearly shown under the captions of “don’t go” and “unspecified.” Even in Walworth County where two years previously one of the organizations boasted of a very large membership and influence, the enumerator could work for whole days without finding a farmer who would confess to member- ship in the movements. COMPETING SERVICE CENTERS The City Center. That the farmer does have trading relations with the larger city is clearly shown by such results as are included in Table XXXIII which includes the answers from all of the 787 farm families. The so-called better grades of clothing very clearly stand out as the leading item for which the farmer looks to the city. The reason for this buying is because he believes he has a “larger variety and selection.” When the Madison buying is separated from the above tabulation the results are seen in Table XXXIV. The percentages are almost identical when this Madison trade is localized and distributed with reference to the town trade areas in which the farm families live. In all the trade areas, “good clothing” is the thing which bulked largest. The percent- age which this item of trade has to all purchasers in Madison, is very uniform for all the areas but the proportionate number of families from Service Relations of Town and Country 71 Table XXXIII— What the Farm Families in All Areas But in the Large City Centers and Why Reasons for buying in large city centers Kinds of Merchandise Purchas sed in Per Cent and in Amount IN PER CENT IN NUMBER All kinds Better clothing * 1 b Q Groceries 1 Furniture and furnishings Luxuries All kinds Better clot hing | Dry Goods | Groceries Furniture and furnishings Luxuries Total 100 100 100 100 100 100 18534 16334 5M 3 934 334 Larger variety and selection.. Better price Friend or rela- tive 30.2 18.2 10.8 9.2 8.2 8.0 15.4 30.0 16.1 11.6 8.9 9.0 8.3 16.1 47.8 26.1 58.3 33.3 25.6 10.3 20.5 10.3 26.7 46.7 56.0 33M 20.0 17.0 1534 U% 2834 49.0 2634 19.0 1434 14M 1334 2634 234 134 134 334 234 1.0 2.0 1.0 1 134 Not available locally Better quality and style Nearest Unspecified 4.3 8.7 13.1 25.0 16.7 13.3 13.3 34 34 % H 34 34 34 Table XXXIV. — What the Farm Families of Eastern Dane County Buy in Madison and Why Kinds of Merchandise Purchased in Madison Reasons for buying in Madison iN PER CENT IN NUMBER All kinds Good clothing | Dry goods | Groceries House furnishings Luxuries All kinds Good clothing | Dry goods Groceries House ( furnishings | Luxuries Total 100 100 100 100 100 100 139 12134 434 3 8 2 Larger variety and selection 29.5 30.0 33.3 37.5 41 3634 134 3 Not available locally 21.6 20.6 50.0 50.0 30 25 4 1 Nearest ... 18.3 18.8 22.3 50.0 2534 23 1 134 Better quality and st le 10.1 10.3 11.1 50.0 14 1234 34 1 Better price 3.9 4.1 16.7 534 5 34 Friend or rela- tive 2.9 3.3 4 4 Unspecified 13.7 12.8 33.3 33.3 12.5 19 1534 134 l 1 the Stoughton area who come to this city for any trading is very much smaller than from the other areas. In Figure 18 the distribution of these zones of influence is pictured. The lines marking off these zones have the appearance of isothermic lines swinging away from certain of the small town centers. Stoughton is able to ward off all the competition in its immediate area excepting the direct mail advertising. The first three zones including the more or less regular trade in groceries, work clothes and the banking swing out away from Madison 72 Wisconsin Research Bulletin 58 about eight or nine miles on the average. The next set of services, furniture and good clothes extend their zones on the average from • nineteen to twenty-one miles, while the advertising is extended beyond the bounds of the county. The Open Country Stand. On the service area maps the open country centers have always been indicated by a circle although no at- tempt has been made to outline the areas of their influence. The presence of these open country stands has been much more in evidence in the non-commercial lines of service such as church, club and social activity although they appear in the grocery areas in the form of cross- roads general stores. Many of these stores, despite general opinion to the contrary, report that their business is good ; the character of the trade has changed a good deal, to be sure, due to the trunk system of highways and the automobile. Many farm families reported that they traded at these open country stores because they were “nearest” and that they went there for what was often termed “convenience” goods and Service Relations of Town and Country 73 Table XXXV. — What tHE Farm Families in All Areas But in the Small Open Country Centers and Why Kinds of Merchandise Purchased in Per Cent and in Number Reasons for buying in the small country centers IN PER CENT IN NUMBER All kinds Gro- ceries Dry goods Cloth- ing All kinds Gro- ceries Dry goods Cloth- ing Total IDO 100 100 100 287^ 275 4 8J4 Nearest Most convenient Market for produce 46.8 35.5 5.9 5.2 2.6 4.0 46.3 35.8 6.2 4.9 2.6 4.2 25.0 50.0 70.6 17.6 134 H 102 17 15 127M 98 M 17 1314 ,7 ny 2 1 2 6 Good line of goods Better price Unspecified 12.5 12.5 11.8 Vi Vi i of course groceries formed the bulk of such orders. Table XXXV shows this relationship decisively. Some of the people when pressed for detail about their relation to this small center admitted that probably they could get along just about as well without it but that while it was there, they would continue their patronage because it was “so handy.” The Mail Order House. Another channel by which the farmer and his family are connected with the outside world beyond his neighborhood and his local town, is the mail order house. Much debate has centered upon this issue of the farmer, the small town merchant and the mail order house. Table XXXVI should therefore prove in- Table XXXVI. — Proportion of Families Trading at Mail Order Houses and the Annual Amount of Their Purchases Families Doing Mail Order Business and Amount of Purchases The town trade area in which families live FAMILIES TRADING WITH MAIL ORDER HOUSES ANNUAL AMOUNT OF MAIL ORDER BUSINESS In per cent In number Average per family Total amount Total 38.8 305 $58.91 $17,968.50 Cambridge 29.8 VA 94.12 800.00 Cottage Grove 6.8 2 82.50 165.00 Deerfield 25.0 9 31.89 287.00 Stoughton 13.8 13 29.24 380.00 Elkhorn 69.3 65^ 54.20 3,550.00 Waupaca 45.9 54 61.34 3,312.00 All other $reas 39.5 153 139. .2 9,474.50 teresting as indicating that 38 per cent of the 787 families interviewed bought on the average $58.91 worth of goods from mail order houses during the year just previous to the time of the study which was dur- ing the summer of 1922. The families claiming Elkhorn as their trade town showed the greatest percentage of mail order buying although not the highest in money value of such purchases. There was every ev.dence that the volume of such purchases had been considerably re- duced during the year under observation as compared to the more 7 4 Wisconsin Research Bulletin 58 Table XXXVII. — What the Farm Families in All Areas Buy from Mail Order Houses and Why Kinds of Merchandise Purchased in Per Cent and in Number Reasons for buying from mail order houses A IN PER CENT All kinds Good clothes Miscel- laneous supplies Work clothes Hard ware supplies Gro- ceries Dry goods Furniture and furnishings Total 100 100 100 100 100 100 100 100 Better price 63.8 68.1 49.6 72.6 73.1 66.0 82.9 66.7 Good quality 15.3 10.6 27.8 2.4 9.0 17.0 4.9 16.7 Most convenient 6.8 8.4 6. ' 4.8 1.5 14.9 9.8 Not available locally 5.5 2.3 9.0 2.4 14.9 8.3 Unspecified 8.6 10.6 7.4 17.8 1.5 2.1 2.4 8.3 B IN NUMBER Total 385 131** 122 42 3334 2334 203* 12 Better price 245 3* ■ 89 34 60 34 303* 2434 153* 17 8 Good quality 59 14 34 1 3 4 1 2 Most convenient 26J4 11 7 34 2 34 33* 2 Not available locally 21 3 11 1 5 1 Unspecified 33 14 9 7 3* 3* 3* 3* 1 prosperous years of the war period. In kinds of purchases and the reasons Table XXXVII tells the story. Good clothes heads the list as to kinds of purchases followed by a somewhat ill-defined group of miscellaneous supplies and sundries. Groceries is fifth down the scale. The reasons for such purchase are very clearly “better prices”, and this is uniform for all the different kinds of purchases. The Store at Your Door. During the summer months in the Dane County area a grocery truck called a “Store at Your Door” travels over all the good roads selling a very complete line of groceries, taking eggs in exchange where it is desired. Some of the trucks also carry fresh meats or will take orders one day to be delivered the next day. Many of the farm women testified to the real convenience of this service during the busy summer months when so frequently extra help had to be served. The influence of this competition was especially evident in the Deerfield and Cottage Grove areas. DEVELOPING THE SERVICE AREAS “The Advertising Game.” Not every town retailer believes in de- veloping better relations throughout his service area by advertising. Two clothiers in different towns were especially pronounced in this opinion since they contended that it reflected* upon the farmers’ good judgment if they did not trade at their stores for had they not been in business thirty years and should not the farmers know of the superiority of their merchandise? This policy was not general however and Table XXXVIII shows the advertising activity of the various centers in bid- ding for farmers’ trade in their territory. This is an analysis of the direct mail advertising simply among those families who claimed the town as their trade center and not in the wider area beyond which could well be considered as prospective territory. Obviously from this Service Relations of Town and Country 75 Table XXXVIII. — Families Receiving Direct Mail Advertising from Trade Town and Other Cen • ters Compared Centers from which These Families Received Direct Mail Advertising in Per Cent and in Number Trade towns claiming town as trade town IN PER CENT IN NUMBER Trade towns Nearby com- peting town Large city Small center Trade towns Nearby com- peting town Large city Small center Cambridge 28 3.6 50.0 78.6 1 14 22 Cottage Grove . . 32 6.2 28.1 50.0 2 9 16 Deerfield 37 56.8 64.9 21 24 Stoughton 98 35.7 41.8 53.1 35 41 52 Elkhorn 128 75.8 84.4 25.8 97 126 33 Waupaca 124 92.7 72.2 62.9 5.6 115 92 78 7 Table XXXIX. — The Use to Which the Direct Mail Advertising Is Put in the Waupaca Area Total. Pay no attention Read and use Read and don’t use. Not. specified Use made of the advertising Comparisons in Per Cent and in Number In per cent 100 In number 268 43.3 20.9 11.2 24.6 116 56 30 66 table the towns in Dane County at least still have opportunity for ad- vancement in this line. It is evident that many of the families are re- ceiving advertising material from several other centers and that the- nearby competing towns and the larger cities are actively working for trade in the service areas of the local trade towns. Waupaca has the best record with nearly 93 per cent of those designating it as their trade town receiving the advertising of its merchants. A special effort was made to study the effectiveness of their rather intensive advertis- ing in this territory and Table XXXIX summarizes the results. When those “paying no attention” to the advertising are added to those who claim to receive no benefit even though reading it, the percentage of seeming ineffectiveness goes well over the 50 per cent mark. A special study would be required to answer the question of reasons for this ap- parent failure. The Most Popular Store. Another factor in developing greater business activity in the service community is the regard which the customer or the potential customer has for the service institution its- self, the store. Reasons for considering any store, regardless of the line carried, as the “best store” in town are shown in Table XL. When all areas are taken together “price”, “service” and “have everything” bulk about equally in the minds of the 50 per cent of the people who had opinions on the subject at all. The item of “having everything”, of 76 Wisconsin Research Bulletin 58 Table XL. — Reasons fob Considering a Particular Store as “The Best Store” in Town “The Best Stores” Distributed bt the County Areas in Per Cent and in Number Reasons for store being L . the “best store” All areas Dane county Wal- worth county Wau- paca county All areas Dane county Wal- worth -county Wau- paca county H Total 100 100 100 100 787 307 212 268 Better price 12.1 6.8 12.7 17.7 %y 2 21 27 4734 Best service 11.8 ‘ 16.8 7.1 9.9 93 51 34 15 2634 Have everything 10.2 20.2 1.2 6.0 so 34 62 234 16 Best goods 8.3 12.2 1.2 9.3 65 37K 234 25 Always traded there 4.6 6.2 .2 6.3 3634 19 34 17 Friend or relative 3.2 5.4 1.7 1.9 25 16 34 334 5 Unspecified 49.8 32.4 75.9 48.9 39134 9934 161 131 course, refers to the general or department store where purchases of almost every description can be made within the one building and with the aid of a single clerk. In Dane County this factor appears to be most important. Patronizing the “Home Town”. Many campaigns have been in evidence recently where the appeal has been made to the motive of loyalty to the local or home town. When the question is raised as to which town is the home town for this or that farmer, the problem is not so simple as it may appear. To appeal to a farmer to be loyal to his home town in his furniture trading for example when the town does hot possess a furniture store is certainly rather a tar cry. Two con- clusions seem evident from a study of the trading, social and church going habit of the 817 families, 387 from town and 430 from the country, of high school children in Dane County exclusive of those Table XLL— Trade Relations of Families of All High School Children in Dane County Kinds of Merchandise and Where Families Live in Per Cent and in Number Place of buying in per cent IN NUMBER Groceries Furniture Clothing Groc series Furn iture Clotl ling In town On farm In town On farm a £ o el On farm In town On farm In town On farm In town On farm Total 100 100 100 100 100 100 387 430 387 430 387 430 High school town 87.1 70.7 70.0 54.6 57.1 38.4 337 304 271 235 221 165 Other towns 8.0 17.7 18.9 31.4 28.7 43.7 31 76 73 135 111 188 Divided between school town and other towns 3.9 10.9 5.9 8.6 12.7 15.3 15 47 23 37 49 66 Not specified 1.0 .7 5.2 5.4 1.5 2.6 4 3 20 23 6 11 Table XLII. — Location Preference as Between Town and Country of Certain Social Institutions Service Relations of Town and Country 77 78 Wisconsin Research Bulletin 58 sending children to the Stoughton school. Table XXX already re- ferred to above shows the distribution of the church and social activity and Table XLI, the trading relation of the town and country families with their home town. The first conclusion is that loyalty campaigns are needed for others than merely the farmers of the community, and second, that there is a varying degree to which the town can hope to command trade and social patronage depending upon the type of the service. In other terms, it would seem that the town must specialize with reference to the kinds of service which it can perform most efficiently and most acceptably and not expect to furnish everything to everybody all the time. “Groceries,” “furniture”, and “clothing”, is the order of most frequent trading in the local town and the farm families are not far behind those of the village in the percentages of their purchases and activities in this town. Wives of the town merchants and professional men are frequently seen in the stores of urban center buying special articles. Wives of city merchants and prolessional men are frequently seen on the trains bound for metro- politan centers. After much the same fashion the farmer’s wife dis- tributes her “loyalties” by this principle of specialized service. If the Farmer Had His Way. Opinion of course, is subject to change but if the services of a center are to be developed it is often wise to know the disposition and wishes of the people to be served. In answer to the question of whether they preferred their social in- stitutions located in town or country the 787 families reported to the field workers as follows : Church, town 46 per cent, country 28 per cent, consolidated school, town 19 per cent, country 52 per cent ; high school, town 59 per cent, country 18 per cent ; social club, town 30 per cent, country 29 per cent; store, town 68 per cent, country 14 per cent. The 100 per cent in each case is made up in “no preference” or “no opinion” classification as are detailed in Table XLII. When asked as to whether they preferred to have these institutions located in the same town or distributed among several towns, the answers were almost unanimoush in favor of centralization at the same center. Still another question was asked in regard to how far people would be willing to go to such a town center provided the roads were normally good. A frequency curve for the answers shows the mode falling at 6 miles for Waupaca, and Dane Counties while in Walworth County it rises to 9 miles. There was an exceedingly striking correlation between the higher per- centages of town preference and the ages of those answering the question The younger people with few exceptions voted for the town location. The inferences from these expressions should be plain — an expanding opportunity for service faces the local town. CONTENTS Page Part I. — The Service Organization of Town and Country — A Brief of Findings and Principles 1 Town and country interdependencies 3 The town meets an opportunity 4 Business and living emphasis for agriculture 9 Part II. — The Town has Its Farmers — How Important is the Farmer in the Life of the Town 15 The town, an aggregation of service agencies 15 The commercial agencies 22 The non-commercial agencies 31 The town as a town and service station 34 Part III. — The Farmer has his Town — What Kind of a Town Does the Farmer Really Want? 39 The economic service 43 The communication service 58 The educational service 65 The religious service 65 The social service 66 The organization service 68 Competing service centers 70 Developing the service areas 74 3 0*7 Research Bulletin 59 May, 1924 Anthracnose of Cane Fruits and Its Control on Black Raspberries in Wisconsin LEON K. JONES Agricultural Experiment Station of the University of Wisconsin Madison Contents Page The disease • 1 Description 1 Economic importance 2 The causal organism 4 Cultural studies ... 1 4 Germination studies 5 Pathogenicity 6 Life history 7 Seasonal development 7 Production of spores 7 Source of inoculum in nature 8 Control Measures 9 Sanitation : 9 Summary 24 Literature cited 25 Anthracnose of Cane Fruits and Its Control on Black Raspberries in Wisconsin 1 A SURVEY of the cane fruit industry of Wisconsin by the writer (1920) in 1919 showed that anthracnose was one of the chief factors responsible for the drastic decline of the black raspberry acreage of the state in the last decade. Consequently, it was deemed advisable to make a careful study of this disease, with special reference to control measures. Field and laboratory experiments were conducted at Madison, Wisconsin, in the period of 1920-23. The results of the writer’s studies as they relate to history and geo- graphic distribution of the disease, pathological histology, taxonomy, and morphology confirm those reported by Burkholder (1917). Burk- holder’s account of these subjects is so satisfactory that it appears un- necessary to treat them in this publication. THE DISEASE In the United States the disease caused by the fungus Plectodiscella vencta Burk, ( Gleosporium venetum Speg.) appears to be widespread throughout the north and also in hilly southern regions, coinciding with the ranges of its hosts. It is generally distributed on the following hosts in Wisconsin as shown by the writer (1920) : red raspberry ( Rubus idaeus var. aculeatissimus (Mey.) Regel and Tiling), black raspberry ( Rubus occidcntalis L.), purple-cane raspberry ( Rubus neglectus Peck), and black- berry ( Rubus sp.). The relationships of the pathogens causing anthracnose on the above named hosts have not been definitely determined by cross inoculations, except as reported by Burkholder (1917) that the organism isolated from purple-cane raspberry through inoculations pro- duced infection on black and red raspberry. Description The symptoms manifested on the various hosts of P. veneta are some- what similar, although they vary in the color, shape and size of the lesions produced, depending on the host and the severity of the attack. Canes, leaves, petioles, peduncles, pedicels and fruits may be attacked, although the symptoms on the canes are usually the most noticeable. Descriptions here given refer to symptoms on Cumberland black raspberry unless otherwise noted. On canes. Elliptical to circular greenish-brown lesions one-half to one millimeter in diameter usually appear on the young shoots in early spring when the latter are eight to ten inches high. These lesions are slightly sunken, with a stromatic development of the fungus in the center, which is somewhat darker and raised. Under binoculars the 1 The writer wishes to express his indebtedness to Dr. G. W. Keitt, of the University of Wisconsin, under whose direction the work was performed. 2 Research Bulletin 59 affected tissue has a slightly water-soaked appearance. The lesions enlarge slowly and the centers become a pale buff to white, while the advancing margin is raised and reddish-brown to purple in color. Mature lesions (Plate I) are circular to oval and seldom become more than one centimeter in diameter, although they often become confluent, making large irregular patches that may encircle the cane. In cases of severe attack, the canes may crack longitudinally (Plate I, B). These cracks, usually small, may split the cane to the pith for a distance of two to three inches. The lateral branches often become seriously infected. The resultant lesions are similar to those on the canes, but smaller, and often cause the death of the young branches. On leaves. The first spotting of leaves in the spring appears about the same time as that on the canes, although not so abundantly. The lesions first appear as yellow or straw-colored oval to irregular spots one-half to one millimeter in diameter. The center of the lesion is raised and brownish, while under binoculars the veins of the leaf at the outer edge of the lesion are slightly purple. The mature lesions are one to two millimeters in diameter with light colored centers and purple margins. These spots may drop out, giving the leaf a ragged or “shot-hole” appearance. In cases of severe infection of red raspberries, the leaves may turn yellow and drop. On the leaves the symptoms of this disease are often confused with those of common leaf spot ( Mycosphaerella rubi Roark). The Mycosphaerella leaf spot lesions differ from those of anthracnose in being irregular in outline and somewhat larger, with minute black pycnidia usually present. On petioles, peduncles and pedicels. Anthacnose has been found commonly on these plant parts. The lesions produced are similar to those on the canes, although smaller and often without the purple margin. On the peduncles and pedicels especially, they may coalesce into white scab-like patches (Plate II) that cause these parts to become brittle. These white patches often retard growth on one side of the attacked part, causing it to curl and crack. On fruit. Infection occurs less frequently on the fruit than on the other parts of the host. On one occasion the writer observed fruit in- fection on the variety Plum Farmer in Wisconsin. One or several drupelets become brown and sunken. Frequently the whole fruit be- comes brown, dry and woody, while the healthy berries are still green. Economic Importance Anthracnose is one of the most serious diseases of black raspberry and blackberry, although it seldom causes much injury to red rasp- berry. Burrill (1882) cites an instance of a plantation that had yielded a profit of $400.00 per acre, on which one attack of this disease re- duced the proceeds so that expenses were not met. Scribner (1888) estimates the losses in southern Missouri due to anthracnose on black raspberries at 10 to 12 per cent. Burkholder (1917) states that in certain localities in New York state growers have been obliged to discontinue berry growing due to anthracnose, and that it is evident Anthracnose of Cane Fruits 3 that ahthracnose is correlated with reduction of yields. Anderson (1920) states that “Anthracnose has entirely eliminated the growing of raspberries in some sections of Illinois, and many growers are com- pelled to renew their patches after two /ears of bearing.” He also estimates that in Illinois in 1908 the loss from anthracnose was 50 per cent of the crop, and that 25 per cent of the berry crop is lost there annually because of this disease. In a survey of Wisconsin the writer (1920) found that anthyacnose was one of the most important diseases of cane fruits, and was found wherever raspberries were grown, al- though it was of very little importance on blackberries and purple-cane raspberries. Red raspberries usually show a light spotting of the canes but the writer has not noted important anthracnose injury on red varieties in Wisconsin except in the vicinity of Eau Claire. In this district in 1919 there was considerable spotting of the leaves, which caused yellowing and dropping of the foliage. The disease is most important in the state on black raspberries and, in association with crown gall injury, it is a limiting factor in the black raspberry industry. The disease affects the canes and leaves in the first season of growth, thereby weakening the plant and causing a decrease in fruit yield the ensuing year. The diseased canes are also more subject to winter injury than healthy ones. A very important injury in Wisconsin is caused by the lesions on peduncles and pedicels. Abundant infection on these host parts causes the fruit to be small or to dry up before maturity (Plate II). In order to obtain data on the decrease in fruit yield due to anthrac- nose the writer made counts and weighed the .fruit harvested from sprayed and unsprayed plots of Cumberland raspberries in 1921. The sprayed plot, consisting of 24 plants, had received two applications of lime-sulfur with gelatin as a spreader, as outlined on later pages. The disease had been very satisfactorily controlled on this plot the previous season, while the unsprayed plot, consisting of 12 plants in the same planting, had never received any spray treatment and the plants were abundantly infected with the disease. A summary of the data obtained is presented in Table I. Table I. — Comparison of Fruit Yields of Sprayed and Unsprayed Cumber- land Raspberry Plants, H. Fischer Planting, Madison, Wis., 1921 Plot No. Treatment 11 Average number of beriies Average weight of berries per plant Per plant Per pint No. No. Pound ' 1 Unsprayed 143 227 0.42 3 L-S. -{-gelatin, 1, 2 217 239 0.62 a L — S=liquid lime-sulfur (1) delayed-dormant spray, 1-10, (2) second ap- plication of spray about one week before blooming, 1-40. One-half pound of gelatin was added to each 100 gallons of spray. These data show that the sprayed plants on which the disease had been controlled satisfactorily for two consecutive years produced about 4 Research Bulletin 59 32 per cent more fruit by weight than abundantly infected plants that had never received any treatment for the control of the disease. The loss caused by the disease was very noticeable during the season of 1921, due to the fact that the disease was very severe and was followed by a long dry period prior to and during harvest. The average number of berries per pint was slightly higher on the sprayed plants than on the unsprayed since the smaller berries of the latter part of the season on the unsprayed plants dried up and were not harvested, while all the small ones ripened on the sprayed plants. For a comparison of the control of the disease obtained with various treatments during the season of 1921 see Table VI. Photographs taken at the time of harvest in 1921 show the condition of fruiting branches which had been sprayed as compared with that of unsprayed fruiting branches seriously injured by anthracnose (Plate'VIII). THE CAUSAL ORGANISM Cultural Studies Considerable difficulty has been experienced by investigators in ob- taining pure cultures of the anthracnose organism. Stoneman (1898) states that the fungus “....does not adapt itself readily to artificial cul- ture.” The growth of the organism is so slow that contamination is likely to occur, but it is possible to obtain pure cultures by placing frag- ments of diseased tissue in poured agar plates. The exterior of the tissue from which the isolation is to be made is best sterilized by being dipped into 100 per cent alcohol and flamed. Fragments of tissue may then be removed from below the surface with a sterilized scalpel. On a 15 per cent dextrose-potato agar a straw-colored growth may be detected with a hand lens at the side of the fragment of tissues in five to seven days and may be transferred to an agar slant. The easiest method of isolation, however, is to place on the inside of the lid of a Petri dish fragments of a cane lesion bearing ascocarps. At Madison, Wisconsin, it has been possible to obtain mature ascocarps in the field from early March through June. If the cane tissue is moistened with water the spores are shot onto the agar in the lower part of the Petri dish. Germination of single ascospores may be watched and the resultant growth transferred to an agar slant in a test tube by the method described by Keitt (1915). The growth after 14 days on a 15 per cent dextrose-potato agar is light russet-vinaceous to maroon, with a reddish discoloration of the medium. There is very little aerial mycelium produced in the young cultures. The colonies are formed by a piling up of cells that have a glistening appearance. As the cultures become older, however, fine aerial hyphae are formed over the compact growth. In cultures that are a month old these hyphae give a white, downy appearance to the maroon mass of cells underneath. Conidia are seldom produced in culture. However, a sudden increase in humidity usually stimulates their production. The writer has been able to produce conidia by transferring culture fragments from dextrose-potato agar to the side of sterile sweet Anthracnose of Cane Fruits clover stems in tubes, the stems resting on a small amount of absorbent cotton and in an abundance of water. After three days this culture may be removed to a tube of sterile distilled water, in which the conidia drop off readily. In order to obtain an abundance of conidia for inoculation work it was found ad- visable to pour a spore suspension on poured agar plates. After ten days large pieces of agar bearing the fungus may be transferred to a sterile glass slide in a sterile moist chamber, consisting of a Petri dish lined with moistened filter paper. After three days the cultures may be removed to sterile distilled water and the conidia shaken from the fungal growth. lucida drawing of germinating coni- dia of P. veneta after 16 hours in sterile distilled water at 24° C. A study has been made of the relation of temperature to the growth of the organism on dextrose-potato agar. The most rapid growth was obtained at 22° to 26° C. while no growth occurred at 10° or at 32° C. Plate III, A shows the growth that was made in seven days at constant temperatures ranging from 11° to 32° C. The platings were made from a suspension of conidia in water, one loop of the sus- pension being removed to the center of each Petri dish, into which a 5 per cent dextrose-potato agar had been poured. Germination Studies Conidia germinate readily in sterile distilled water or on dextrose- potato agar or “water agar” (2 per cent agar in water). In sterile distilled water on slides in moist chambers at 24° C. the conidia be- come twice their original size in 12 to 24 hours and a few may become one-septate, or produce short germ tubes (Fig. 1). During the next 24 hours elongation occurs and three or four septa and possibly a small amount of branching may be observed. Conidia are budded off from any of the cells, most abundantly, however, from those at the ends (Fig. 2). Further growth takes place with profuse branching and piling up of cells, forming a stromatic mass about 50 microns in diameter after 96 hours. From this mass of cells filaments radiate for 25 to 50 microns. There is seldom any further growth in sterile water. The germination on dextrose-potato agar is somewhat similar ex- cept that true germ tubes and conidia are pro duced after 44 hours seldom produced and that there is a greater i n sterile distilled water tendency towards the massing of cells. After at 24 C. and the pro- 96 hours at 24° C. the colonies on this medium duction of secondary have an average diameter of about 200 microns, conidia. drawing of germinated conidia of P. veneta showing the growth 6 Research Bulletin 59 Experiments on conidial germination have been conducted at cpn- trolled temperatures in sterile distilled water and on dextrose-potato agar and “water agar.” Six series have been run on dextrose-potato agar, five series in sterile distilled water, and two series on “water agar,” at the following temperatures: 4, 8, 11, 15, 17, 19, 22, 24, 26, 30, 32, and 34° C. No germination has been observed on any medium at temperatures below 11° C. and only slow germination with slight growth at 15° C. The optimum for germination and growth was found to lie between 22° and 26° C. Germination takes place readily at 30° while no germination has been observed at either 32° or 34° C. The ascospores germinate readily in sterile distilled water and on dextrose-potato agar. In sterile water the spore becomes slightly swollen and from five to seven conidia are usually budded off within 16 hours. These conidia have not been observed to germinate in sterile water. On dextrose-potato agar the five to seven conidia are budded off and produce germ tubes 20 to 30 microns long within 24 hours (Fig. 3). Within 72 hours the germ tubes become branched and pro- duce masses of cells, making a colony about 50 microns in diameter with numerous strands of branching mycelium growing for a distance of about 35 microns from the outer edge of the stromatic mass. Modes of germination of the spores of this organism are very variable, depending on such factors as temperature and media. The writer projects making this problem the subject of a future publication. Pathogenicity Lawrence (1910) inoculated fruit of blackberries with conidia from leaves and canes and reports obtaining typical lesions after an incubation period of 15 to 48 hours. Burkholder (1917) inoculated young shoots with a water suspension of conidia from lesions on canes and from pure cultures. He obtained infection in 18 out of 56 inoculation trials, with an inoculation period of three to seven days. experiments on Cumberland raspberry plants The apical foot of each young cane was placed in a bag made of partially water-proofed, translucent “glassine” paper for seven days immediately preceding the inoculation, in order to preclude possibility of natural infection. At the time of inoculation the bags were removed and the young canes atomized with sterile distilled water (controls) or a water suspension of conidia from pure cultures of a single ascospore isolation from a black raspberry cane lesion. The drawing of germinating ascospores of P. veneta showing production and germination of secondary conidia. The writer made inoculation at Sturgeon Bay, Wisconsin. Anthracnose of Cane. Fruits 7 bags were replaced and the canes were kept moist by hanging inside of the bags Erlenmeyer flasks of water from which cheese cloth wicks were wound around the canes. The bags and wicks were removed five days after the inoculations were made, at which time no disease was apparent on the young bagged parts. Observations were made every day after the wicks were removed and the number of resultant lesions re- corded. The results of these inoculations (Table II) show an incuba- tion period of six to nine days. Table II. — Results of Inoculations made with P. veneta on Cumberland Raspberry Canes at Sturgeon Ray, Wis., June 6, 1921 Inocu- lation No. Inoculum Number of lesions o stated date bserved o is n June 12 June 13 June 14 June 15 June 16 1A Control 0 0 0 0 0 IB Spore suspension 0 8 8 8 8 2 Spore suspension 0 0 8 16 16 3 Control 0 0 0 0 0 4 Spore suspension 12 17 17 17 17 5 Spore suspension 0 2 2 2 2 6 Control 0 0 0 0 0 Black and red raspberry plants have been grown in the greenhouse in early spring and attempts to inoculate them have met with little success. Eight series of such inoculations have been made with coni- dial suspensions from cultures obtained through single ascospore iso- lations. Only one of these series gave positive results. Life history studies made during the seasons of 1921 and 1922 indicate that only the young growing canes are susceptible to the disease (Tables V and VII). The plants which were grown in the greenhouse during winter and early spring did not produce a succulent type of growth, which probably accounts for the lack of positive results from inoculation ex- periments with these plants. LIFE HISTORY Seasonal Development of Disease The disease first appears on the young growing canes and leaves in early spring, usually when the canes are eight to ten inches high. At Madison, Wisconsin, the first lesions have been observed on the following dates during the four years of observation: May 20, 1919; May 13, 1920; May 15, 1921; and May 20, 1922. The lesions on the canes, leaves, laterals and fruiting branches continue to increase in number on the young growing tissue throughout early summer. There appears to be little or no increase in disease after the middle of July as is shown in data obtained during the seasons of 1921 and 1922 (Plates V and VI). Production of Spores The immature ascocarps are first observed during the latter part of August. Some of the ascocarps are mature at Madison, Wisconsin, as 8 Research Bulletin 59 early as March 1, as the writer has been able to cause the discharge of mature ascospores from freshly collected cane lesions at this time of the year. The asci and ascospores, however, continue to mature through the spring and early summer. Conidia are produced during the spring on the old cane lesions and abundant production of conidia follows the development of lesions on the new growth during spring and summer. On the fruiting canes the fungus probably dies after the production of conidia in the spring, as the writer has been unable to obtain conidia or make cultures from the lesions on these canes through the summer. Source of Inoculum in Nature The primary sources of natural inoculum are the ascospores, which are ejected forcibly from the asci, and the conidia from the overwintered lesions on the canes. The ascospores continue to be a source of inoculum through the spring and early summer. Burkholder (1917) reports that the ascocarps are very rare and probably do not play an important part in the dissemination of the disease. The writer has, however, observed an abundance of ascocarps (Plate III, B) on black and red raspberries in Wisconsin and considers the ascigerous stage an important factor in the overwintering of the disease and its spread in the spring under Wisconsin climatic conditions. The conidia produced in the lesions (Plate III, C) on the current year’s growth form the source of secondary infection through the spring and summer. Experiments have been conducted at Packwaukee, Wisconsin, in order to obtain more definite information relative to the spread of the dis- ease. Three rows, each 250 feet long, of Cumberland raspberry plants were set out April 15, 1920. The rows were 12 feet apart and the plants were spaced five feet apart in the rows. The planting was one-half mile from any other raspberry planting, on land where grain had been grown for 15 years. A careful survey of the surrounding country showed no wild hosts of the disease within one-half mile of the new planting. These plants were obtained from layered tips that were removed from the vicinity of the old plants one month before the appearance of the disease in the spring, and before the new shoots had appeared above the ground. Care was taken to remove all of the old cane stubs from the new plants, in order to avoid carrying any source of inoculum to the new planting. All of the soil was removed from the young plants by washing, after which they were dipped into mercuric chloride solu- tion, 1-1000, and then rinsed before they were planted. On April 14, 1921, one year after the planting was made, these plants were entirely free from anthracnose lesions. Observations made April 13, 1922, two years after the planting was made, showed an abundance of disease on these plants. Of the 125 plants that were living, 99 were diseased, 13 of them being severely in- fected. The remaining 26 plants, which were not infected with anthrac- nose, were scattered among the diseased ones. The ascigerous stage of the fungus had been found abundantly on the diseased canes in the old planting, one-half mile from the ex- Anthracnose of Cane Fruits 9 perimental planting. Ascospores carried by winds that blew over the old planting toward the new one were probably the source of infection for the new planting. Conidia are chiefly water borne, as emphasized by Burkholder (1917). The writer has endeavored to blow the conidia free from the coni- diophores with air from an aspirator, but with little success. However, the conidia drop off readily when the stromatic mass is placed in water. Consequently, when the fungus mass was atomized with water an abundance of spores was washed off. CONTROL MEASURES Sanitation Most writers have emphasized the importance of keeping the planta- tion free from badly diseased canes. Longyear (1904), Jackson (1913), Cook (1918) and Swartwout (1921) recommend cutting out all old canes and the most severely diseased new ones soon after harvest. This is a good cultural practice but it is of little value in checking the dis- ease during the current season, since the writer’s observations show that little or no infection takes place after harvest. When thinning out the canes in the spring, it is advisable to prune out the more severely diseased ones, thereby reducing the source of early inoculum. Good cultural practices during the growing season are advisable in order to remove weeds from the rows. Weeds and compact growth of canes interfere with air drainage, and facilitate the collection of mois- ture, which is favorable to the spread of the disease. Six to twelve inches of the old canes are left on black raspberry nursery stock by nurserymen to facilitate handling. The disease is often abundant on these old cane stubs, and is therefore disseminated to the new plantings. Before nursery stock is planted these old canes should be carefully removed. Young plants obtained from the vicinity of old plants in the spring should be removed to the new planting before they are four to six inches high, since infection of the young plants usually occurs soon after they have attained this much growth. There is little possibility that the fungus lives over on fragments of plants on the ground. Observations by Burkholder (1917) and by the writer show that on the old fruiting canes the organism probably dies after the conidia have been produced in the spring. Therefore, it would appear that new plantings wodld not necessarily have to be made on land formerly free from the disease. However, it would be advisable to make plantings on new soil, because of the prevalence of the crown gall organism in soil previously used for cane fruit culture. Spraying Spraying for the control of anthracnose has been recommended by many writers, but most of the numerous attempts to control the dis- ease in this manner have given questionable or conflicting results. 10 Research Bulletin 59 Burkholder (1917) reviews the earlier literature on spraying and re- ports that a dormant application of lime-sulfur, 1-8, proved to be of no benefit in the control of raspberry anthracnose. After considerable experimental work he states that : “More data relating to the effect of diseased canes on the yield of fruit are needed, and until they are obtained no conclusive proofs can be furnished that spraying to combat the anthracnose of raspberry is a profitable practice.” Dutton (1918) reports control of the disease from three applications of lime-sulfur before the blooming period, and further reports that in 1915 one dormant spray of lime-sulfur, 1-20, gave good control. There is considerable controversy as to the possibility of spray injury from the use of lime-sulfur and Bordeaux mixture. Most writers agree that raspberry plants are very susceptible to spray injury. With- out doubt, a considerable portion of the difference in the amount of injury reported as occurring in spraying experiments has been due to the fact that in their reports most workers have not differentiated between red raspberries, black raspberries, purple-cane raspberries and blackberries. There is certainly a difference in susceptibility to spray injury among these different kinds of cane fruits. Goff (1891) experimented with ammoniacal copper carbonate, and with mixtures of ammoniacal copper carbonate and copper sulfate. These had an injurious effect on the foliage of Cuthbert, Tyler and Gregg varieties, and Bordeaux mixture, 4-6-50, caused great injury to the foliage. He concludes that the foliage of the raspberry is very delicate and can not endure applications of a corrosive nature, and that the foliage of the blackberry, though more resistant than that of the rasp- berry, is more susceptible to injury than that of the apple.* On black raspberries the foliage of old canes and fruiting branches is more susceptible to injury than that of young shoots, and injury is likely to occur in case either Bordeaux mixture or lime-sulfur is applied to the plants in hot, dry weather. From observations made in Wisconsin, foliage injury is to be expected if lime-sulfur or Bordeaux mixture is ap- plied to the plants after blooming. The writer has not observed injury to black raspberry plants from summer strength of Bordeaux mixture or lime-sulfur applications before the blooming period of the plants. The dormant strength of these sprays, applied to the plants after the leaf buds on the old canes had opened in the early spring and only a few leaves had unfolded (Plate IV, A), occasioned no material injury to the plants in the experiments conducted by the writer. In view of the conflicting evidence that has been presented regarding the effectiveness of spraying it was deemed advisable to carry on com- prehensive spraying trials. Preliminary reports of the results of these investigations have been made by the writer (1922 and 1923). A sum- mary of experimental treatments during the seasons of 1920, 1921 and 1922 appears in Table III. Anthracnose of Cane Fruits 11 Table III. — Summary of the Treatment of Experimental Plots of Cumberland Rasp- berry for the Control of Anthracnose, H. Fischer Planting, Madison, Wis. Plot No. No. plants treated Treati nent of plots in stated years a 1920 1921 1922 1 12 Unsprayed Unsprayed Unsprayed 2 24 L-S. + glue, 1, 2, L-S. + glue, 1, 2 L-S. + glue, 1, 2 3 24 L-S.+gelatin, 1, 2 L-S. + gelatin, 1, 2 L-S. + gelatin, 1, 2 4 24 L-S.-j- gelatin, 1 L-S. + gelatin, 1 L-S.+gelatin, 1 5 18 L-S. -j-glue, 1 L-S.+glue, 1 L-S.+glue, 1 5A 6 L-S. -j-glue, 1 L-S. + glue, 1 Unsprayed 6 24 I^S., 1 L-S., 1 L-S., 1 6A 24 L-S., 1 L-S., 1, 2 L-S., 1, 2 7 24 B.M.-fcal-cas., 1 B.M. + cal-cas., 1 B.M. + cal-cas., 1 8 12 B.M. + cal-cas., 1, 2 B.M. + cal-cas., 1, 2 B.M. + cal-cas., 1, 2 9 12 B.M. -(-milk, 1 B.M. + milk, 1 B.M. + milk, 1 9A 12 B.M.-|-milk, 1 B.M. + milk, 1, 2 B.M. + milk, 1, 2 10 12 B.M. -(-gelatin, 1 B.M. + gelatin, 1 B.M.+gelatin, 1 11 12 B.M. -(-gelatin, 1, 2 B.M.+gelatin, 1, 2 B.M. + gelatin, 1, 2 12 24 B.M. -j-glue, 1 B.M. -j-glue, 1 B.M. + glue, 1 13 24 B.M. + glue, 1, 2 B.M.+glue, 1, 2 B.M.+glue, 1, 2 14 12 B.M., 1 B.M., 1 B.M., 1 14A 12 B.M., 1 B.M., 1, 2 B.M., 1, 2 15 20 I^S., 1, 2 L-S., 1, 2 15A 20 L-S., 1 L-S., 1 16 24 L-S. + glue, 1, 2 L-S.+glue, 1, 2 16A 24 L^S. -j-glue, 1 L-S. + glue, 1 17 18 L-S.+gelatin, 1, 2 L-S. + gelatin, 1, 2 17A 24 L-S.+gelatin, 1 L-S.+gelatin, 1 17B 6 L-S.+gelatin, 1, 2 Unsprayed 18 24 L-S.+saponin, 1, 2 L-S.+saponin, 1, 2 18A 12 L-S.+saponin, 1 L-S.+saponin, 1 18B 12 L-S.+saponin, 2 L-S.+saponin, 2 19 24 L-S. + gelatin, 2 L-S. + gelatin, 2 20 24 B.M. + gelatin, 2 B.M. + gelatin, 2 21 12 B.M., 1, 2 B.M., 1, 2* 21A 12 B.M., 1 B.M., 1 22 12 B.M. + glue, 1, 2 B.M.+glue, 1, 2 22A 12 B.M. + glue, 1 B.M. + glue, 1 23 12 B.M.+gelatin, 1, 2 B.M. + gelatin, 1, 2 23A 12 B.M.+gelatin, 1 B.M. + gelatin, 1 24 12 B.M. + cal-cas., 1, 2 B.M. + cal-cas., 1, 2 24A 12 B.M. + cal-cas., 1 B.M.+cal-cas., 1 25 8 Scalecide, 1; B.M.+gela- L-S.+gelatin, 1, 2, 3 tin, 2 25A 8 Scalecide, 1 L — S. + gelatin, 1, 2, 3 b 26 8 Scalecide, 1; B.M.+gela- B.M. + cal-cas., 1, 2, 3 tin, 2 27 8 Scalecide, 1; L-S.+gela- B.M.+cal-cas., 1, 2, 3° tin, 2 28 48 Unsprayed Unsprayed 30 20 L-S., 1 31 20 I^S., 1, 2 32 24 L-S.+glue, 1 33 24 L-S.+glue, 1, 2 34 24 L-S. + gelatin, 1 35 24 L-S. + gelatin, 1, 2 36 24 L— S. + saponin, 1 37 24 ] L— S.+saponin, 1, 2 38 24 .. 1 L— S. + cal-cas., 1 39 24 ] L-S. + cal-cas., 1, 2 a L-S. = liquid lime-sulfur. B.M. = Bordeaux mixture. Cal-cas. = calcium caseinate spreader. Spray applications designated as: 1 = delayed-dormant, using lime-sulfur, 1-10, or Bordeaux mixture, 6-6-50; 2 = application about one week before blooming period, using lime-sulfur, 1-40, or Bordeaux mixture, 3-3-50: 3 = application one week after blooming, using lime- sulfur, 1-40, or Bordeaux mixture, 3-3-50, except as noted in footnotes following. For discussion of spreaders see page 12- b Lime-sulfur, 1-80, plus gelatin was used in application 3. 0 Bordeaux mixture, 1 Vi - 1 H-50, plus calcium caseinate was used in application 3. 12 Research Bulletin 59 Preparation of Sprays Lime-sulfur. Commercial liquid lime-sulfur testing 33° Baume was used in the experiments. The required amount of liquid was added to the water to make the strengths outlined in the summary of treatment (Table III). Bordeaux mixture. Pound to gallon “stock solutions” of lime and copper sulfate were prepared. To make the Bordeaux mixture of the 6-6-50 formula, six gallons of the lime “solution” was diluted to 25 gallons, and six gallons of the copper sulfate solution was diluted to 25 gallons after which the two were mixed with agitation. Bordeaux mixtures of other formulae were made in a corresponding manner. Spreaders with Lime- Sulfur Gelatin. One-half pound of white gelatin was used to each 100 gal- lons of spray. The gelatin was placed in solution with a small amount of water aided by slight heating. This solution was added to the diluted spray mixture and agitated. Glue. One pound of finely ground high grade glue was added to each 100 gallons of spray. The glue was placed in solution and added to the diluted spray mixture in the same manner as the gelatin. Saponin. One ounce of soap tree bark was placed in one quart of water and boiled for 15 minutes. The liquid was strained and used at the rate of eight ounces to ten gallons of spray mixture. Calcium caseinate. A proprietary preparation of casein and lime was used at the rate of one pound to each 100 gallons of diluted spray. The powdered material was added to the diluted spray mixture slowly, with agitation. Spreaders with Bordeaux Mixture Gelatin. Added as outlined above. Glue. Added as outlined above. Calcium caseinate. During the season of 1922 the proprietary prepara- tion was used as outlined above. During the seasons of 1920 and 1921 this material was made as follows : 200 grams of powdered casein was mixed thoroughly with 480 grams of hydrated lime. The dry mixture was added to the spray, slowly and with agitation, at the rate of 150 grams to 25 gallons of the diluted spray mixture. Milk. Milk was added to the diluted spray mixture at the rate of two gallons to 100 gallons of the spray, as recommended by Lecomte (1913). Condition of Plots The experimental plots were located in the H. Fischer planting near Madison, Wisconsin. In 1920 four rows of 78 plants each were selected in a four-year-old Cumberland raspberry planting. The plants were four feet apart in the rows and the rows five feet apart. Plots 1 to 14, with the number of plants shown in Table III, were arranged consecutively in Anthracnose of Cane Fruits 13 the four rows. During the seasons of 1921 and 1922 an adjacent plant- ing of Cumberland raspberries was selected for additional plots, the planting being four years old in 1921 (Plate IV, B). The plots were square or rectangular and contained the number of plants shown in Table III. Previous to 1920 no spraying had been done for the control of the disease in the H. Fischer planting, which was heavily infected with anthracnose (Plate I, A). EXPERIMENTS IN 1920 Treatment A summary of treatment appears in Table III, and supplementary data follow. The delayed-dormant spray was applied on April 26 to plots 2, 3, 4 and 6, but a heavy rain washed off most of the spray before it could dry and made it necessary to discontinue the work On April 29, a partly cloudy day, all plots were sprayed, including the ones that had been sprayed on April 26. A “Perfection” hand sprayer was used. Since there was no foliage on the canes at this time it was easy to cover them thoroughly with a low pressure. An average of one-half pint of spray per plant was used. The buds on the old canes were showing from one-quarter to one-half inch of green tissue with no leaves unfolded. The second application of spray was made on May 26, a bright, clear day. A double-action pressure pump with a barrel attachment was used, and a pressure of 150 to 200 pounds was maintained on a single disc nozzle. An average of three-quarters of a pint of spray per plant was used. The plants were grown in the hill system and tied to stakes. The foliage was so dense that it was hard to cover the old canes thoroughly. Buds were forming on the fruiting branches and the new shoots were 12 to 15 inches high. Results Observations made May 4 showed no apparent injury from the dormant strength spray. Observations made May 26 showed very little infection on the plots. Primary infection occurred during the rain of May 10, appear- ing as lesions on May 13, although the infection was very light at this time. In order to obtain comparative data on the effectiveness of the different spray applications, a count of the number of lesions on each of 20 canes and 20 fruiting branches per plot, chosen at random, was made on June 17 and again on July 13. A summary of this count work appears in Table IV. As the plants were heavily pruned in the latter part of July, it was impossible to obtain further data. However, very little infection occurred after the last counts had been made. The results of the counts made on the canes and fruiting branches are discussed in accordance with the objects of the experiments. Table IV.— Results op Spraying Experiments for the Control of Anthracnose on Cumberland Raspberries, H. Fischer Planting Madison, Wisconsin, 1920 a C0 ctf T) T) 0) 03 -u m a o a .2 July 13th On fruiting branches Research Bulletin 59 1 *OOOOCOt>in> 04 On canes by feet Total 159.2 6 2 5.1 9.6 57.8 20 5 57.8 28.3 33 9 31.5 18.5 28.6 16 5 45.8 4 th foot 27.3 1.3 0.9 6.0 19.2 7.4 38.2 15.2 21.4 15.9 11.6 16.7 9.8 21.5 3rd foot 59.7 3.2 3.4 3.2 30.5 8.7 13.1 8.6 9.2 11.7 4.6 7.1 2.5 15.2 2nd foot 38.7 1.0 0.5 0.2 5.1 2.8 3.7 1.9 2.0 2.7 1.1 3.4 1.8 5.7 lst c foot 32.5 0.7 0.3 0.2 3.0 1.6 2.8 2.6 1.3 1.2 1.2 1.4 2.4 3.4 t ° 1 £ c 0) M c3 S > < June 17th On fruiting branches 26.8 0.0 0.0 0.0 0.2 1.9 0.1 0.0 0.5 0.7 0.4 1.0 0.3 1.1 Total 87.8 0.4 0.4 0.7 5.8 5.4 3.4 4.4 3.5 5.2 3.4 11.2 3.2 15 2 On canes by feet 4th foot 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ro 3rd foot 10 0 0 2 0 o' 0 0 0 0 0 0 0 0 0 80 0 0 0 0 0 0 00 9 ' II 2nd foot 31.2 0.0 0.0 0.3 3.5 4.1 1.6 0.8 1.9 2.6 0.6 3.4 0.3 6.2 1st® foot 44.6 0.4 0.4 0.4 2.3 1.0 1.8 3.6 1.6 2.6 2.8 6.6 2.9 8.8 Treatment b Unsprayed L-S.+glue, 1, 2__ _____ __ _ L-S. + gelatin, 1, 2 _ _ _ _. L-S. + gelatin, 1 _ _ L-S. + glue, 1 L-S., 1 B.M.+cal-cas., 1 B.M. + cal-cas., 1, 2 __ _ _ _ _ _ __ B.M.+milk, 1 B.M. + gelatin, 1 B.M. + gelatin, 1, 2_ B.M. + glue, 1 _ _ B.M.+ glue, 1, 2 _ __ __ ___ B.M., 1 Plot No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 a A summary of counts made on twenty canes and twenty fruiting branches per plot, chosen at random. b See Table III for details of treatment. Spraying dates: 1 (delayed-dormantl, April 29; 2, May 26. 0 Basal foot of cane. Plate I A. — Anthracnose lesions on two-year-old canes of Cumberland raspberry from the H. Fischer planting, Madison, Wisconsin, showing the severity of the disease when the control experiments were begun in 1920. B. — Longitudinal cracking of Cumberland canes following severe anthracnose infection. C. — Anthracnose lesions on a Plum Farmer raspberry cane. Plate II Cumberland raspberry fruiting branches from unsprayed plants, July 7, 1921. The abundance of anthracnose lesions on peduncles and pedicels caused a reduced yield of fruit. Plate III A. — Seven-day-old growth of P. vcneta on dextrose-potato agar (x .65). Cultures made from a conidial suspension and held at various constant temperatures (Centigrade). B. — Photomicrograph of a cross section of an ascocarp of P. veneta from a cane lesion, showing one globular ascus with ascospores. C. — Photomicrograph of a cross section of a lesion on a young cane, showing collapsed host tissue and the production of conidia. Plate IV A. — Development of the foliage on raspberry canes in the H. Fischer planting on May 2, 1922. It is recommended that the delayed-dormant spray be applied after a few leaves have unfolded from the buds on the old canes, as shown by the cane marked 1. B. — General view of the H. Fischer Cumberland raspberry planting where the control experiments were carried on, April 19, 1921. APRIL MAY JUNE JULY A correlation of disease and host development with meteorological records and dates of spray applications, raspberry anthrac- nose experiments, H. Fischer planting, Madison, Wis., 1921 (see pagel5). > '<*XiOCOtO'-<* Tj 3 0 ! bfi O y COCOCOCOmt-XCOXt-O-tfCS-tfOSincOin "*C5 t>c-co P*z «, 3.3 <£.13 00 O 05 05 00 whmon x oo tj< o m HWC'^HW m o x m © o HOlTfOCOH (U 6 ® 3 O* 3 m CD S boo 2 "53 S.2 > inint-Ht'O ,o o HNinw 00 ICO ICO — SUts-s r ooocst«cc22£ a) 3 3 32 32 ,2 a o, >.30 3 O l) O (J (J pjbObObObObOaicoai - i-H -—I CO ' 3 3 3 2 2 3 5 s V X^'aj’aj'o'rt’rt'c'c 5 ' 5 bcbObObObOyoBB - ^73 m J] O CD O 1 2 2 2 A summary of counts made on twenty canes and twenty fruiting branches per plot, chosen at random, August 15, 1921. See Table III for details of treatment. Spraying dates: 1 (delayed-dormant), April 19; 2, May 10. Plots 1-14 had received the same treatment in 1920, while plots 15-28 had received no previous treatment for the control of the disease. Anthracnose of Cane Fruits 19 The effectiveness of a delayed-dormant spray followed by a second application about one week before blooming. On plants which had been treated the previous season satisfactory control was obtained from the use of lime-sulfur alone or with gelatin or glue as a spreader, and from the use of Bordeaux mixture with gelatin or calcium caseinate as a spreader. Although no satisfactory control was obtained on the plants that had not been treated the previous season, lime-sulfur with glue, and Bordeaux mixture with calcium caseinate were more effective than the other spray combinations. The effectiveness of a single spray application about one week before blooming. Lime-sulfur with saponin or gelatin as a spreader, and Bor- deaux mixture with gelatin as a spreader showed little effectiveness in controlling the disease when only the one application of spray was made, about one week before the plants were in blossom. Th6 effectiveness of adding spreaders to the above sprays. Added effectiveness was obtained by using spreaders with the sprays during this season, which was one of extremely abundant infection. Greater benefit was obtained from the use of gelatin or glue with lime-sulfur, and from calcium caseinate with Bordeaux mixture than from any other spreader used with either of these sprays. EXPERIMENTS IN 1922 Seasonal Development of Host The seasonal development of the host was followed as in 1921, and the results are shown graphically in Plate VI in relation to the development of other factors important in the control of the disease. 1 On April 18 the buds on the old canes were showing about three- quarters of an inch of green tissue, but no leaves had unfolded. The first leaves were unfolded on April 22 and the new shoots began to appear above the ground May 1. The development of foliage on the old canes on May 2 is shown in Plate IV, A. From a study of Plate VI it will be seen that the canes continued to increase in length until August 1, and that the most rapid growth occurred between May 15 and June 12. The fruiting branches had obtained their maximum length about May 27. Seasonal Development of Disease The seasonal development of the disease was followed on 20 canes and 20 fruiting branches as in 1921. The data are recorded graphically in Plate VI, and supplementary data are to be found in Table VII. From a study of Plate VI it will be seen that the disease first developed in the field on May 20 and that no increase in number of lesions was observed after August 1. The greatest development of disease occurred during the early part of June when the plants were making their most rapid growth. *The climatological data are from the records of the Madison station of the United States Weather Bureau as in 1921 (Climatological data. U. S. Dept. Agr. Weather Bur. Wis. Section 27: 17-32. 1922). 20 Research Bulletin 59 The disease continued to develop through a longer period in 1922 than in 1921, which may be correlated with the fact that the growth of the host plants continued for a longer period in the season of 1922. The fact that the temperature seldom reached 90° Fahrenheit during June and July of 1922 may have had some effect in favoring the longer period of in- fection. As in the previous season the greatest amount of disease de- veloped when the host plants were growing most rapidly. As in 1921 the older portions of the canes developed resistance to the disease while the younger portions were being infected (Table VII), which further in- dicates that the rapidly growing portions of the raspberry plant are the most susceptible to the disease and that resistance to the disease is de- veloped as the growth ceases and the plant tissues harden. Table VII. — Average Increase in Number of Anthracnose Lesions on Canes and Fruiting Branches of Unsprayed Cumberland Raspberry Plants, H. Fischer Planting, Madison, Wis., 1922. Dates observed On canes by feet a On fruiting branches b 1 st° foot 2nd foot 3rd foot 4th foot Total No. No. No. No. No. No. May 17 0.0 0.0 0.0 0.0 0.0 0.0 20 0.3 0.0 0.0 0.0 0.3 0.1 22 1.0 0.0 0.0 0.0 1.0 0.2 26 1 .0 0.0 0.0 0.0 1.0 0.1 29 2.5 0.0 0.0 0.0 2.5 0.8 June 1 5.7 1 .6 0.0 0.0 7.3 2.3 4 2.0 10.3 0.3 0.0 12.6 1 .0 7 0.5 11.3 1.5 0.0 13.3 0.3 10 0.1 3.9 1 .4 0.0 5.4 0.3 12 0.0 10.9 0.7 0.0 11.6 0.0 16 0.0 0.4 0.8 0.3 1.5 0.0 19 0.0 0.9 0.0 0.0 0.9 0.0 22 0.0 0.3 0.3 0.0 0.6 0.0 24 0.0 0.0 0.0 0.5 0.5 0.0 29 0.0 0.0 0.2 0.0 0.2 0.0 July 2 0.0 0.0 0.0 0.3 0.3 0.0 5 0.0 0.0 1.4 0.4 1 .8 0.0 11 0.0 0.0 1 .0 0.0 1 .0 0.0 18 0.0 0.0 2.0 1.8 3.8 0.0 24 0.0 0.0 1.0 3.0 4.0 0.0 Aug. 1 0.0 0.0 0.0 1 .3 1 .3 0.0 11 0.0 0.0 0.0 0.0 0.0 0.0 Oct. 7 0.0 0.0 0.0 0.0 0.0 0.0 a Average increase in number of lesions on twenty canes tagged on May 17. b Average increase in number of lesions on twenty fruiting branches tagged on May 17. c Basal foot of cane. Treatment A summary of treatment appears in Table III, and additional data follow. The delayed-dormant spray was applied on May 2, a cloudy day with a light easterly wind. A wheelbarrow spray outfit was used, and a pressure of 100 to 150 pounds was maintained on a single disc nozzle. An average of one-half pint of spray per plant was used. The stage of de- velopment of the foliage on the old canes at the time when this spray was applied is shown in Plate IV, A. The new shoots were beginning to appear above ground. Anthracnose of Cane Fruits 21 The second application of spray was made on May 17, a cloudy, cool day with a light breeze from the southeast. The wheelbarrow spray out- fit was used, and a pressure of 75 to 100 pounds was maintained on a single disc nozzle. An average of pints of spray per plant was used. The young canes were eight to nine inches high, and the fruiting branches seven to eight inches long with the blossom buds well formed. The third application of spray was made on June 1, at the end of the blooming period of the plants. The wheelbarrow spray outfit was used, and a pressure of 75 to 100 pounds was maintained on a single disc nozzle. An average of \ l / 2 pints of spray per plant was used. The new canes were 22 to 23 inches high. Results Counts were made of the number of lesions on canes and fruiting branches on the various plots, as in 1920 and 1921, a summary of which appears in Table VIII. The results of the counts are discussed in accordance with the objects of the experiments. Unsprayed. The disease was fairly abundant on the unsprayed plants, although not so abundant as in the previous season. Plants which had been sprayed in 1920 and 1921 but left unsprayed in 1922 showed con- siderable decrease in the amount of infection on them as compared with the amount of infection on plants which had been left unsprayed the three seasons (plots 5A and 1). This cumulative benefit from spraying is not evident, however, in a comparison of results from plots 17B (sprayed in 1921, unsprayed in 1922) and 28 (unsprayed the two sea- sons). The effectiveness of lime-sulfur as compared with Bordeaux mixture. In general, lime-sulfur gave slightly better control of the disease than did Bordeaux mixture. The effectiveness of a delayed-dormant spray only. Commercial con- trol of the disease was obtained from the use of lime-sulfur alone or in combination with glue, gelatin or calcium caseinate as a spreader, and from the use of Bordeaux mixture with calcium caseinate as a spreader. Lime-sulfur with saponin as a spreader, and Bordeaux mixture alone or in combination with glue, gelatin or milk failed to control the disease commercially. The effectiveness of a delayed-dormant spray followed by a second ap- plication about one week before blooming. Very satisfactory control of the disease was obtained from the use of each of the spray combina- tions, with little difference in their effectiveness. The effectiveness of a delayed-dormant spray followed by two appli- cations; (A) one week before the blooming period, and (B) at the end of the blooming period. Excellent control of the disease was obtained from the use of lime-sulfur with gelatin, and Bordeaux mixture with calcium caseinate, but extreme injury to the foliage resulted from appli- cation of the sprays after the blooming period. Little or no reduction Table VIII. — Results of Spraying Experiments for the Control of Anthracnose on Cumberland Raspberries, H. Fischer Planting, Madison, Wisconsin, 1922 a 22 Research Bulletin 59 S-g « »•£ c © 6 g-g g > 3 O 3 J; C'M'*-' 0 © u 2 “© a a, S-g 2 fi © £ 3 cs > ~ o © o 6 PhJz; g,© Q) § § > 3 ° 3 tj <3 c Sc © a §1,0 2 S c m c © c c c3 >30© <3 c ’3 o o PhZ So © a gUs ” c » +3 c > 3 O 3 <3 g-35£^ 0) >-i X So © a ™ c 2 c © 5 c s > 3 o w <3 o o OhZ ■, >»— ^©©©©©cflrtrtcece' to eg t- 05 eg eoi-ieocg eo^ouHto <3 <3 05 05^' eoeo eg* eg T H,-iCCC®w©® r ”' T *r^ tc ce ce © .. 221 ^ 111^3 5 Z’©’©’©'e 5 'c 5 ts’rt’c'c cg boMbflbobo©y©wCC * C cWMCQWMCQOQOQMCQGQMOQCqSSSSSSSSSSSSIS to i-i a 3 2 © be £ ~ 3 3 j ~ £ ” ~ l^s rt >> c * i ’g § eg* ta § ... 2 £ N ^ “ >* £ ° I I a *oo >- +3 eg to c © £ to w o o C y "a 2 T3 . £ s. ^ ■ - >» eg 05 .3 © 1-1 •S '3 >> So T) ll cs eg na 05 be 1-1 C c 2 a © « w £ O tii Hi ^ 03 ds CO £ C 5 .2 * I 'l IS « £ C ,3 -5 ~ U 8 3 n to 3 .5 S t5 ©_ c jS T3 “ v 6 ‘3 ® „r © ^ B 3 .► - g © © a 3 -3 y © o u ©.2 © O M T 3 ° a 1-2 >. s ^ 5 ® ^ ’s 6 *° P r O 3 ^ 5 * ” 8 ^ § <3 co Ph © ce js o o< ; o 2 oo 3 3 ^ 3 | ■p 2 11 J PQ Anthracnose of Cane Fruits 23 in the amount of foliage injury was brought about by reducing the strength of the summer sprays by one-half in this third application. The effectiveness of a single spray application about one week before blooming. Lime-sulfur with gelatin or saponin as a spreader, and Bordeaux mixture with gelatin failed to control the disease commer- cially when only the one application of spray was made, about one week before the blooming period of the plants. The effectiveness of adding spreaders to the above sprays. Very little benefit was obtained from the use of spreaders with the sprays during this season, which was one of only moderately abundant in- fection. 24 Research Bulletin 59 SUMMARY Anthracnose, caused by the fungus Plectodiscella veneta Burk., mani- fests itself in purplish to white spotting of the canes, leaves, petioles, peduncles, and pedicels, and in drying up of the fruit. The disease appears to be widespread with its hosts in the United States, and has been reported as common to blackberries and raspberries in Canada. The black raspberry has been observed to be more susceptible to the disease in Wisconsin than any other host. No difference in susceptibility of the different varieties of black raspberry has been ob- served. Anthracnose is one of the most serious diseases of black raspberries and blackberries. It is reported as entirely eliminating the growing of rasp- berries in some sections of the United States, and estimates of the an- nual loss in fruit yield due to this disease in various sections of the United States range from 12 to 63 per cent of the crop. The writer obtained data in 1921 showing a 33.2 per cent decrease in fruit yield caused by this disease on black raspberries. The minimal temperature for growth of the fungus on dextrose-potato agar is about 11° C., the optimal, between 20° and 26°, and the maximal, about 31° C. Conidia are not produced readily in culture, but are obtained abundantly upon the transfer of suitable fragments of cultures fr-om a dry to a very moist atmosphere. Conidia germinate readily in sterile distilled water and on nutrient media, and secondary conidia are often budded off. Ascospores on cultural media germinate usually by the production of five to seven conidia, which in turn produce germ tubes. The period of incubation on the canes has been shown by inoculations and observations to be from three to nine days. The disease first appears on the young growing canes and leaves in the early spring, usually when the canes are eight to ten inches high, which has been between May 13 and May 20 during the last four seasons at Madi- son, Wisconsin. The lesions continue to increase in number on the young growing tissue throughout early summer, and as the plants cease growth during July resistance to .the disease is developed. Ascospores and conidia form the source of natural inoculum in the spring and early summer. Ascospores, which are forcibly ejected from the asci, may be carried by the wind for a distance of at least one-half mile from old plantings and cause infection in new plantings. Good cultural practices during the growing season are advisable. Weeds should be kept in check, as they increase the humidity around the canes. In making new plantings care should be taken to remove the old canes from the young plant roots, thereby eliminating a possible source of in- oculum. Anthracnose of Cane Fruits 25 During the seasons of 1920, 1921 and 1922 anthracnose on black raspberries was satisfactorily controlled by spraying, lime-sulfur giving somewhat better results than Bordeaux mixture. Only one application of spray, about one week before blooming, failed to control the disease in any case. The use of a spray after blooming increased the effectiveness of lime- sulfur and Bordeaux mixture in controlling the disease. Injury to the foliage from this spray application was sufficient, however, to preclude its use. The results indicate that fair control may be obtained by applying only the delayed-dormant spray each year, using lime-sulfur, alone or with glue or gelatin as a spreader, or Bordeaux mixture with calcium caseinate or gelatin as a spreader. The use of spreaders increased the effectiveness of the sprays, especially in seasons of abundant infection. Glue and gelatin gave the best results with lime-sulfur ; gelatin and calcium caseinate, with Bordeaux mixture. It is doubtful, however, whether the use of these spreaders is warranted when careful spraying is done. To control anthracnose on black raspberries under Wisconsin climatic conditions it is recommended that two applications of spray be made each season as follows: (1) after a few leaves have unfolded in the spring (Plate IV, A 1), using lime-sulfur, 1-10; and (2) about one week before the blooming period of the plants, using lime-sulfur, 1-40. LITERATURE CITED Burrill, T. J. 1882 Blackberry and raspbern^ canerust (Anthracnose). Agr. Rev. 2 4 : 89-92. Scribner, F. L. 1888 Anthracnose of the raspberry and blackberry. U. S. Dept. Agr. Rept. 1887 : 357-361. Goff, E. S. 1891 Experiments in the treatment of the Septoria of the rasp- berry and blackberry. Jour. Myc. 7 : 22-23. Stoneman, Bertha 1898 A comparative study of the development of some anthracnoses. Bot. Gaz. 26 : 69-120. Longyear, B. O. 1904 Fungous diseases of fruits in Michigan. Mich. Agr. Exp. Sta. Sp. Bui. 25 : 1-68. Lawrence, W. H. 1910 Anthracnose of the blackberry and raspberry. Wash. Agr. Exp. Sta. Bui. 97 : 1-18. Jackson, H. S. 1913 Diseases of small fruits. Ore. Agr. Exp. Sta. Bien. Crop Pest Rept. 1911-12 : 261-270. 26 Research Bulletin 59 Lecomte, Antoine ' 1913 Contribution a la recherche d’une bonne bouillie mouillante. Rev. Vit. 40 1027 : 225-228. Keitt, G. W. 1915 Simple technique for isolating single-spore strains of certain types of fungi. Phytopath. 5 : 266-269. Burkholder, W. H. 1917 The anthracnose disease of the raspberry and related plants. Cornell Agr. Exp. Sta. Bui. 395 : 155-183. Cook, M. T. 1918 Common diseases of berries. N. J. Agr. Exp. Sta. Cir. 88 : 1 - 12 . Dutton, W. C. 1918 Spraying to control anthracnose on black raspberries. Mich. Agr. Exp. Sta. Sp. Bui. 88 : 1-8. Anmerson, H. W. 1920 Diseases of Illinois fruits. 111. Agr. Exp. Sta. Cir. 241 : 1-155. Jones, L. K. 1920 Diseases and insect injuries of cane fruits in Wisconsin, 1919. Wis. State Dept. Agr. Bui. 33 : 149-157. Swartwout, H. G. 1921 Small fruit growing in Missouri. .. Mo. Agr. Exp. Sta. Bui. 184 : 1-27. Jones. L. K. 1922 A preliminary report on the control of raspberry anthracnose. Phytopath. 12 : 57-58. Jones, L. K., and Vaughan, R. E. 1923 Control anthracnose on black raspberries. Wis. Agr. Exp. Sta. Cir. 159 : 1-4. EXPERIMENT STATION STAFF E. A. Birge, President of the Univer- sity H. L. Russell, Dean and Director F. B. Morrison, Asst. Dir. Exp. Sta- tion W. A. Henry, Emeritus Agriculture S. M. Babcock, Emeritus Agr. Chem- istry A. S. Alexander, Veterinary Science F. A. Aust, Horticulture B. A. Beach, Veterinary Science R. A. Brink, Genetics L. J. Cole, In charge of Genetics May Cowles, Home Economics E. J. Delwiche, Agronomy (Ashland) J. G. Dickson, Plant Pathology Bernice Dodge, Home Economics J. S. Donald, Agricultural Econom- ics F. W. Duffee, Agr. Engineering J. M. Fargo, Animal Husbandry E. H. Farrington, In charge of Dairy Husbandry C. L. Fluke, Economic Entomology E. B. Fred, Agr. Bacteriology W. D. Frost, Agr. Bacteriology J. G. Fuller, Animal Husbandry W. J. Geib, Soils E. M. Gilbert, Plant Pathology L. F. Graber, Agronomy E. J. Graul, Soils F. B. Hadley, In charge of Veterin- ary Science J. G. Halpin, In charge of Poultry Husbandry E. B. Hart, In charge of Agr. Chem- istry E. G. Hastings, In charge of Agr. Bacteriology C. S. Hean, Librarian B. H. Hibbard, In charge of Agr. Economics A. W. Hopkins, Editor, in charge of Agr. Journalism R. S. Hulce, Animal Husbandry G. C. Humphrey, In charge of Animal Husbandry J. A. James, In charge of Agr. Educa- tion J. Johnson, Horticulture E. R. Jones, In charge of Agr. Engi- neering L. R. Jones, In charge of Plant Path- ology G. W. Keitt, Plant Pathology F. Kleinheinz, Animal Husbandry J. H. Kolb, Economics B. D. Leith, Agronomy T. Macklin, Agr. Economics Abby L. Marlatt, In charge of Home Economics P- E. McNall, Agr. Economics J. G. Milward, Horticulture J. G. Moore, In charge of Horticul- ture Moore, -In charge of Agronomy £,• E* Morrison, Animal Husbandry G. B. Mortimer, Agronomy F. L. Musbach. Soils (Marshfield) Helen T. Parsons, Home Economics W. H. Peterson, Agr. Chemistry Griffith Richards, Soils R. H Roberts, Horticulture •y o Sammis, Dairy Husbandry u tt o VAGE> Animal Husbandry H. H. Sommer, Dairy Husbandry J. A. James, Asst. Dean K. L. Hatch, Asst. Dir. Agr. Exten- sion Service H. Steenbock, Agr. Chemistry H. W. Stewart, Soils A. L. Stone, Agronomy W. A Sumner, Agr. Journalism J. Swenehart, Agr. Engineering W. E. Tottingham, Agr. Chemistry E. Truog, Soils R. E. Vaughan, PlantPathology H. F. Wilson In charge of Economic Entomology A. R. Whitson, In charge of Soils A. H. Wright, Agronomy W. H. Wright, Agr. Bacteriology 0. R. Zeasman, Agr. Engineering and Soils A. R. Albert, Soils H. W. Albertz, Agronomy Freda M. Bachmann, Agr. Bacteriology Ann G. Braun, Home Economics Olive Cooper, Home Economics Ellen H. Craighill, Home Econom- ics W. H. Ebling, Assistant to the IDean N. S. Fish, Agr. Engineering W. C. Frazier, Agr. Bacteriology A. A. Granovsky, Economic Ento- mology A. J. Haas, Executive Secretary R. T. Harris, Dairy Tests E. D. Holden, Agronomy C. A. Hoppert, Agr. Chemistry L. K. Jones, Plant Pathology Alcie Kinslow, Home Economics C. Kuehner, Horticulture Clifford Lamp man, Poultry Husbandry Grace Langdon, Agr. Journalism Samuel Lepkovsky, Agr. Chemistry V. G. Mtlum. Economic Entomology Marianna T. Nelson, Agr. Chemistry G. T. Nightingale, Horticulture Eva Schairer, Home Economics S. D. Sims, Animal Husbandry R. B. Streets, Plant Pathology L. C. Thomsen, Dairy Husbandry L. P. Whitehead, Economic Ento- mology M. Wood (Mrs.), Home Economics Geo. Arbuthnot, Agr. Engineering Archie Black, Agr. Chemistry Conrad Elvehjem, Agr. Chemistry Edith Haynes, Agr. Bacterioogy O. N. Johnson, Poutry Husbandry J. H. Jones, Agr. Chemistry Ruth Myrland, Asst, to Director of Home Ecenomics E. G. Schmidt, Agr. Chemistry M. E. Smith, Inst. Administration D. G. Steele, Genetics Henry Stevens, Genetics Frances W. Streets, Plant Path- ology M. N. Walker, Plant Pathology Memoranda Research Bulletin 60 June, 1924 , ^ t « w* Rural Religious Organization A Study of the Origin and Development of Religious Groups J. H. KOLB and C. J. BORNMAN / AGRICULTURAL EXPERIMENT STATION OF THE UNIVERSITY OF WISCONSIN AND UNITED STATES DEPARTMENT OF AGRICULTURE CO-OPERATING MADISON Present Outlines of Religious Organization Complex Nature of Religious Organization; Dane County Repre- sentative; Religious Bodies Enumerated and Major Groupings Outlined; Distribution and Location of Churches and Membership; The Parishes Mapped; Methods of Study Detailed. Social History of the Religious Groups The Streams of Early Settlement The Original Sources of the Groups and Their General Distribution in the County. The Lutheran Stream The English, German, and Norwegian Lutherans Together with Integrating Movements. The Roman Catholic Stream Father Inama’s Colonization Project in Roxbury Township; The European Origins and Recent Rural Life Emphasis. The “Reformed” Stream A brief sketch of 11 bodies represented in the county. National Federation Movements among these “Reformed” Bodies. The Streams Flow Together The groups have been spread out together over the whole county. Tendencies and Problems of Readjustment Organisation T endencies Present Organization of Parishes : Open-Country Parishes and Neigh- borhood Groups Compared : Village Parishes and the Trade Area ; Rela- tion Between Open-Country and Village Parishes; The Church Mem- bership; Size of Congregations; The County's Clergy; Location of the Churches and their Circuit Arrangements. Problems in Readjustment Abandoned Churches; Over-lapping Parishes; Unchurched Territory; The Non-resident Clergy; Nationality and Language Factors. The Future in Rural Religious Organization Strategic Location of Churches ; The Principle of “Sufficient Volume of Business”; Over-churched and Under-churched Territory; Absentee and Migratory Pastors; the Language Question; Need for a State-Wide Consciousness. Rural Religious Organization PRESENT OUTLINES OF RELIGIOUS ORGANIZATION A VAGUE IDEA often prevails that present-day divisions, dupli- cation, overlapping and lack of co-ordination in rural religious organization are caused by denominational rivalry. This is only a partial explanation and, by no means, the main reason. The complex and seemingly conglomerate nature of such modern or- ganization is the result of historical and sociological processes. The religion of people is usually a social heritage. They hold to its various forms not ordinarily as a matter of persuasion but more often by force of inherited custom and tradition. Complex Nature of Religious Organizations. — This study is an at- tempt to shed some light on the present-day complexity of rural re- ligious organization, the term “religious organization” being used as a designaton for the whole network of activity and influence of various religious bodies over a given area. The endeavor is to show, from the point of view of religious organization in this larger and more general sense, how things came to be as the}' are and to indicate and discuss some of the problems that have thus arisen. In a research project of this sort, there is, of course, no thought of discussing any purely re- ligious question or of setting forth any special program for action. The purpose is simply to present in as clear a manner as possible, the facts and conditions found in this Wisconsin county, among the re- ligious groups and to outline the factors which lie back of this present day group life. A system of careful group analysis, it is there- fore believed, will most readil}' accomplish this purpose. The settlers of Dane County, Wisconsin, for example, coming from the eastern states and from the countries and provinces of Europe, brought their religious beliefs and practices with them. These beliefs and practices were as varied and different as those of the several states and countries from which they came. The establishment of churches in this county meant the transplanting of an Eastern or European institution into what was often an entirely different social and economic environment. These organizations are still in the midst of difficulties and problems attendant upon their readjustments to this different and constantly changing environment. The result of all this is a complicated and diversified growth, as the almost indistinguishable lines on the laboratory map, Fig. I, indicate. The roots of this growth must be traced in their ramifications not only to the churches of the early settlers and settlements, but also back to the various sectional camps into which Christianity divided itself C. J. Bornman received the degree of Master of Arts from the University of Wisconsin in June, 1922, having done his major work in the Department of Agricultural Economics. He is now' the superintendent of a rural demon- stration center at Pittman Center (Sevierville, P. O.) Tennessee. 2 Wisconsin Research Bulletin 60 Fig. 1. — Church Parish Map Showing Location of Churches, Parishes and Circuit Lines for All the Groups at the time of and following the Reformation, when theological id rife was bitter and the newly acquired freedom of religious thought oc- casioned the rise of many and diverse churches and sects. Dane County Is Representative — Dane County is situated in the south-central portion of Wisconsin and comprises an area equal to twice the size of the ordinary county. For the purposes of this study the city of Madison is excluded. According to the Federal census of 1920, the area studied had a population of 51,120. This county gives a fairly accurate picture of rural religious organization, not only in Wisconsin but also in the Middle-West, particularly in those parts where the new American element is predominant and in those other sections of the country where settlement took place in groups. Religious Bodies Enumerated — Seventeen different constituent re- ligious bodies of the 216 in the United States are represented in the county. There are 117 different religious organizations, as shown in Table I. This classification follows that used by the Federal Census Bureau. These organizations are served by 80 ministers and have a combined communicant membership of 21.504. 1 The Major Groupings — For purposes of the study, certain definite major groupings, or families of religious bodies, namely, the Lutheran, communicant member is a person on the church roll, usually over 13 years of age, who enjoys all the rights and privileges of the oiganization. Rural Religious Organization 3 Table I. — Church Organizations, Ministers, and Communicant Members by Denominational Groups.* Constituent bodies Number of organi- zations Number pastors Number of mem- bers 117 80 21 ,504 Adventist bodies Seventh-Day Adventist 1 2 1 2 17 Baptist bodies Baptists 200 Seventh-Day Baptists 1 j 180 Congregational Churches 7 5 507 Lutheran bodies United Lutheran 3 2 290 Synodical Conference 4 4 530 Norwegian Lutheran 27 15 8,919 860 Joint Synod of Ohio 7 3 Iowa Synod 6 4 619 Methodist bodies Methodist Episcopal 22 16 1,573 57 Primitive Methodist 1 2 1 Moravian Church (Unitas Fratrum) 2 . 251 Presbyterian Church in U. S. A 6 5 371 Protestant Episcopal 1 1 37 Reformed Church in U. S 4 2 453 Roman Catholic Church 22 16 6,590 Universalists 0 50 *City of Madison is not included. Table II. — Churches by Major Groupings, Constituent Bodies and Location* Location of Churches Groupings Total Lutheran bodies Roman Catholic “Reformed” bodies Constituent bodies Total Total, United Lutheran Synod Conference Norwegian Lutheran Joint Ohio Iowa Synod Total 117 47 3 4 27 7 6 22 Total 48 Seventh Day Adven- tist Baptist Seventh Day Bap- tist Congregational church Methodist Episcopal.. Primitive Methodist.. Moravian Presbyterian Protestant Episcopal Reformed Church in United States Universalist 1 2 1 22 1 2 6 1 4 1 Vil- Ham- Open City lage let country 10 49 19 39 4 17 4 22 3 3 1 4 6 4 13 2 5 3 0 3 1 12 1 8 5 22 12 9 1 1 1 1 1 3 3 2 13 1 6 1 1 1 4 1 1 1 2 2 1 No towns in county of 2,501 to 5,000 population. 4 Wisconsin Research Bulletin 60 the Roman Catholic and the “Reformed” 2 , have been followed. The constituent bodies of a family, however, are not to be regarded as necessarily having any organic or corporate union. Tables II and III show the constituent bodies in each major grouping, or family, as they were found in the county. Distribution and Location of Churches and Membership — A summary of the number of churches in each major grouping is found in Table II. It shows their location, whether in city, town, village, hamlet, or open country, giving also the totals for all the churches and for each major grouping. 3 Table III.— Communicant Members by Major Groupings, Constituent Bodies, and Location* Groupings Constituent bodies Locati ion of communicant members Total City Vil- lage Ham- let Open country Total 21 ,504 3,167 9,662 1 ,449 7,226 Total 11,218 2,489 3,342 499 4,888 United Lutheran 290 290 Lutheran bodies Synod Conference 530 390 140 Norwegian Lutheran.. Joint Ohio 8,919 860 2,489 2,003 450 499 3,928 410 Iowa Synod 619 209 410 Roman Catholic Total 6,590 203 4,283 80 2,024 Total 3,696 475 2,037 870 314 Seventh Day Adven- tist 17 17 “Reformed” Baptist 200 100 100 bodies Seventh Das'- Baptist Congregational 180 180 507 75 297 135 Methodist Episcopal.. Primitive Methodist.. 1 ,573 57 250 1 ,098 50 175 57 Moravian 251 169 82 Presbyterian Church in U. S. A. 371 248 123 Protestant Episcopal Reformed Church in U. S 37 37 453 240 213 U niversalist 50 50 *No towns of 2,501 to 5,000 population in county. Similarly Table III presents a total communicant membership for each major grouping, the total number belonging to city, town, village, hamlet, or open country churches, as well as the totals for each major grouping. This compilation indicates that of a total of 47 Lutheran churches, 4 are located in cities, 17 in villages, 4 in hamlets, and 22 in the open country, having a total membership of 11,218, of which 2,489 belong to city churches, 3,342 to village, 499 to hamlet, and 4,888 to 2 The term “Reformed” is not used in its narrow or sectional sense but in its historic and general sense, designating practically all the non-Lutheran Protestant bodies. 3 The definition of city, town, village and hamlet adopted follows that of the Committee on Social and Religious Surveys of New York City, namely: Hamlet— 0 to 250; village— 251 to 2500; town— 2501 to 5000; city— 5001 and over. Rural Religions Organization 5 Ohio rlAiixviu.i Oh toj PRirmott open country. Of tF 22 Roman Catholic Churches, 1 is located in a city, 12 in villages, 1 in hamlets, and 8 in the open country. These 22 churches have 6,5^ members, distributed as follows : City, 203 ; village, 4,283 ; hantet, 80 ; open country, 2,024. The “Reformed” bodies have 48 churche> in the county. Of these 5 are in the cities, 22 in villages, 12 in lamlets, and 9 in the open country. Their combined membership i 3,696. Of these 475 are found in city churches, 2,037 in villages, 81) in hamlets, 314 in the open country. The Parioes Mapped — In order to measure the area of influence of each religious organization and to indicate this graphically on a map, Figures 2 to 7 inclusive, present the Lutheran, the Roman Catholic and 3awm City ! J Parish 0 § * Boundaries 1 1 Overlapping j Parishes { O O Circuit Lines I O Church _ Pastor's x Residence SJ»rcl Church with . » $ Resident Pastor .' I A ABANDONED •. *. • i ~ :V.;vVV-::::-:: : m ■■'o : " • 9 Church iPtRRY Fig. 2. — Lutheran Churches, Parishes and Circuit Lines of Western Dane County 6 Wisconsin .Research Bulletin 60 NinXl DEFOREST ' S.BRISTOI Sufi Prairfe MARSHALL PUNKIN Hollow Sugar' Bosh' 'Dexrfieu Deerfield! [CAMSRIDGI rARLANO «<-»SHKOnONCj OUGH TOI rw I m Norway ? I ■ t Parish Boundaries mmxmm Overlapping Parishes O O Circuit Lines O Church * Pastor's Residence Church with $ Resident Pastor 0 Abandoned CHURCH Fig. 3. — Lutheran Churches, Parishes and Circuit Lines of Eastern Dane County Rural Religious Organization 7 SAimCrrr Dane Roxbury, MAZO MANIC WaUNAI EE Marti isvillei Ashton cross^ PLAINS MlDDLETOl Pine Bluff VERONi Belleville Brooklyn Parish Boundaries Overlapping Parishes O— -o Circuit Lines O Church v Pastor s Residence Church mth 0 Resident Pastor m Abandoned w C hurch VERMONT -[ Oregon Fig. 4. — Catholic Churches, Parishes and Circuit Lines for Western Dane County 8 Wisconsin Research Bulletin 60 East Bristol marshall Son Prairie l WEST PORT COTTAGE y GRO^E IOregom I 1 ^TOUftl £df*rto* STONEtHURCH Porter Township Parish Boundaries mmmm Overlapping Parishes O O Circuit Lines O Church * Pastor's Residence Church with Resident Pastor M Abandoned 9 Church Fig . 5 . — Catholic Churches , Parishes and Circuit Lines for Eastern Dane County Rural Religious Organization 9 Prairitm \ 5a ut\ » Mounds •Creek” glACK. 1 .vJa fm* Cross Plains VERONi Mt VernoI Paoli ORE&ON 1 •EU.EVU.LE Parish Boundaries Overlapping Parishes o O Circuit lines O Church Pastor's Residence Church with Resident 'AST or RAN DO NED HURCH Fig. 6. — “Reformed” Churches, Parishes and Circuit Lines for Western Dane County 10 Wisconsin Research Bulletin 60 CoLurravs I Waters I Sufi^ Prairie MoRav i iRL'ANO IEQ®BAPr. JQCohc. Oregon! sroutHfo! #U 6 k 0r ooA tun Edqtrt o> Parish \ Boundaries OVERLAPPT Parishes Circuit L/NES& O Church * Pastor's Residence Church with 0 Resident Pastor m abandoned w Church Fig. 7. — “Reformed” Churches, Parishes and Circuit Lines for Eastern Dane County Rural RELiqous Organization 11 the "Reformed” parish maps, lye parish boundaries enclose the area within which practically all th^ members of a given church reside. For the sake of clearness, they Are shown on three different maps ac- cording to the major groupings. The cross-hatching on the maps in- dicates overlapping areas or small Parishes within larger ones. The Methods of Study Employtd — A careful and rather extended personal visitation and study was made right out in the field of every church in the county, excepting thoie within the city limits of Madi- son. Pastors and responsible laymen were interviewed and schedules for each church and its parish wert secured. By use of detailed township plat books showing the locations of all the farms, the parishes were laid out during the course of these interviews and then later mapped in the laboratory. All manner of local histories, periodical publications, and documents, as well as local and general church histories were used in forging out the life stories of the various religious groups represented by congregations within the boundaries of the county. The parish schedule covered the following items of inquiry : Name of church ; location of church ; location of parsonage ; name and address of pastor ; years service in this church ; official name of denomination ; nationality or national extraction of members ; frequency of services ; language of services ; number of communicant members ; and remarks. SOCIAL HISTORY OF THE RELIGIOUS GROUPS “Of what national extraction is yorr church membership?” was one of the questions in the original schedule. With surprising readiness one pastor answered : “My people are Rhinelander Germans.” And then with pride he turned to a la'ge township plat map hanging on the wall of the parish reception joom and indicated the farms upon which his people had settled and grown comparatively well-to-do. He described the extent of his parish which consisted of a rather com- pactly settled group of people, who themselves or whose antecedents had come from the same general locality in Europe. Although throughout the entire county a comparatively small proportion of people could be found who are not American citizens, either native born or naturalized, yet whenever this question of racial history was pressed there were usually similar and read}" answers as “Bohemian”, “Yankee”, “Swiss”, or Scotch”. Generally such parishes were found to consist of a group with a common social history which could be traced, if not to the exact, then to the same general locality in the eastern states, or to a similar section or province in Europe. The Streams of Early Settlement It may seem somewhat artificial and perhaps not always absolutely accurate but for purposes of clearness the various streams of popula- tion according to their religious affiliation will be traced back to their sources or reservoirs in Europe or the eastern states. The answers to the inquiry as to the racial history of the various religious groups Fig. 8. — The Streams from Europe and the Eastern States to Dane County, Wisconsin Rural Religious Organization,' 13 have been summarized in Table IV. This table together with the facts of European and American church history make it possible to con- struct a map, Fig. 8, showing graphically the various streams of popula- tion according to religious affiliation that were to distribute them- selves over Dane County. The Original Sources of the Groups — The correlation between nationality and religious affiliation is shown by Table IV. This correlation agrees with the facts of church history, for it will be re- membered that with the Reformation in the sixteenth century, Western Christianity was separated into two grand divisions, namely, Roman Catholic and Protestant or Evangelical Christianity. This is shown very clearly in Chart I. In studying this chart it should be re- membered that the various Protestant confessional groups while separate, were very often strongly influenced by one another. For ex- ample, while it is possible to trace the Moravians back to the Waldensian movement, of the twelfth century, the strongest impetus was given to their movement by John Huss in 1415, and later at the beginning of the eighteenth century two Lutherans, Zinzendorf and Spangenberg resuscitated this church and largely determined its career to the present day. Likewise, the Baptists have been influenced strongly by Calvinism, the Anglicans by Roman Catholics, Methodists by Moravians and Congregationalists by Unitarians. Protestantism, in turn, at the time of the Reformation itself, parted into two streams near its fountain head, that is, the Lutheran and the “Reformed.” While the cause of the Lutheran and the “Reformed” movements was a common one, they were in reality separate and distinct and each had its own peculiar genius. There are, to be sure, certain marked dif- ferences among the religious bodies comprising the Lutheran family of churches as well as among those of the “Reformed” family. Neverthe- less each family of churches has certain distinguishing characteristics, many of which have been retained to the present day. These are the result not so much, perhaps, of original differences as they are due to the fact that in its sweep each movement conquered different sections of Europe. Church and state being so closely allied, the rulers usually decided the forms of religion for their subjects. Each movement stamped its individuality upon the sections thus conquered. Owing to differences in race, location, temperament, and political fortunes, certain groups in turn within each family were formed each having a different sectional development. Thus it is that there are in Protestant- ism two great families of churches the Lutheran and the “Reformed” and within each of these families numerous smaller bodies. The term “Reformed” is used not in the more modern and narrower sectional sense but in its historic and general sense, a usage which is generally adopted by church historians. 4 The Lutheran comprises those churches which bear the name of Luther, while the “Reformed” generally de- ♦Schaff, Philip, History of the Christian Church, Vol. VII, p. 11. 14 Wisconsin Research Bulletin 60 Years 1900 1600 1700 1600 1500 1400 13 00 1200 1100 1000 900 600 TOO GOO 5 00 400 300 ZOO 100 Roman Catholic Old Catholic Anglican (Protestant Episcopal) Catholic Apostolic ' Methodists Lut heran Su/edenborgian Salvation Army Quaker ISZo 1713 1530 1830 1740. 1870 1670 1789 Congregationatist Christian Centuries i42°. .Anabaptist, Mennonite Baptist Brethren Scientist 1879 | Calvinist Zwinglian Socinian Unitarian 1170 1051 ‘Waldensian, Bohemian,. Moravian -Graeco -Russian Orthodox Oriental Orthodox 550 Syrian, Coptic, Abyssinian 4iy ft r . men>Qn Nestorian 43f CATHOLIC TTY WITT xxrr xm xx YTV TUT XU XL X IX VTTT mi XL X m UL re i Chart I. — The Confessional Division of the Churches A. D. 30 to 1911. notes the non-Lutheran Protestant churches. Other general classifica- tions on the basis of liturgical practices, for example, could no doubt have been made and might have been preferred by some. It is also recognized that objection blight be raised to the classification of all non-Lutheran bodies as “Reformed”. Nevertheless from the point of view of historical origin, which naturally cannot be discussed in de- tail here, there are many good reasons for adopting this classification, even though in certain respects it is somewhat arbitrary. Rural Religious Organization 15 Both map, Fig. 8, and Table IV, show the three main streams according to religious affiliation which overflowed from the various reservoirs of population in Europe and in America. The Lutheran stream was supplied by strong currents from Germany and Norway, both of which countries have been predominently Lutheran since the Reformation period. In Germany the Lutheran population elements usually came from one or the other of the northern states. The Roman Table IV. — Nationality or National Extraction of Church Members Grouping Constituent bodies All Numb er of churches Ger- man Nor- we- gian Old Amer- can Eng- lish Scotch Swiss Irish French Mixed Total 117 32 29 16 8 4 4 4 1 19 Lutheran bodies Total 47 19 27 1 United Lutheran 3 4 27 7 6 3 3 Synod. Con- ference 1 Norwegian Lutheran 27 Joint Ohio 7 6 Iowa Synod Roman Catholics 22 I -oc | 4 1 9 “Reformed” bodies Total 48 5 2 16 8 4 4 9 Seventh Day Adventist .... Baptist 1 2 1 7 22 1 2 6 1 4 1 1 2 Seventh Day Baptist 1 Congrega- tional 4 8 1 2 4 Methodist Episcopal .... Primitive Methodist .... Moravian 3 '2 5 1 2 Presbyterian... Protestant Episcopal .... Reformed in U. S 3 3 1 4 Universalist ... 1 Catholic stream was supplied by tributaries from Germany, usually the Rhine Valley and southern German states, and from Bohemia, France, and Ireland. These countries or states are for the most part Roman Catholic in their religious affiliation. The “Reformed” stream which is the thinnest of these three, was made up of numerous small rivulets from England, Scotland, Switzerland, and the New England States, New York, Pennsylvania, and Ohio. Not all of the population elements in the “Reformed” church groups, however, came to the county with this stream but after reaching here they were deflected or thrown off from the Lutheran stream. These elements are found in such groups as the German Methodist, the Evangelical Association 16 Wisconsin Research Bulletin 60 Table V. — Nativity of Dane County for the Year 1870* State or Country of Birth Population Total 53,096 Native Born Total 33,456 Wisconsin 22,738 4,820 802 932 1 ,061 314 2,789 New York Ohio Pennsylvania Vermont Illinois All others Foreign Born Total 19, *640 British America 684 1 ,631 2,955 465 2 6,276 169 6,601 195 216 17 131 298 England and Wales Ireland Scotland Great Britain Germany France Sweden and Norway Bo hem ia Switzerland Holland Denmark All others ♦Ninth Census ol the United States, 1870, Population and Vital Statistics, p. 376. Rural Religious Organization 17 Table VI. — Nationalities of Rural Dane County (Distribution by Town- ships for Years 1895-1905)* County of birth and years ^Townships France Ger- Great Ireland Scandi- Switzer- ma ny Bril tain na 1 via land 1895 1905 1895 1905 1895 1905 1895 1905 1895 1905 1895 1905 Total 99 57 4037 3028 757 498 706 511 5886 4699 10 344 Albion 1 0 45 36 51 41 12 6 275 223 0 0 Berry 0 0 216 117 5 3 0 0 6 4 0 1 Black'*Earth 0 0 26 29 60 30 18 20 0 52 0 0 Blooming Grove 0 3 127 157 2 8 9 13 86 118 0 9 Blue Mounds 0 0 74 45 13 12 18 11 326 155 0 31 Bristol 3 0 156 125 1 2 4 3 113 93 0 0 Burke 0 0 115 74 11 7 16 8 226 196 0 5 Christiana 0 0 32 18 10 11 1 857 641 0 0 Cottage Grove 0 0 84 57 13 3! 32 15 213 174 0 5 Cross Plains 0 0 189 135 17 4 32 21 9 0 0 2 Dane 1 0 190 116 15 5 0 6 65 8 0 0 Deerfield 0 0 196 154 12 10 9 2 322 154 0 2 Dunkirk 0 0 14 9 29 13 42 22 466 317 0 0 Dunn 0 0 19 17 20 14 11 3 265 240 0 1 Fitchburg 0 0 33 43 20 25 58 39 55 50 0 3 Madison 1 1 116 108 45 44 29 19 42 37 0 6 Mazomanie 1 0 146 122 89 61 33 32 7 4 10 4 Medina 0 0 290 208 31 24 21 17 31 21 0 1 Middleton 0 0 389 290 31 12 i 16 4 0 6 0 0 Montrose 65 36 60 49 36 15 23 16 53 27 0 62 Oregon 7 11 18 18 19 33 55 46 119 130 0 2 Perry 0 0 21 12 0 0 0 0 276 240 0 41 Primrose 3 1 8 6 3 4 9 5 164 115 0 57 Pleasant Springs 0 0 2 8 6 1 0 488 471 0 0 Roxbury 0 0 194 119 1 1 5 3 1 1 0 2 Rutland 0 0 12 2 33 21 13 18 309 294 0 0 Springdale 0 0 37 32 15 9 14 11 168 95 0 54 Springfield 0 0 222 137 3 0 1 1 1 5 0 0 Sun Prairie 1 1 228 163 16 10 36 33 170 45 0 2 Verona 3 0 120 143 50 22 20 30 7 73 0 25 Vermont 0 0 43 9 6 0 38 19 151 100 0 12 Vienna 0 0 80 58 28 9 4 1 245 184 0 3 Westport 12 4 230 147 37 26 109 73 1 110 139 0 14 Windsor 1 0 147 122 17 10 9 8 ! 260 287 0 0 York 0 0 158 143 12 8 8 5 0 0 0 0 1 *Kolb, J. H., Rural Primary Groups, Res. Bu!. 51, Agr. Exp. Sta., University of Wisconsin, page 25. 18 Wisconsin Research Bulletin 60 Table VII. — Religious Bodies in Dane County by Number of Churches, Aggregate Accommodations and Total Value for the Years 1850, 1860 and 1870* Number of Aggregate Accom- Denominations churchesf modations by Total value by by years years years 1850 1860 1870 1850 1860 1870 1850 1860 1870 All denominations .. 14 39 84 1 ,694 12,830 22,250 5.500 $113,900 $163,200 Baptist 3 2 10 189 900 1 ,700 16,000 1 ,300 Congregational 1 1 6 300 150 2,000 850 2,500 Evan. Association ... 9 Lutheran 1 10 6 500 3,350 1 ,500 1 ,500 18,900 Methodist Episcopal 4 7 32 332 2,250 1 ,900 7,000 1 ,000 500 12,300 11,700 Presbyterian 4 7 3 173 Protestant Episcopal 1 4 400 1 ,000 7,000 ! 24,000 27,700 2,000 Rom an Catholic 1 10 14 200 3,580 300 1 ,000 Universalist 1 1 *This table was compiled from the following sources: Seventh c’ensus of the United States (1850) pp. 934-936; Eighth Census of the United States (1860), Mortality and Miscellaneous Statistics, pp. 489-493; Ninth Census of the United States (1870) Population and Social Statistics, pp. 559 'ff. No figures are available for the period ending in 1880. fin 1850 and 1860 the caption “Numcer of Churches” is used; in 1870 the caption Number of Church- Organizations.” and the Moravian. Naturally some other elements, too small to be of striking or general significance, were drawn from one group to the other by intermarriage, convenience of location and by other causes. As a rule, then, it is found that each parish has an old American or an old world social and historical background of its own. General Distribution in the County — The settlement took place in groups, such groups consisting of an aggregation of people who had much in common with respect to nationality, language, customs, re- ligious traditions, and economic and social interests. The Lutheran stream, consisting as it did of Norwegian and German elements, spread pretty well over the entire county. Tables V to VII show the distribution at various periods. The Norwegians first settled in the southeastern part of the county in the vicinity of Lake Koshkonong. Later settlements followed in Primrose, Perry, and Vermont town- ships while still other parts of the county were later settled by them. The German settlers generally divide their allegiance between the Roman Catholics and the Lutherans. With respect to their church relations they grouped ‘ themselves as “German Lutherans” and , “German Catholics”. As the map, Fig. 9, shows, they are found in considerable numbers in all parts of the county. Possibly about one- third of the German settlers were Lutherans. The Roman Catholic stream, composed of German, Irish, Bohemian, and French groups, is represented in a large majority of the townships. There are a good many townships, however particularly where the Norwegian elements are strong, in which few, if any, Roman Catholics are found. The large majority of Germans in the county are Roman Catholics. The Rural Religious Organization 19 Fig. 9. — Distribution of the Streams in Dane County in 1895 20 Wisconsin Research Bulletin 60 other elements arrayed in point of numbers are the following: Irish, Bohemian, and French. The “ Reformed ” stream was the earliest and thinnest. A large proportion of these early settlers, as may be seen in Table V, were natives of Vermont, New York, Pennsylvania, and Ohio, and came from the farms and villages of these states. They were pretty well scattered over the entire county. Many of the local townships were organized by them, being named after the localities from which they came. Roxbury, for instance, now overwhelmingly German, was named by an Easterner whose native town in New York bore the same name. Another example is Utica, named by settlers from the vicinity of Utica, New York. A goodly proporton of the Anglo-Saxon immigration came from England and some from Scotland. The former were particularly strong in Albion, Mazomanie, Black Earth and Verona townships. The Swiss are represented in these nativity tables for 1870, disappear almost entirely in that of 1895 and reap- pear in 1905. These elements are all tributaries of the “Reformed” stream. The various groups within the general church families will now be considered in greater detail. The Lutheran Stream The Lutheran stream, it will be remembered, had its original reservoirs in the northern states of Germany and in the Scandinavian countries. These emigrations were attributable in part to a religious strain between church and state in all these areas although another very important factor in the movement was the quest of economic fortune. The English Lutherans — One of the constituent Lutheran bodies represented in the county is the Lutheran General Council. Locally it is referred to as the English Lutheran Synod of the Northwest, this latter being one of the synods constituting the United Lutheran Church. Its adherents are mainly Engfish-speaking, or Americanized descendants of Lutheran immigrants, as can be noted in Table IV. Churches of this body are found in the towns and villages of the county rather than in the open country. This General Council is one of the oldest of American churches. Henry Melchior Muehlenberg was the pioneer of American Lutheran- ism and began his work among the Germans of Pennsylvania in 1741. The first Lutheran synod was the Synod of Pennsylvania, founded in August, 1748. Active missionary work was carried on by itinerants in what was then the West and Southwest. In due time synods were organized by states and in 1820 these were united into the General Synod. The period of Americanization, or the transition from the use of the German to that of the English language, occasioned much bitterness and animosity, and together with other causes led to serious disruption within this synod. During the Civil War the southern churches broke away and formed what is now known as the Rural Religious Organization, 21 United Synod South. Doctrinal controversies which had 36611 carried on for years led to a second, more serious, rupture in 844 and re- sulted in the formation of the Lutheran General Council 1 1866 by the more conservative elements. The liberal party remaitd i n an d had full control of the General Synod. This latter group i-not represent- ed by churches in the county. As will be indicated l er > these three bodies came together again in 1919 and are now ^nown as the United Lutheran Church in America. The Norwegian Lutherans — The first Norwegiar settlers came to Dane County in 1840. From this time on there a steady influx partly occasioned by religious persecutions in No^Y* Considerable friction existed in that country between the not>ty and the masses of the people. The church was an institution of *e state and belong- ed to the Lutheran family of churches. Dissents were punished for their religious activities, compelled to attend cotnunion services and their children were taken by force and baptize These settlers or- ganized their first church in the county in JfA. It was called the East Koshkonong church and is located in Clhstiana Township. This church has the distinction of being the mot'^r church of the Nor- wegian Lutheran Churches in America. The sne year the West Kosh- konong church was founded in Pleasant Spngs township. In both these townships the population is almost excisively of Norwegian ex- traction. Later there were settlements in Sringdale, Perry and Blue Mounds Townships and the following were tfe more important churches established: East Blue Mounds (1850) ; Springdale (1852); Perry (1852). Following close upon settlement irother parts of the county, churches were organized as follows:' Yo*c (1855); Vermont (1857); Primrose (1870) ; West Blue Mounds (180) ; Mount Horeb (1887). Only the earlier and more important chrches are here mentioned. Names prominent in those early beginning were those of C. L. Clausen, A. C. Preus, H. A. Preus, U. V. Korn, J. A. Ottesen, and P. L. Larsen. C. L. Clausen, by some conslered the pioneer among the Norwegian ministers, was chaplain duing the Civil War of the famous Norwegian regiment, the 15th Yisconsin. The Norwegian Evangelical Lutheran .ynod of North America was founded in 1853, under the leadership o these men, who represented the more conservative of the Norwegiai Lutheran elements. Among Norwegians this body is often called singly “The Synod Church.” In the early days they were in close sympahy with the Missouri Synod, an influential German Lutheran body, a whose seminary they had professors of their own, namely, Larsei, Preus, and F. A. Schmidt. Indeed, in 1872, the Norwegian Synod, ogether with other Lutheran synods, had united with the Missouri.ns in forming the strong Synodical Conference. Soon, however, the predestin^rian contro- versy was at its height within the Syncdical Conference. The two parties within this Conference were the ‘Missourians” and the “anti- Missourians”. Both these parties were represented in the Norwegian 22 Wisconsin Research Bulletin 60 Synod and iiprder to avoid a rupture within itself, the latter synod withdrew fro\ the Synodical Conference in 1884. During this con- troversy Prof^or Asperheim of the Seminary of the Norwegian Evangelical Lt ieran Synod, who was an “anti-Missourian” resigned and left the sV^d. The rupture, however, could not be permanently avoided and oc\ rre d in 1887, giving rise to the United Norwegian Lutheran ChurcK n North America. The United Nokegian Lutheran Church in North America was or- ganized in 1890 an\ w as promoted by the “anti-Missourians” who had left the Norwegiar^nod. The latter attempted to ally with them- selves other Lutherl bodies, namely, the Hauge Synod, the Norwegian Augustana Synod an the Norwegian-Danish Conference. All but the Hauge Synod joinecyhis bod}'. These controversies were projected down into the local togregations as is shown by the following story of the Perry townshiprhurch. A meeting was called on Dec. 29, 1882 at which an outline y the discussion of the doctrinal question of “election” was present^ and a vote taken. Again in July, 1887, this same congregation by vte withdrew from the Norwegian Evangelical Lutheran Synod of Nori America owing to the so-called “Naadevalg Striden” (the predestinlion controversy) later to join the United Norwegian Lutheran Ch^ch in North America. 5 The N orwegian Evangecal Lutheran Hauge Synod, first organized according to various churl historians in 1846, or 1850, gained its ad- herents in Dane county a\er the first Norwegian Lutheran churches' in the southern part were \tell established. Although the churches and circuits had been lately orinized and the pioneers were enjoying the peace of a quiet and orderHchurch life, there appeared in their midst the eloquent lay preacher, \lling Eielson, a disciple of Hauge. The latter had likewise been a la\ preacher and a revivalist and the leader of a popular religious movement in Norway. He was several times imprisoned on account of hij religious activity. Under the influence of Eielson, rival churches, oien in immediate proximity of the older churches, sprung up. This, tqether with the fact that differences had arisen over the much debatec question of predestination, accounts for the phenomenon that right at in the open country two substantial churches with prosperous standing within a stone’s th illustrated in the cases of both well as in Perry Township, throughout the northwest wl of the three described, wa sympathizing with Hauge’s ^ sensions in this synod were :ongregations may today be found )w of one another. This is strikingly East and West Koshkonong churches, as I fact this picture is rather characteristic re Scandinavian settlements are found. This synod, apparently the olest in America but not in the county, organized by Norwegian immigrants ews under Eielson’s leadership. Dis- frequent and after several parties had seceded, reorganization took Iplace in 1876 under the name of “The Norwegian Evangelical Lutheran Hauge Synod.” “Ruste, C. O., Sixty Years of berry C Congregation, pp. 71 and 91. Rural Religious Organization 23 The German Lutherans — The German immigration began very early. The settlers came largely from Ron\an Catholic provinces of Germany and some from Bohemia and Austria. Others were members of the Lutheran church or of the official State Church. The German Lutherans in most places superseded earlier groups, one of which was known as the EvangelicaJ Association. Some of the outstanding churches of this Lutheran group were the following: Berry (1860); Blue Mounds, (1858); Cottage Grove, Deerfield Township; Middleton; Perry; Springdale and Westport. Typical of one of these settlements is t\at of the former Lutheran congregation in the neighborhood known a\ Mud Lake. The settlers came in 1854 and built their church in 1861. One of the early settlers is said to have had a definite colonization plai\ which he put into effect by writing consistently to friends and relatives in a similar neighbor- hood in one of the provinces in Mecklenburg. Why some groups were Lutherans and othen Catholics is illustrat- ed in the case of the Springfield Corners neighbo hood. The Lutheran group has its church at what is known as Lutheran or Hickory Hill. Dr. Otto Kerl and Louis Martini, the latter an intimate friend of Carl Schurz, were leaders among these early settlers ccming from Saxony. The other group is said to have come from Cologne, and partly from Luxemburg. These settlers united . and organizec a Catholic con- gregation and built a church just to the west of Springfield Corners. It was later moved to what is now known as Martinsville. The Evangelical Lutheran Synodical Conference is c^ie of the three German Lutheran bodies found in the county. It w«s organized in 1872. It originally embraced the following synods : Missouri, Ohio. Wisconsin, Minnesota, Illinois, Michigan, and the Noitvegian Synod referred to above. Some of these synods have in the m-antime with- drawn. The churches in this county are affiliated through vthe Synod of Wisconsin with this Synodical Conference. Like othei^ synods the Wisconsin Synod had a corporate existence prior to the organization of the Synodical Conference and before this belonged to tl^e Lutheran General Council, which it left in 1872 to join the Synodical Conference. Owing to the influence which the latter organization has^had upon German Lutheranism in the West, and therefore also upon th churches in this county, a little more detail as to its origin and histdty will be given. \ * 'i The organization of the Synodical Conference was the result of a movement on the part of German Lutherans belonging to the various synods referred to above, to unite and form a stronger an,^. larger body. In order to avoid confusion it must be remembered that (a, con- gregation belonging to a given synod is not always located in the; state, whose name that particular synod bears. For example, the corporate title of one of these is “The Evangelical Lutheran Synod of Missouri, Ohio and other States”. It extends from the Atlantic to the Pacific and from Canada to the Gulf of Mexico and the Argentine. It so 24 Wisconsin Research Bulletin 60 happens that this particular synod has had a most powerful influence upon the Synodical Conference, wfrose chief promoter it was from the very beginning. In the determination of the policies and the doctrinal position of the Synodical Conference the Missouri Synod has always played the leading part. This has been due in large part to the energetic and capable leadership of C. F. W. Walther, the founder of so-called “Missouri Lutheranis/n” and its most conspicuous proponent. For these reasons any one t/elonging to one of the Synodical Con- ference churches, of whatevyf synod, is in popular usage, a “Missouri Lutheran.” It will also be /emembered that the theological questions which agitated the Norwegian Lutherans were those propounded by the so-called “Missourians'/. Most of the Lutherans in the west and the Missouri Synod in particular, trace their origin to the reaction which followed the union of the Lutheran and Reformed Churches in Germany into one new Evangelical Church, popularly known as the State Church. This church was a tax-suiported institution. The movement which con- summated in this unyon was inaugurated by Frederick William III of Prussia, urged by the desire of the Hohenzollerns to reconcile the re- ligious differences /f their subjects. The strict, orthodox Lutherans violently opposed /his union and organized separate Lutheran Free Churches. The rt/ictionary movement spread to Saxony and Bavaria. Efforts to suppre/s this dissent were made but the only result was to stiffen in their ^opposition those who considered themselves the champions of trVe Lutheranism. Partly as a result of such persecution in the Fatherlaid, thousands of Lutherans emigrated to America. In 1839 one such /arty of 500 immigrants streamed into Milwaukee while another consi^ing of 700 Saxons settled in Perry County, Missouri, and in St. Lot/s. Under the influence of C. F. W. Walther the churches which the Sa/ons formed became the nucleus of the Synod of Missouri. No doubt di£ to the experience in Europe of its founders this synod from the beflnning rejected all kinds of unionism with those of another faith. Without question also much of the success of this synod was due to its exctf)tionally able leaders, many of whom were theologically trained in/Serman universities and were former pastors in Germany. The Iom Synod has several congregations in the county as is shown in Table/L It was founded in 1854 at Dubuque, Iowa and like the Missouri Synod it is directly a product of German old Lutheran orthodoxy. The story of its European backgrounds would be largely a repetiti^ of those already told in connection with the Missouri Synod. There is, however, this important difference, namely, that Wilhelm Loehe/the venerable theologian and philanthropist was its sponsor in Germany, and had a remarkable influence upon its policies and activities. Loeha was a strict Lutheran and considered impurity of doctrine as bad ;/s immoral conduct. Upon the request of the German settlers at Fort/ Wayne, Indiana, Loehe sent missionary pastors to America. The.de were trained in Neuendettelsau, Bavaria, the seat of his labors in Rural Religious Organization 25 Germany, which had two stately buildings devoted to the training of missionary pastors. At first these preachers identified themselves with the Ohio Synod, which will be described presently, but later owing to the “unionistic tendencies” of the latter they left and in 1845 establish- ed a theological seminary at Fort Wayne, Indiana, where missionaries, who had received their preliminary training at Neuendettelsau, completed their preparation. In this undertaking Loehe and his„ friends rendered financial support. Loehe advised them to identify themselves with the Saxon Lutherans in Missouri. This was before the Missouri Synod had been organized. Indeed, many of Loehe’s missionaries in 1847 joined with the Saxons in forming the Missouri Synod. The two parties, the “Saxons” and the “Bavarians” could, however, not agree concerning various points of doctrine, particularly those of the Church and the ministry. Representatives were sent to confer with Loehe but in spite of the latter’s efforts at conciliation, no agreement could be reached. Loehe’s adherents, therefore, went west to carry on their missionary work and in 1854 together with several other ministers founded the Synod of Iowa. Since then, although some attempts have been made to harmonize differences, this, synod has maintained a separate corporate existence. The Evangelical Lutheran Joint Synod of Ohio and Other States is the third and last synod with which several of the German Lutheran congregations of the county are affiliated. It was formed in 1818 by itinerant preachers connected with the Pennsylvania Ministerium. The latter later took the initiative in the organization of the General Council. This synod is therefore historically quite closely related to this body. In 1833 it adopted its present name. It had some connection, also, from almost the very beginning with Lutheranism in Germany. The missionary pastors sent by Wilhelm Loehe at first identified them- selves with this synod. This caused it to develop in a more decidedly Lutheran direction than was the case with the older American Lutheran groups. Indeed, in 1872, it had united with the Missouri Synod, said to be the most conservative of Lutheran bodies, in the formation of the Synodical Conference. The controversy on predestination led to its withdrawal in 1881. Recent Lutheran Integrating Movements — Within more recent years the desire on the part of outstanding Lutheran leaders for closer as- sociation has crystallzed in the merging of the more homogeneous bodies into larger organizations. Chart II indicates the general scheme of relationships together with these integrating movements. What the Synodical Conference attained in 1872 by the union of a number of strong German Lutheran synods, making it for man}'- years the strong- est Lutheran body in America, the English or old American Lutherans and the various Norwegian Lutherans accomplished by their respective mergers of more recent date. True to the heritage cherished by them, union among Lutherans is, as a rule, upon the basis of their common doctrines whereas among the “Reformed” bodies as in the case of the 26 Wisconsin Research Bulletin 60 Federal Council of Churches, for example, questions of doctrine are carefully avoided. Their association is rather for purposes of co- operation in practical programs. This accounts for the fact that the only Lutheran body, the United Lutheran Church in America, related to this Federal Council, is associated in a consultative rather than a . voting capacity. The United Lutheran Church in America was formed in 1919. It is a corporate union of the General Council, the General Synod and the United Synod South. The latter two are not represented in the county. In Dr. H. K. Carroll’s religious statistics for 1922, this new body was reported as the largest Lutheran body in America, outstripping the Synodical Conference, however, by only a very small margin. 6 The Norwegian Lutheran Church of America is a merger of the three Norwegian Lutheran bodies represented in the county and al- ready described, namely the Norwegian Evangelical Lutheran Synod of North America, the United Norwegian Lutheran Church in North America and the Norwegian Evangelical Lutheran Hauge Synod. This union took place in 1919. A small minority of the Nor- “Carroll, H. K., “Religious Statistics,” Christian Herald, April 7, 1923. Chart II.- — Integrating Movements of Lutheran Bodies Represented in County with Dates of Organization* General Council (1867) General Synod (1820) United Synod South (1863) The United Lutheran Church in America (1918) Hauge’s Synod (1876) Norwegian Church in America (1853) Norwegian Augustana Synod ► United Norwegian Church (1890) Norwegian Danish Conference 1 : National Lutheran Council ; (1919) Iowa Synod (1854) Ohio Synod (1818) Other Synods not in county Missouri Synod (1847) Wisconsin Synod (1849) Other Synods not in county Synodical Conference (1872) ‘This chart is based on material found in Warburton S. R., Yearbook of the Churches, 1920, pp. 59-71, Schaff-Herzog, Op. Cit. Vol. VII, pp. 83-93, Rememsnyder, J. B., Op. Cit., p. 22 and upon information courteously furnished by the Lutheran Bureau of New 7 York City. The dotted line indicates a federation rather than a complete merger. Rural Religious Organization wegian Synod did not join in this union, and is now a part Synodical Conference. Thus it will be noted that Lutheran bodies with a more national or racial experience are moving together. Among the r maining bodies, language and nationality are lines of cleavage. The National Lutheran Council was formed in 1919. This is not merger, but at present a qiore or less loose federation for cooperation in practical work. It has general headquarters in New York City, known as the Lutheran Bureau, “a national medium for information and service”, together with a Reference Library. The constituent bodies also compile together their statistical and annual church re- ports. The Roman Catholic Stream The Roman Catholic group is next to be considered. Their organi- zation also followed the German immigration as well as the Irish and the French. Table V shows that in eight of their churches the nationality or national extraction of the membership is German, in four Irish, in one French and in nine it is mixed. This latter characteristic is stronger in the case of the Roman Catholics than of any other single religious body. The earliest organizations appear to have been the following: Berry (1853); Mt. Horeb ; Bristol; Dane (1875); Deerfield; Dunkirk (1868); Dunn; Mazomanie (1856); Marshall; Paoli ; Roxbury (1845); Sun Prairie; Westport. Typical of these settlements is the story of Father Inama and the Roxbury church. Father Inama’s Colonization Project in Roxbury Township — An out- standing example of how the county was originally settled, and how people came to group themselves religiously as they have, is that of Father Adelbert Inama, a highly trained young German Catholic priest. He settled in Roxbury township in 1845 building for himself a small log cabin in one of the dells several miles from the Wisconsin River. He entered a large tract of land. Hither he invited his fellow country- men to whom he sold the land for the original government price. In this undertaking he was very successful and in process of time a compact German Catholic settlement had been formed. Although at the begin- ning the Americans were in the majority in this township, “today” as one writer puts it, “the town is as free from people of English extrac- tion as Germany itself.” 7 The European Origins — Christendom as a whole may be divided in- to three grand divisions, Greek, Latin and Protestant as outlined in Chart I. Of these the Latin, or Roman Catholic division is the largest. Christianity originated in Palestine at the beginning of the Christian era, and had firmly established itself in all parts of the Roman Empire 7 History of Madison, Dane County, and surroundings, p. 500. Wisconsin Research Bulletin 60 the beginning of the fourth century. It then became the religion of 'ie Empire. After a struggle lasting several centuries between the East nd West, the “Great Schism” occurred, splitting Christendom into rival ^atin and Greek churches following political cultural lines of partition. The final and permanent separation came in 1053 A. D. Protestant Christianity, the third grand division, began with the rise of the Protestant Reformation at the beginning of the sixteenth century. It spread rapidly and established itself firmly in northern Germany, Scandinavia, England and Scotland. Latin Christianity, or the Roman Catholic Church, maintained itself after a fierce struggle in the Romance lands, southern Germany, Bohemia, Moravia, Poland and Hungary. In most of these countries Protestantism had gained a considerable following, but a movement within the Roman Catholic Church itself known as the Counter-Reformation saved these parts of Europe to the old church. In the minds of the Irish, Protestantism and the domina- tion of the English were closely related, and for this reason their country remained overwhelmingly Roman Catholic, only about one- fourth of the population being Protestant. It should again be recalled in this connection that in those days religious beliefs and forms were matters not so much of the individual as of the state. Hence, whole states or provinces became either Protestant or remained Roman Catholic. It is from these Roman ' Catholic sections of Europe that many of the settlers of Dane County came, particularly southern Germany. Bohemia, Ireland and France. They organized here the churches of their native land. They brought with them the legacies of their re- ligious life, and these they transmitted to their children; among these are a love of the artistic and a reverence for beauty as seen in their stately church buildings, an appreciation of the mystical elements of worship in religion, a veneration together with an intense loyalty and zeal for their church with its rich and remarkable past. Recent Rural Life Emphasis — Although in the past the Catholic Church has been considered pretty largely a city church, nevertheless strong rural parishes have been built up. These have recently been receiving a larger measure of attention. In 1919 the National Catholic War Council was re-organized under the name of the Catholic Welfare Council, with headquarters in Washington, D. C. Among other depart- ments in this Council is one known as the “Social Action Department”, Its purpose is reported to be that of furnishing information and spread- ing Catholic principles and ideals in citizenship and industrial and social service. Within this department in turn is a Rural Life Bureau, directed by the Rev. Edwin V. O’Hara of Eugene, Oregon. “St. Isidore’s Plow,” devoted to the promotion of rural welfare, is the monthly publi- cation issued from this office. Two statements from a bulletin recent- ly put out by this Bureau may give some indication of the present emphasis. 8 “The Church is called to rural leadership not merely from 8 A Program of Catholic Rural Action, County, Oregon, pp. 4 and 18. based on a religious survey of Lane Rural Religious Organization 29 legitimate motives of self-interest, but also because of its peculiar fit- ness in helping to solve the rural problem. In a rural religious program strong emphasis must be laid on colonization. Every effort should be made to have Catholic farmers settle within a radius of five miles of a religious center. * * * Where Catholic farmers live out of touch with their co-religionists, there is not only danger of indifference, but the almost certainty of mixed marriage.” The “Reformed” Stream A large proportion of the very early settlers in the county were natives of New York, Ohio, Pennsylvania and Vermont, as is shown by Table III. These were, as a matter of fact, the first settlers. A small proportion came directly from England and Scotland. They were scattered in their settlement over the entire county. Even where they have been superseded by later settlers, they have left their marks in the names of the villages and townships. Most of these were organized by the “New Englanders” and named after the localities in the East from which they came. As these settlers were the pioneers in the county, so their churches seem to have been the first to establish themselves, for Table IV indicates that the “Reformed”, or nOn- Lutheran Protestant churches, were greatly in the majority in point of number of organizations, and also in the number of adherents which they had. These again were the churches to which they had become ac- customed in their former Eastern or European homes. Adventists — The Adventist bodies, two of which have figured in the religious history of the county, trace their origin to William Miller, a farmer by occupation. He became interested in prophetical Bible studies, which he believed taught premillennialism, the doctrine that the millennium is to be introduced by the personal, visible return (advent) of Christ. He further believed this return was near at hand. He was active from 1831-1849. The Advent Christians appear in the religious census for 1890, the only trace of this body which was found in the county. They are a corporate body dating from 1861. Among their distinguishing tenets may be mentioned the belief that the dead remain in an unconscious condition until the resurrecton at the advent, when those who had faith will begin an endless life upon this earth while the rest will suffer complete extinction. They believe that Sunday is the Christian Sabbath. The Seventh-Day Adventists have one small or- ganization in the county at Dane. They hold to the general teachings of the Adventists, and in addition they believe that Saturday, the seventh day of the. week, is the Sabbath obligatory upon Christians. Their organization dates from 1845 when a body of Adventists adopted this belief. The Baptists — Three different Baptists denominational groups appear in the county’s religious history. The designation “Baptist” is used for various bodies of Christians who hold to the common doctrine that immersion is the only Christian baptism. The Northern Baptist Con- 30 Wisconsin Research Bulletin 60 vention is the body to which several of the churches of the county be- long. It is commonly called simply the Baptist Church. They were active or had adherents in the following villages and townships: Berry; Stoughton (1861); Mazomanie (1857); Marshall (1860); Belleville (1856); Rutland; Sun Prairie. American Baptists trace their origin to Roger Wiliams and John Clark. The churches at Providence and New- port both contend for the honor of being the mother church. Their form of government is congregational, and they acknowledge no ecclesiastical authority. In history they are known for their evangelistic fervor, educational work and foreign missionary activity. Among the most important educational institutions founded by them are the Uni- versity of Chicago and Brown University. They were early champions of religious liberty. The Seventh Day Baptists constitute a group which is still a settle- ment in Albion township. The first of these settlers came from Alleghany county, New York, about 1842. One of their early and very important institutions was an academy organized in 1853. Prominent among the early leaders in this educational institution were Dr. C. R. Head and Professor A. R. Cornwall. Among the prominent alumni are the late Senator Knute Nelson of Minnesota and Prof. Rasmus B. Anderson, Congressman H. C. Adams, Chief Justice C. V. Bardeen and Dairy and Food Commissioner J. Q. Emery all of Wisconsin. This church was organized at Albion in 1843. Another church was organized near Utica in 1850. The latter has not persisted, as has the Albion settlement. The Seventh Day Baptists in the United States organized their first church at Newport, Rhode Island in 1671. They are a dis- tinct constituent body. As their name indicates they teach the keeping of the Jewish Sabbath as the day of rest and worship rather than Sun- day. The Freewill Baptists organized preaching stations at Medina (1845) ; Belleville (1853) ; and Rutland. They date from the year 1780 and originated in New Hampshire. In distinction from other Baptists, they are anti-Calvanistic and open communion Baptist. They were very active in anti-slavery agitation. The Adventist movement drew away many of their members. Recent efforts have been made to unite with the Northern Baptist Convention but thus far without success. None of their churches remain in the county. The Congregationalists — Congregational societies were organized at Black Earth (1853); Stoughton (1863); Mazomanie (1853); Medina; Sun Prairie; Windsor; York. As a denomination, they owe their origin to the Pilgrim Church at Plymouth, Massachusetts. The Plymouth colonists brought with them an organized congregation, that formed at Scrooby, England, in 1606. The germs of Congregationalism may be found in the Separatist movement of Queen Elizabeth’s time. The followers of this movement believed that each congregation was a separate religious community and should be ruled independently of the state or any other church. To the present day this is one of the un- Rural Religious Organization 31 derlying principles of Congregationalism. They hold, nevertheless, to cooperation and fellowship among the churches. The Congrega- tionalists were strict Calvanists. Calvin, it will be remembered, es- tablished his model “theocracy” at Geneva, which in many respects became the model of American Puritan commonwealths. They have been very active in missionary and especially educational work. Important educational institutions founded by them are Harvard, Yale and Oberlin Universities, as well as Mt. Holyoke, Smith and Wellesley, among women’s colleges.. The Evangelical Association — In the early days this denomination performed a useful service in ministering to the unchurched German immigrants. The Evangelical Association missionaries were the first to hold services among them in Madison and the surrounding country. They began this work in 1844. In 1853 they organized a church in Cottage Grove township and about 1858 at Mazomanic. In all they had nine organizations in the county, all of which have been d>banded, excepting the church in Madison. An example of what has happened to these organizations is the story of the abandoned church near the present German Lutheran Church at Hope, in Cottage Grove township. The Evangelical Association was the first to begin work among the German settlers in that vicinity who had been Lutherans in their native land. Later the Lutherans established an organization. Although the Evangelical Association maintained itself until very recently, they were compelled to abandon the church and the Lutherans now hold the field. This denomination was formed under the leadership of Jacob Albright, a layman trained in youth in the Lutheran faith, later becoming a member of the Methodist Church. He was ordained by his own followers in 1803, in which year the first general meeting was held. Like the Methodists in England, followers of Hauge in Norway and the Pietists in Germany, with their emphasis upon the need of con- version and an inner experience of the religious life, so these “Albrechtsleute” (Albright people), as they were often called, in- augurated a reactionary movement among the Germans in America against the formalism and frigid orthodoxy prevalent in their day. The Methodists — Three separate corporate bodies active at one time or another in the county come under the classification of Methodists. Methodism was set on foot by John Wesley during the first half of the eighteenth century. He and some of his friends were first called “Methodists” in derision of their methodical habits of life at Oxford. In theology nearly all Methodists hold to Arminianism as opposed to Calvinism. 9 They came out of the Episcopal church in England, as a revolt against its formalism and neglect of the masses. 9 Calvinism as defined in the Standard Dictionary consists of five points : (1) God elects individuals to be saved. (2) He designs complete redemption for these elect only. (3) Fallen man is of himself incapable of true faith and repentance. (4) God’s grace is efficacious for the salvation of the elect. (5) A soul once regenerated and converted is never ultimately lost. Arminianism by the same authority is opposed to these five articles and the following are its five points: (1) Conditional election; ( 2 ) Universal redemp- tion; (3) Salvation by grace, with which grace, however, man can cooperate; (4) Grace not irresistible; (5) Falling from grace possible. 32 Wisconsin Research Bulletin 60 The Methodist Episcopal Church with its characteristic pioneer zeal held stated services or had organizations at the followng places : Black Earth (1844) ; Horeb Corners (Mt. Horeb) and Blue Mounds (1854); Cambridge (1848); Cottage Grove; Stoughton (1867); Mazo- inanie (1856); Marshall (1869); Dane; Medina Township (1876); Belle- ville (1847); Paoli (1850); Oregon (1848); Primrose Township; Brooklyn; Syene Prairie; Springdale Township; Sun Prairie; Vermont Township; Verona Township; Windsor Township ; York: Township. Some of their work . is conducted in the Norwegian and German languages. The church at Cambridge is the oldest Norwegian Methodist Episcopal church in the world. The first religious organizations among the Germans appear to have been those of this church and of the Evangelical Association. The Methodist movement early spread to America, the first American Conference being held at Philadelphia in 1773. It is one of the oldest and strongest American churches and spread to Wisconsin with the westward tide of immigration. The church is connectional in organization and episcopal in government. 10 It has had a phenomenal growth in the United States. The missionary and educational work of this church are noteworthy, Northwestern, Syracuse and Boston universities being among the most important of its higher institutions of learning. The Primitive Methodists are of English origin. This body is the church of the English settlers of the eastern part of Albion Township. Another, no longer in existence, was organized near Mazomanie. This body organized in England in 1810 as a result of an effort to revive the spirit of the ancient days of Wesley and Whitfield, hence the name “Primitive” Methodists. The Free Methodists at one time had an organization in Medina Township. This denomination dates from 1860 and was formed as a result of an agitation in the Genesee Conference of the Mbthodist Episcopal Church. It was declared that the church was too lenient in tolerating worldly practices and contradictory in its teaching of entire sanctification. In government and general practices it is simlar to the mother church. The Moravian Church — A church among the German settlers in Windsor township was established by the Moravian denomination in 1885, and in the following year in the village of London. The Moravian church, in church history often called the Unity of the Brethren, traces its origin to Bohemia and Moravia, where the Ancient Unitas Fratrum was organized in 1567 by followers of John Huss. Persecuted in the land of its origin, its members fled to the estate of Count Zinzendorff in Saxony. Under his leadership this church, which had 10 This policy denotes (1) that the various conferences are presided over by bishops, one bishop having charge of several conferences, the denominational funds, however, being administered not by bishops but by independent boards; i the title of the local property is vested not in the local church corporation but in national governing body. Rural Religious Organization 33 been almost wiped out, was resuscitated. Thus a new religious movement was inaugurated, which spread to other parts of Europe, to England, and in 1741 to America. Bethlehem, Pennsylvania was settled by them and, became the center of missionary activity among the early German settlers and among the Indians. Its missionary work among the German immigrants in the Western States resulted in the establishment of the churches in the county. In religious history it is chiefly known for its pioneer missionary zeal. The Presbyterian Church in the United States— The Presbyterian Church began its service at Blue Mounds (1852) ; Cambridge (1847) ; Verona and Belleville (1847) ; Oregon. Scotland is the native home and John Knox, leader of the Scotch reformers, is the father of Presbyterianism. Some of the present members of the Cambridge church were born in Scotland. There are many varieties of Ameri- can Presbyterianism for it is as diverse as are the people who have blended to form the American nation. The first churches in America originated in New England, Maryland, Delaware and Virginia. They were in large part of English origin, their pastors being Church-of- England ministers holding Presbyterian views. This church is still represented in various parts of the county as the maps indicate. The Protestant Episcopal Church — English settlers organized Protestant Episcopal churches at Mazomanie and at Black Earth. This church does not appear in the United States census report for Dane county before 1860. This organization, however, established itself in Madison in 1839, the preliminary steps toward its organization being taken on July 25 of that year, and complete organization taking place in the spring of 1840 with 16 signers. As far as can be ascertain- ed this was the first church organized in Dane County. The Protestant Episcopal church is the lineal descendant and successor in America of the Church of England. Its history in America begins with the voyages of Englishmen in this direction and is one of the very oldest of Ameri- can churches. The Reformed Church in the United States — To the Swiss settlers in the southern part of the county belongs the Reformed Church. Although the Swiss immigration is of comparatively recent date, un- like that of the Germans and the Norwegians, these settlers did not establish separate constituent organizations, but affiliated themselves with the Reformed Church in the United States. This body is, however, of Swiss and German origin, but it is also one of the oldest American Protestant denominations. The first congregation was formed at Germania Ford, on the Rapidan, Virginia, in 1710, and the first synod meeting was held at Philadelphia, Pennsylvania, in 1747. It has always felt a special affinity for the Presbyterian Church since John Knox, the Scotch reformer, was a contemporary and an admirer of Calvin, whose doctrines have become the special heritage of the Reformed Church, using the term now in its narrower sectarian sense. 34 Wisconsin Research Bulletin 60 The Universalists — A congregation at Stoughton was formed by Uni- versalists. This denomination arose in America with the preaching of John Murray. It espouses the doctrine of final salvation for all, hence the name. Murray began preaching in America in 1770, itinerat- ing from Virginia to New Hampshire. ' The United Brethren in Christ — At one time the United Brethren had an organization in Rutland Township. This body originated in Pennsylvania in 1800, under the leadership of Philip Otterbein, a missionary of the German Reformed Church. In some respects its government is similar tto that of the Methodist Episcopal Church. In doctrine it is Arminian. The United Brethren were strenuous op- ponents of slavery. Federation Movements Among the “Reformed” Bodies — The most significant step in the direction of closer cooperation among these churches was taken w T ith the organization of the Federal Council of Churches of Christ in America made up of the constituent bodies shown in Chart III. It was formed in 1908 when representatives of thirty- one denominations ’ and communions, elected and authorized by their highest ecclesiastical judicatories, bound themselves together in a National Federation, with a constituency of one hundred and fifty thousand churches, with something like seventeen million members, and representing, through family relations, about fifty millions of people. It is especially to be noted that the Council is an official, or- ganized instrumentality just as definite in its constitution as the de- nominations themselves. 11 It includes practically all the more im- portant non-Lutheran or the “Reformed” churches, as they are characterized by church historians. Of the Lutheran churches, the General Synod was formerly a constituent body of this Federal Council. 12 When the United Lutheran Church was formed the question of its relation to the Federal Council came up and is now under consideration by its Executive Committee. Its present status is that of a consultative body. All the “Reformed” churches of the county, with the excepton of the Seventh-Day Adventists and the Universalists are constituent members of the Federal Council. It should be noted that the Protestant Episcopal Church is affiliated with the Federal Council only through its Commissions on Christian Unity and Social Service. The Streams Flow Together The settlement period in the county lasted until about 1870. Successive waves of people came on, each with its own social heritage, a part of which was to be contributed to the common life. As the years have n Tlie Quadrennial Session of the Federal Council, Homiletic Review, Dec., 1912, p. 429. ’-Warburton, S. R., Op. cit., p. 238. Rural Religious Organization 35 Chart III. — Federal Council of Churches of Christ in America, Constituent Bodies* Churches in county are marked with a cross. Baptist Churches, North National Baptist Convention Free Baptist Churches Christian Reformed Church in North America Churches of God in N.A. (General Eldership) Congregational Churches Disciples of Christ Friends Evangelical Synod of N.A. Evangelical Association Methodist Episcopal Church Methodist Episcopal Church, South African M. E. Church African M. E. Zion Church Colored M. E. Church in America Methodist Protestant Church x Moravian Church x Presbyterian Church in the U. S. A. Presbyterian Church in the U. S. A., South x Primitive Methodist Cnurch x Protestant Episcopal Commissions on Christian Unity and Social Service Reformed Church in America x Reformed Church in the U. S. Reformed Episcopal Church Reformed Presbyterian Church (General Synod) x Seventh Day Baptist Churches United Brethren Church United Evangelical Church Welsh Presbyterian Church 1 Federal Council > of Churches of Christ in America (1908) CONSULTATIVE BODY x United Lutheran Church J *This chart is based on information found in Year Book of the Churches, 1920, by Stacy R. Warburton. moved along these streams have run together, some have intermingled, some have been superseded, but all have been spread out together over the whole county. Frequently group identity has been maintained, but frequently also one group has slowly replaced another or has given of its life in the formation of new groups. The study of the neighbor- hood or primary groups in the county shows this tendency rather definitely. The 121 groups found were classified according to the predominating nationality first, when the group was originally recogniz- ed and, second, on the basis of its present composition. This tabulation shows that only nine fell in a “mixed” classification when considered originally, while sixty-four or over fifty per cent fall in this class at the present time. 13 Among the religious groupings a similar tendency of intermingling or coming together is clearly distinguishable. For example, by 1890 all the denominations which were destined to play a part in the religious 13 Kolb, J. H. Rural Primary Groups, Bui. 51, Agr. Exp. Sta., Univei*sity of Wisconsin, page 26. 36 Wisconsin Research Bulletin 60 history of the county had pretty firmly established themselves. In this year, moreover, a larger number of groups was active than at any other period. It is shown clearly in Table VIII that there were twenty-nine religious bodies in existence. ' At present this number is reduced to seventeen. Table VII also shows that what have been Fig. 10. — The Streams Flow Together in Western Dane County Rural Religious Organization 37 Fig. 11. — The Streams Flow Together in Eastern Dane County 38 Wisconsin Research Bulletin 60 termed the “Reformed” churches were the first to establish them- selves in the county. In 1850 only one Lutheran and one Roman Catholic church were reported. Even in 1870 sixty-four of the eighty- four, or 76 per cent were “Reformed” churches. While no figures are available to show the number of organizations in 1890, Table VIII in- dicates the marked growth of Lutheran and Roman Catholic organiza- tions. By this time the membership of either of these groups was larger than the “Reformed” group. Thus the religious groups of today are in the midst of readjustments occasioned by their past attempting to grapple with the present. This becomes the subject of the next chapter. TENDENCIES AND PROBLEMS OF READJUSTMENT What has happened in the past not only helps to explain the present but also furnishes clues to understanding certain trends in the future. The past, furthermore, has moulded the present. It has determined and passed on the structure of modern rural religious or- ganization. The religious organization of an area like Dane County, then, is an accretion of groups which cling to the forms and traditions they have inherited. The result is not only an interesting social phenomenon but a practical situation which presents difficult problems to churchmen, both clergy and laity. Table VIII. — Dane County Churches by Denominations and Communicant Membership for 1890, 1906 and 1916. Communicant Membership Years 1890 1906 1916 Total membershiD 22,353 32,073 39,507 Adventist bodies Advent Christian 23 Seventh Day Adventist 55 117 128 Baptist bodies Baptists 570 685 827 Free Will Baptists 70 Seventh Day Baptists 362 Congregational churches 898 1412 1 ,938 285 Evangelical Association 438 256 Lutheran bodies Synodical Conferences (Synod of Wisconsin) United Lutheran 120 1 ,041 576 58 762 431 Lutheran United Danish Church 144 35 Norwegian Lutheran Church Hauge’s Synod 239 377 809 Norwegian Church in America 1 ,574 4,088 2,242 5 ,584 2,530 6,565 United Norwegian Church Norwegian Free Church 900 German Augsburg Synod 936 Iowa Synod 1 ,240 1 ,394 Joint Synod ol Ohio . 1 ,226 906 Methodist bodies Methodist Episcopal 1 ,372 60 2,361 3,006 Primitive Methodist . . Free Methodist 56 Moravian Church 120 218 309 Presbyterian Church in U. S 854 1 ,069 1 ,649 Protestant Episcopal Church 180 494 811 Reformed Church in United States 50 50 449 Roman Catholic Church 9,043 150 12,068 15,626 U nitarians United Brethren in Christ 24 22 Uni versa list 30 Other Protestant bodies 894 Jewish Congregations 40 300 All others 62 725 Statistics of Churches of the United States, 11th. Census 1890. Special Reports, Bureau of Census, 1906, Religious Bodies, Part 1, pp. 371-372. Special Reports, Bureau of Census, 1916, Religious Bod is. Part 1, pp. 327-328. 40 Wisconsin Research Bulletin 60 Organization Tendencies The past, present and future are not separate stages but a continu- ous process. Change is not necessarily a sign of weakness or instability, but it is often the very thing which sustains life. These changes take more or less definite forms and can be examined as tendencies in group organization. The Present Organization of the Parishes — The composite laboratory map, Figure 1, gives a graphic picture of the agglomeration of churches, pastors’ residences, parish boundaries, and circuit lines, together with Z0 40 60 60 WO \Z0 Grand Total 117 1- 100 members 55 Lutheran 1 6 Roman Catholic z | Reformed 37 ;.v ; : 101-200 Lutheran 6 1 Roman Catholic 7 Reformed 11 P 201 and over 38 Lutheran 25 Roman Catholic 13 “Reformed” 0 Chart IV. — The Number of Church Organizations in Dane County by Number of Members and Major Groupings the official names of the denominations. The parishes are first to be considered. The parish boundaries enclose the areas of influence of the various churches. On this map they appear to be a maze of unrelated, overlapping lines. In order to understand their relation, the parishes must be dissected according to the major groupings, namely, the Lutheran, the Roman Catholic and the “Reformed". If it is remember- ed that there were three main streams of settlers who peopled the county, roughly corresponding to these three major branches of Western Christianity, the apparent enigma of the conglomerate nature Rural Religious Organization 41 oi rural religious organization may be particularly explained. On the separate parish maps, Figure 2 to 7 inclusive, for example, there is a surprising lack of overlapping. The apparent confusion is the result mainly of the intermingling throughout the county of the various parishes of these three main branches of the Christian Church having as they do such different historical backgrounds. Open County Parishes and the Neighborhood Groups — Various factors have entered into the formation of what have been called the Fig. 12. — Lutheran Open-Country Parishes and the Neighborhoods Compared for Western Dane County 42 Wisconsin Research Bulletin 60 Township Lines Neighborhood Boundaries Open Country Parishes W7///////A Overlapp/ng Parishes Fig. 13. — Catholic Open-Country Parishes and the Neighborhoods Compared for Western Dane County Rural Religious Organization 43 WEST DANE COUNTY Fig. 14. — “Reformed” Open Country Parishes and the Neighborhood Compared for Western Dane County 44 Wisconsin Research Bulletin 60 rural primary or neighborhood groups, which are the first socially significant groups beyond the family having some sense of local unity. 14 Religious, educational, nationality, economic, social and topographical factors largely determine these groups. Of these factors the religious is one ot tne most important. This is brought out very clearly in Figures 12, 13 and 14 in which the parish boundaries have been super- imposed upon the neighborhood boundaries. The close relation is particularly marked in the case of the Lutheran and Roman Catholic parishes. Examples in point are those of the Lutheran parishes of Primrose, Springdale, East Blue Mounds, German Valley, First Luther. Fitchburg and Danz in the western part of the county. All these are parishes of open country churches. In eastern Dane County the parishes are usually larger than the neighborhood of the corresponding name, as for example, Norway Grove, Pumpkin Hollow, Hope, Liberty Prairie, Deerfield, West Koshkonong and East Koshkonong. The Roman Catholic parish lines, Roxbury, Ashton, Pine Bluff and Spring Valley in the western half of the county, follow the neighborhood lines closely. In the eastern half, as at East Bristol and Westport, the parishes are larger than the neighborhoods. Only rarely do the parish boundaries of the “Reformed” bodies follow those of the neighbor- hood. Montrose and Mounds Creek in the western part and Token Creek, North Windsor, York and Albion in the eastern are instances. This is due no doubt to the fact that there are very few open country churches of the “Reformed” bodies. It is clear then that in the case of the Roman Catholic and particularly of the Lutheran churches in this county, religion is a powerful groupmaking factor. In the creation of these groups, of course, nationality also played a leading part es- pecially in the early days. The “Reformed” parishes are located mainly where the eastern American and English population elements settled in the early days. At that time many of these groups were of considerable size. These historical neighborhoods have in many cases disappeared and the present parishes are often smaller than the original neighborhoods, which represent what is left of the early group. The Lutheran and Roman Catholic parish groups have grown and expanded and are crowding or have crowded out beyond these original smaller groups. In contrast to the Roman Catholic and Lutheran parishes, those of the “Reformed” bodies are not very dense and often represent only a few scattered families. In the case then of the “Reformed” bodies, religion today is not playing such a leading role in the holding of these neighborhood groups as is the case with the parishes of the other major bodies. The breaking up of these smaller religious groups often presents a peculiarly difficult local problem. The Brerton church at Acorn is such an example. In that locality there was an early Ohio settlement. 14 Kolb J. H. Rural Primary Groups . Res. Bui. 51 Agr. Exp. Sta., University of Wisconsin. Rural Religious Organization 45 Many of the pople were graduates of Oberlin. They had a well-or- ganized church, a resident pastor and an excellent school. Later on German settlers came in, who were divided religiously between the Lutheran church at Hickory Hill and the Catholic church at Martins- ville. The Brerton church is now practically abandoned because of lack of members. Farms have changed hands and maiiy of the original settlers have moved away. The problem is one of incomplete readjustment. One of the families, for instance, is at a loss to know iBlANCHARD I . mie 1 Brooklyn Fig. 15. — Lutheran Village Parishes and the Trade Areas Compared for Western Dane County 46 Wisconsin Research Bulletin 60 New Glarus iBLANCHARD I vill Li i Fig. 16. — Catholic Village Parishes and the Trade Area Compared for Western Dane County Rural Religions Organization 47 iftlANCHARD I V'LUtj 1 Fig. 17. — The “Reformed” Village Parishes and Trade Areas Compared for Western Dane County 48 Wisconsin Research Bulletin 60 whether to try to keep the old church going by encouraging the son to remain on the farm, or to throw all its allegiance to the Lodi church, sell the farm and encourage the son to settle at a place where he will be among people with religious traditions more nearly like his own. Village Parishes and the Trade Areas — There is a rough sort of correspondence between the parish boundaries of churches located in villages and small towns and the trade areas of these various centers. Often the single parishes do not include the entire trade area as at Verona, Mt. Horeb, and Cross Plains. Taken together, however, all the parishes of churches located in a population center usually in- clude all of the trade area or that part which is not included within the limits of open country neighborhood groups. In this county a church is the church of a group rather than of the community for Table IX. — Correlation Between Size of Church and Location Size of Membership Major Grouping Numb er of cl [lurches with respect to location All City Vil- lage Ham- let Open country Grand total 117 10 45 23 39 Total 55 4 22 14 15 1-100 Lutheran 16 6 6 4 6 Roman Catholic 2 0 0 2 0 “Reformed” 37 4 16 8 9 Total 24 1 14 5 4 101-200 Lutheran 6 0 3 1 2 Roman Catholic 7 0 5 0 2 “Reformed” 11 1 6 4 0 Total 38 5 9 4 20 201 and over Lutheran 25 4 4 2 15 Roman Catholic 13 1 5 2 5 “Reformed” 0 0 0 0 0 even where the trade area boundaries and the parish lines correspond, many of the people within these limits often go to different churches. The tendency of the "Reformed” parishes to center about some hamlet or village is striking. Their parishes are in the main com- paratively small, although in quite a large number of cases they ex- tend to the limits of the trade area but the membership of such parishes is sparsely distributed. Relations Between the Open Country and the Village Parishes — How efficiently a church in some population center ministers to the open country population is a question often discussed. In this county, ac- cording to Table IX. there is a surprisingly large number of large open country churches for of the 38 churches with a membership of over 200, twenty are located in the open country. With what success the remaining 18, located in some center, reach the farmer cannot be Rural Religious Organization 49 determined with absolute accuracy from the data in hand, but it is obvious that many of their members live in the village or city ; never- theless these parishes have a large farmer constituency. Apparently in this county churches located in some center are just as successful in reaching the farmer as are those in the open country, although in one town a certain church was referred to as a farmers’ church in con- trast with another which drew its members more exclusively from the residents of the town. The small church seems to thrive best in the population centers. This is perhaps because the old American element of which most of the small churches are composed has drifted to these centers. It is also significant that of the open country churches slightly over half have a membership of over 200. The Church Membership — The combined membership of the rural churches in the county numbers 21,504, as appears in Table III. Of these 7,226 belong to open country churches, 1,449 to hamlet, 9,662 to village and 3,167 to city^ churches. The Lutherans comprise over half of the total membership ; the Roman Catholic nearly one-third and the “Reformed” slightly more than one-sixth. The Lutherans have a combined membership of 11,218. Approxi- mately half of these belong to open country churches. The Lutheran population belonging to open country churches comprises one-fourth of the total church membership of the county, and of this population four-fifths belong to the Norwegian Lutherans. Of the Norwegian Lutherans in turn, about one-half belong to open country churches. The German Lutherans, namely the Synodical Conference, the Joint Ohio and the Iowa Synods, are divided in about the same proportion while the English Lutherans have their churches and most of their constituency in the villages. In Perry Township the Norwegian Luther- ans have an open country church of over 1,000 members. The Roman Catholics have about one-third of their constituency in the open-country churches. The remainder belongs to churches in vil- lage or hamlet. These latter have also a very strong farmer con- stituency. For example, at Cross Plains, a hamlet, there is a church with over 800 members. The “Reformed” churches have their main strength in city, village, and hamlet, less than one-tenth of their members belonging to open country churches. About two-fifths of their total membership are Methodists. The Congregationalists, the Reformed Church in the United States, and the Presbyterian Church in the United States fol- low in order in point of numbers. Size of the Congregations — The churches of the county were ar- rayed according to size in Table X. Of the 117 churches nearly one- fourth have a membership of 50 or less ; nearly one-half 100 or less, and nearly two-thirds 150 or less. The largest number of small churches falls in the lot of the “Reformed” bodies, and none of these are larger 50 Wisconsin Research Bulletin 60 Table X. — County Churches Classified According to Size and Denominational Groups Classes of Membership Num ber of churches and the major gi oups All churches Lutheran Roman Catholic “Reformed” Total 117 47 22 48 1-50 28 8 0 20 51-100 28 8 3 17 101-150 14 4 4 6 151-200 11 2 4 5 201-250 11 11 0 0 251-300 9 5 4 0 301-350 0 0 0 0 351-400 4 2 2 0 401-500 3 2 1 0 501-600 3 1 2 0 601-800 3 2 1 0 801-1000 0 0 0 0 1001-1200 3 2 1 0 In Table XI the relation between the size of congregations and the frequency of Sunday services is shown. In general there seems to be a tendency on the part of the smaller “Reformed” congregations to have more frequent services. Table XI. — Frequency of Sunday Services in Relation to Size of Church Membership and Major Grouping Size of Churches Major Groupings Number of Churches Twice a week Once a week Once in 2 weeks Once a month Grand total ... 117 13 61 32 11 Total 55 8 25 15 7 1-200 Lutheran 16 0 6 8 2 Roman Catholic 2 0 2 0 0 “Reformed” 37 8 17 7 5 Total 24 5 10 8 1 101-200 Lutheran 6 0 1 4 1 Roman Catholic 7 0 5 2 . 0 “Reformed” 11 5 4 2 0 Total 38 0 26 9 3 Lutheran 25 0 13 9 3 201 and over Roman Catholic 13 0 13 0 0 “Reformed” 0 0 0 0 0 Rural Religious Organization 51 than 200. Their churches are situated in centers of population and what material there is for membership is divided among competing churches. The most frequent size of the Lutheran churches lies be- tween 201 and 250 members. Over half of their churches have more Table XII. — Churches with Resident and Non-Resident Pastors and Dis- tances Traveled by the Non-Resident Pastor Grouping Constituent bodies Churches with resident and noi with distances traveled by nor i-resident pastor i-resident pastor All churches Churches with resident pastor Churches with non-resi- dent pastor Distances in miles traveled by non-resident pastor 1-4 5-9 10 and over Total 117 64 53 20 21 12 Total 47 19 28 12 10 (T United Lutheran 3 0 3 1 1 i Lutheran Synod. bodies Conference .... 4 2 2 0 0 2 Norwegian Lutheran 27 13 14 7 6 1 Joint Ohio 7 2 5 3 1 1 Iowa Synod 6 2 4 1 2 1 Roman Roman Catholic Catholic 22 17 5 1 3 1 Total 48 28 20 7 8 5 Seventh Day Adventist 1 1 0 0 0 0 “Reformed” Baptist 2 2 0 0 0 0 bodies Seventh Day Baptist 1 1 0 0 0 0 Congregational 7 4 3 2 0 * 1 Methodist Episcopal 22 12 10 5 5 0 Primitive Methodist 1 1 0 0 0 0 Moravian 2 2 0 0 0 0 Presbyterian 6 3 3 0 1 2 Protestant Episocpal 1 0 1 0 0 1 Reformed in U. S 4 1 3 0 2 1 Universalist 1 . 1 0 0 0 0 than 200 members. The Roman Catholics have no churches with 50 members or less and over half of their number have over 200 mem- bers. There are two Lutheran and one Roman Catholic chuurch of over 1000 members. 52 Wisconsin Research Bulletin 60 Table XIII. — Average Number of Members Served by a Pastor and the Average Size of the Congregations Major groups Constituent bodies Average number of and for members for pastors churches Average Member- ship per pastor Average Member- ship per church Average of totals .. 263 .D 179.9 Average of totals.... 385.3 229.5 Lutheran bodies United Luth. 96.6 96.6 Synod, Conference 132.5 132.5 Nor. Luth 576.3 320.1 Jt. Ohio 315.0 90.0 Iowa Synod 173.7 115.8 Roman Catholic Roman Catholic 411.9 299.5 Average of totals. .. 101.8 76.4 Seventh Day Adventist 17.0 17.0 "Reformed'’ Baotist 100.0 100.0 bodies Seventh Day Baptist 180.0 180.0 Congregational 101.4 72.4 M. Bp. 90.2 69.7 Prim. M. E 57.0 57.0 Moravian 125.5 125.5 Presbyterian 95.2 63.5 Prot. Epis 37.0 37.0 Ref. in U. S 226.5 113.2 Universalist 50.0 Table XIV. — Length of Pastorates by Years of Service* Grouping Constituent bodies Number of pastors Yea irs of service in parish 1-3 4-6 7-9 10 and over Total 70 37 11 6 16 Lutheran bodies Total 22 9 4 9 United Lutheran. .. Synod. Conference.. Norwegian Luth Joint Ohio Iowa Synod 1 5 1 1 2 3 4 2 4 1 6 1 1 Roman Catholic Roman Catholic 15 4 3 3 5 ‘‘Reformed’’ bodies Total 33 24 4 3 2 Seventh Day Adventist 1 2 1 5 16 1 2 4 1 1 Baptist 1 Seventh Day Baptist 1 1 Congregational Methodist Epis Primitive Meth M ora vi a n 4 13 2 3 1 1 1 Presbyterian Protestant Epis.. . 1 Reformed in U. S... U ni versalist 1 1 ♦Of the 80 pastors in the county, information regarding years of service in par- i-hcs was secured for 70. Rural Religious Organization 53 The County Clergy— The total number of pastors serving churches in the county is 80. Of these 70 reside in the county, and 64 live within the area of their parishes and are therefore resident pastors. In Table XII the distribution of churches with resident and non-resident pas- tors is shown. Those churches which have a resident pastor number 64 and those whose pastor is non-resident, 53. The Lutherans have 19 churches with a resident pastor and 28 with a non-resident pastor ; the Roman Catholics have 17 churches with a resident pastor and 5 with a non-resident pastor, while the “Reformed” bodies have 28 churches with a resident and 20 churches with a non-resident pastor. Table XV.- — Correlation of Membership and Length of Pastoratfs members served by one pastor Total 1-3 4-6 7-9 10-12 13-15 16-18 19-21 22 and over Total 70 37 11 6 5 0 3 2 6 1-100 23 19 2 2 0 0 0 0 0 101-200 16 8 2 1 3 0 0 0 2 201-300 8 5 1 1 1 0 0 0 0 301-400 7 2 3 0 0 0 1 0 1 401-500 8 3 1 0 1 0 0 2 1 501-600 0 0 0 0 0 0 0 0 0 601-700 ] 0 0 0 0 0 0 0 1 701-800 3 0 1 1 0 0 0 0 1 801-900 0 0 0 0 0 0 0 0 0 901-1000 0 0 0 0 0 0 0 0 0 1001-1100 0 0 0 0 0 0 0 0 0 1101-1200 2 0 1 1 0 0 0 0 0 1201 and over 2 0 0 0 0 0 2 0 0 Table XVI. — Correlation of Membership and Length of Pastorates Grouped by Different Classes and then Cumulated Number of mem- bers served by one pastor Length of pastoral te by years and the cumulations 1 to 3 years 4 years and over 1 to 3 years cu- mulation on “less than” plan 4 years and over cumulated on “less than” plan Total 37 33 1-200 27 12 27 12 201-400 7 8 34 20 401-600 3 5 37 25 601-800 0 4 37 29 801-1000 0 0 37 29 1001-1200 0 2 37 31 1201 and over.. 0 2 37 33 The economy with which pastors are made use of is shown in Table XIII in which it appears that the Lutherans have one pastor for an average of 229.5 members ; the Roman Catholics one for an average of 299.5 members ; and the “Reformed” bodies an average of 76.4 mem- bers per minister. The tenure of office of the clergy is an important consideration in 54 Wisconsin Research Bulletin 60 rural religious organization. It appears from Table XIV that 47 per cent of the pastors remain with their churches three years or less ; 20 per cent serve from 4 to 6 years ; 12 per cent from 7 to 9 years ; while sixteen pastors or 21 per cent were on their charges 10 years or more. In this respect the Lutheran and the Roman Catholic churches, having larger congregations, have a better record than the “Re- formed” churches. There is an interesting and significant correlation between the num- ber of members a minister serves and the length of time he stays with his congregation. The small pastoral charge, which may consist of several congregations, is the least successful in holding its pastor for a considerable number of years. This appears in Tables XV and XVI. In Table XVI the classes of membership were grouped into larger classes and the length of pastorate combined in two groups of 1 to -3 years and 4 years and over and then they were cumulated. These tables show that of the 70 cases available in the data collected, 23 pastors served a constituency of 100 or less. Of these 19 had been on their charge only three years or less and only 4 remained for more than 3 years. In other words, judging by the data collected a congre- gation with 100 members or less has about one chance in six of keep- ing its pastors longer than three years. Let this be considered from another angle. There were 37 pastors who remained less than 3 years or about half the total number of pastors. Of these 37 pastors, 19 or about one-half served congregations of 100 or less ; 27 or over two-thirds served congregations of 200 or less and 34 or about 97 per cent served congregations of 400 or less. It would seem, then, that the smallness of these charges or the constituency served a single pastor is one of the causes of the short-time pastorate. This is due, undoubtedly, either to the inability of the small charge to furnish ade- quate support or to offer a sufficently large field of service. Or, on the other hand, it might be said that the short term of office is re- sponsible for the smallnesss of the charge. The former conclusion is probably the more nearly correct. If now the larger congregation be considered, the above conclusion will be substantiated and its corollary will be established. There were 47 pastors who served more than 100 members. Of these 29, or about three-fifths, had a tenure of office more than three years. There were 31 pastors who had a constituency of more than 200 members and of these 21, or about two-thirds, remained over three years. In other words, the likelihood that a pastor will remain for a longer term of office increases with the size of the congregation. Then, once more, there were 33 pastors who remained longer than three years. Of these 29, or about nine-tenths, ministered to more than 100, and 21, or about two-thirds, ministered to more than 200. Hence the larger the constituency, the more successful it is in holding the pastor for a long and presumably a more efficient pastorate. The Churches, Their Location and Circuit Relations — The Norwe- gian Lutherans have the largest number of churches of any single Rural Religious Organization 55 corporate body, namely 27, or nearly one-fourth of the total number. The German Lutherans, representing the Synodical Conference (Synod of Wisconsin), the Joint Synod of Ohio and the Iowa Synod have a total of 17 churches, or nearly one-seventh of the total. The Roman Catholics have 22 churches, or a little more than one-fifth of the total. The “Reformed” bodies have 48 churches or about two fifths of the total. The latter majQr grouping has the largest number of churches but the smallest constituency. The location of churches, as indicated in Chapter I, is further shown on the various church maps, Figures 2 to 7 inclusive. According to Table II two-thirds of the 117 county churches are located in city, village or hamlet, while one-third are located in the open country. The maps give a graphic picture of this tendency to locate in some center of population. For example, Stoughton has 10 churches, Marshall, Mazo- manie and Sun Prairie, each have five churches; Belleville, Cambridge, Dane, Middleton, Mt. Horeb, and Verona each have four churches ; Black Earth, Cottage Grove and Oregon have three churches apiece ; Blue Mounds, Cross Plains, McFarland, Paoli and Waunakee each have two churches; Albion, Deerfield, London and Mt. Vernon each have one church. In both Christiana and Pleasant Springs Townships there are two Norwegian Lutheran churches standing in close prox- imity to each other. This unusual phenomena is the result of the division occurring among the Norwegian Lutherans described pre- viously. The Hauge Synod was formerly represented in the county. The maps show that three of its churches have been abandoned, namely one at Cambridge, one at Deerfield and one at Primrose. This is the result of the corporate overhead union of the three Norwegian Lu- theran bodies. The circuit lines on the church maps show how va- rious churches are organized into one charge under the care of a single pastor. Problems in Readjustment The tendencies described bring the churchman face to face with problems which press for solution. Abandoned Churches — Thirteen abandoned churches were found in the county, to which several others have been added since the field study was made. Conversation with pastors and laymen leads to the belief that the process of consolidation and elimination will continue for some time. While more detailed investigation on this point might be desirable, it seems, nevertheless, that some of these churches have been abandoned because their membership has died out or moved away, as is the case of the Evangelical Association church at Hope. In other cases a small church group is broken up by reason of the greater present day facility of transportation and communication. Peo- ple in such groups follow the better routes of travel to the larger centers. One pastor in the county testified that during the winter months some of his members can travel six miles to Madison with 56 Wisconsin Research Bulletin 60 greater ease than the two miles by which they are removed from his open country church. Thus the larger groups grow at the expense of the smaller. A definite illustration is the case of the church at Swan Creek. This has been disbanded and the membership divided among nearby larger churches. In still other cases an abandoned church is the result of consolida- tion. Examples of this tendency are the Lutheran churches in Daley- ville, Primrose and Deerfield. Following the overhead organization of the Norwegian Lutheran churches, some of the local units have also merged to form larger local organizations. Overlapping Parishes — The composite church parish map, Fig. 1, shows a maze of overlapping and intertwining parish boundaries. System or purpose there seems to be none. After the parish maps have been dissected, however, as has been done in Figures 2 to 7 in- clusive, for each major grouping, there is a surprising lack of over- lapping. To be sure, there is some, but not nearly as much as in the composite parish map. Where there is an overlap, it usually repre- sents the meeting of heterogeneous population elements. This is illus- trated at Cottage Grove on the Lutheran parish map, where German and Norwegian Lutheran parishes as well as several Norwegian Lu- theran parishes overlap. There is considerable overlapping of Lutheran parishes in the Stoughton area. At Marshall there is a small English Lutheran parish within a larger German Lutheran parish. The Roman Catholic groups do not compete with one another and hence prac- tically no overlapping is found. Most of the overlapping is found on the “Reformed” parish map. This is due to the greater variety of denominations found in this family of churches. Thus it becomes clear that much of duplication in religious organization is due to the fact that each population element, whether German, Norwegian, English, American, Swiss, or Scotch, usually established the church to which it had traditionally become accustomed in its former place of abode. Unchurched Territory — Fig. 18 gives a graphic representation of the unchurched areas in the western half of the county. A similar con- dition exists in the eastern portion of the county. While these areas should be more intensively studied before final conclusions can be drawn, it is significant that most of these areas lie outside the neigh- borhood boundaries, which condition denotes a lack of local, social cohesion. In many cases, it may be true that the people who live in these unchurched areas are affiliated with some church but usually they are scattered families living at considerable distance from their churches. They are, therefore, outside the real parish influence and. presumably, not in effective relationship with their church. In some cases, such as Malone Valley, McPherson Valley and Scotch Lane, these unchurched areas represent small neighborhood groups, pock- eted between larger and uncongenial groups, and too weak to maintain their own religious institution. In other cases the abandonment of a church has left some people Rural Religious Organization 57 WEST DANE COUNTY ; - t s *~ ~ -par" J Township Lines Sf i 1 rf |oUNLM» hoi| / / 1 __ i / rt / unu ui\nuw Boundaries rr ^ — 3 |- LI Ji RINGF1EL9 1 rnmMM Unchurched areas Abandoned Churches Fig. 18. — Unchurched Areas and Abandoned Churches for Western Dane County 58 Wisconsin Research Bulletin 60 without local affiliation and therefore, the unchurched territory is the result. South of Middleton two abandoned churches are located, the one within and the other in very close proximity to unchurched ter- ritory. The Non-Resident Clergy — Fifty-three churches in the county are served by non-resident pastors, as shown by Figures 2 to 7 inclusive, of which, however, as is shown in Table XII, 20 live within a four mile radius and 21 within a nine mile radius. Twelve travel ten miles or more to reach their churches and are, therefore, presumably not in very close touch with their membership. Among the Roman Catholics the proportion of local resident priests is the largest. While the propor- tion of non-resident pastors among the Lutherans is large, inspection of the parish and church maps will show that often the group of par- ishes which one man serves constitutes one large compact parish or charge, affording an adequate opportunity to keep in close contact with the parishioners. Thus the ministers who in classification appear tech- nically as nonresidents may in many cases be regarded as virtually resi- dent for two or three churches. Many of the pastors travel long distances to their churches. For example the Presbyterian pastor living at Oregon travels 18 miles to serve a church at Waunakee. Many of the pastors of the “Re- formed” churches live at Madison and sometimes travel more than 20 miles to their charges. Nationality and Language Factors — Table IV shows the correlation between nationality, or racial origin, and the various denominations. The Lutherans have Norwegian and German elements in their con- stituency within the county. The Roman Catholic is the church of the Germans, the Irish and the French. In the “Reformed” grouping there is found every nationality element except the Irish and the French and one-third of the churches have an American constituency. Neither the Lutheran nor the Roman Catholic churches of the county have an old American constituency. By this is not meant that they are not American in spirit or citizenship, but that they still retain some- thing of the old-world heritage taking the form of tradition and lan- guage. Of the “Reformed” grouping, the Methodist Episcopal church ministers to the largest variety of the population. The Roman Catho- lics seems the most successful in bringing various elements into the same church, as is the cases of Paoli, Verona, Dane and Mt. Horeb. The Norwegians are usually Lutherans. The Germans divide their allegiance among the three groupings, the Lutheran, the Roman Catho- lic, and the “Reformed”, depending usually upon the province in Europe from which they or their ancestors came. Over half of the German churches are Lutheran, about one-sixth belong to the “Reformed” bodies while the remaining one-third are Roman Catholic. The old Americans, the English, the Scotch and the Swiss always hold to one or the other of the “Reformed" bodies. Rural Religious Organization 59 The languages used in the church services represented in Table XVII are the English, the Norwegian and the German. English only is used in 46 per cent of the churches ; 31 per cent use German at some or all of their services ; and 23 per cent use Norwegian at some or all of their services The Germans retain more generally the language of their fathers, for, although tfrey settled here at about the same time as did the Norwegians, one-third of their churches use the German language only, while one out of twenty-seven of the Norwegian churches use the Norwegian language entirely. Of the “Reformed’’ churches 79 per cent use the English language exclusively. » Table XVII. — Languages Used in Church Services Languages used by number of churches Grouping Constituent bodies All Eng- lish only Norwegian German only % A. H Only 3 4 A. Total 117 54 1 7 16 3 12 5 15 4 ~ , Total 47 6 1 7 14 3 8 3 5 Lutheran United Lutheran 3 3 bodies Synod. Con- ference 4 1 1 1 1 Norwegian Lutheran.. 27 2 1 7 14 3 Joint Ohio ... 7 5 2 Iowa Synod 6 2 2 2 Roman Catholic 22 10 2 6 3 Total 48 38 2 2 j 4 1 Seventh Day Ad- ventist . 1 1 Baptist 2 2 All others or “Reformed” Seventh Day Baptist .... Congrega- tional 1 1 bodies 7 7 Methodist Episcopal Primitive Methodist 22 18 2 j 1 1 1 Moravian .... 2 1 1 Presbyterian Protestant Episcopal Reformed in U. S 6 6 1 1 4 1 3 Universalist 1 1 THE FUTURE IN RURAL RELIGIOUS ORGANIZATION The story of the developments which led to the present order of things in rural religious organization has been told. A view of things as they now are has been presented. The chief value of such a study lies in the fact that it may reveal some of the problems and may sug- gest possible solutions to those who are responsible for the future of rural religious organization. Strategic Location of Churches — It would seem that the church located in some population center has the better chance of surviving, for of the twelve abandoned churches which are shown in Figures 1 and 2, nine are located in the open country and since the field study was made others have been added to the list. On the other hand, a large proportion of successful rural churches, notably Lutheran and Roman Catholic, are located in the open country and present indica- tions point to their continued usefulness and efficency. It is a ques- tion, well worth further investigation and study, as to how successful farmers’ churches have been when located in a village, and whether a farmer feels as thoroughly at home in a village church as he does in an open country church. A certain farmer and villager psychology must be taken into consideration. It is clear, however, from this study that where there is a reasonably large and homogeneous neigh- borhood group, outside the limits of a population center, held together not only by religious purpose, but by other group-making factors, the church of that group should be located at some convenient point in the open country. The wisdom of such a choice is well illustrated in the county by the following examples : Roxbury, Pine Bluff, Daleyville, Primrose, Norway Grove, Liberty Prairie, West Koshkonong and East Koshkonong. On the other hand, if the village is the natural center of some group, then that center may well become the place for the church. “Sufficient Volume of Business” — The twelve solid circles on the map indicating abandoned churches, show what is happening to the small churches in the county. Table XI indicates that these small churches cannot hold their pastors for any length of time, for 86 per cent of the pastors serving a combined constituency of 100 or less remained on their charges only three years or less. The sayings, “Nothing succeeds like success” or “The destruction of the poor is their poverty,” are well illustrated by the churches of this county. The large church appears to prosper and succeeds in holding its pastor, while many of the small churches are dying out, and cannot hold their pastors for a long period. And yet in the face of such facts as these, churches continue their competition and denominational boards con- tinue to subsidize such churches with good missionary money. Overchurched and Underchurched Territory — The analysis of the rural church situation in the county reveals the fact that many of the small towns are hampered by overchurching and overlapping. This Rural Religious Organization 61 study has told the story of how this all came to pass. Perhaps it could not have been otherwise. Particularly acute cases were found at Mazomanie, Middleton, Marshall, Cottage Grove, Sun Prairie and Cam- bridge. Indeed the village or town that does not suffer on this ac- count is the exception. An analysis of one definite case must suffice. According to the church map, Figures 1 and 2, and the parish maps, Figures 6, 7, 10 to 13, the nine churches in Mazomanie and Black Earth townships including Mounds Creek have areas of influence prac- tically confined to these townships. They are situated within a radius of three miles. The combined population of the two townships includ- ing the villages located within them, in 1920 was 2027, or an average of 225 persons per church. According to the federal Census Report for Religious Bodies in 1916 15 there was one church organization for every 500 people in the state of Wisconsin. The combined member- ship of these churches in question is 865 and if the membership of the two largest organizations is subtracted, the remaining seven churches have a combined membership of 450, material for one good sized or at best two ordinary sized congregations. There are three resident pas- tors and six churches served by non-resident pastors. Surely a situation such as this is entitled to the earnest consideration of churchmen and church leaders for it can hardly be considered a condition of progress. Such churches cannot pay adequate salaries ; inadequate salaries event- ually mean inefficient leadership ; and inefficient leadershiip results in lifeless organization. The problem of caring for the unchurched territory is more easy of solution. Such territory is “no man’s land”, because it lies outside the natural neighborhood groups or because diminutive groups are not accustomed to the churches of the larger groups nearby. The local churches of the county which have adapted their programs to meet the needs of people whose religious traditions are somewhat different, are fulfilling their missionary opportunity. May not the church that feels an obligation to the people in lands across the seas well learn also to feel responsibility for the stranger within its gates. Absentee and Migratory Pastors — Another of the problems of the country church is what has been called “absentee pastorism.” Try as he may, the absentee pastor can only be to the church about what the absentee landlord is to the community. And the migratory pastor may be to the church largely what the migratory tenant is to the farm. This problem is intimately connected with that of size of congrega- tions and with that of overchurching and overlapping. Churches are often small and weak because there have been too many of them, or because shiftings and shrinkings have made growth impossible. Be- cause they are small, they cannot support resident pastors, and if they support them at all, the support is so inadequate that pastors are not willing to remain for any length of time. Table XV shows not only “Religious Bodies, 1916, Bulletin 142. Bureau of the Census, p. 103. 62 Wisconsin Research Bulletin 60 that 31 per cent of the pastors have a combined constituency of 100 or less, but also that 86 per cent of the pastors serving churches of 100 or less, remain but one to three years. Many authorities insist that a rural cleryman should be on his charge at least three years be- fore he can do his most efficent work. The conclusion seems to be that the various church bodies which are most closely related by tra- dition and belief may best serve their respective groups and the cause of the community by getting together on some plan of comity and cooperation and agreeing on a policy of consolidation, elimination or exchange. Another angle of this problem relating to the pastor as important, if not more so, is that of his preparation and his attitude toward his chosen field of work. This most important social institution cannot hope to rise above the level of its leadership. Those pastors in the county who are seeming to meet with the greatest success are not those who regard their tasks as temporary or as stepping stones to city churches. Equally important is the recognition on the part of the local parish, that a well-trained, sympathetic pastor is just as essential to the success of a local community as is the school teacher, the merchant or the physician. Where such a state of mind exists the questions of compensation and tenure are not baffling problems. The Language Question — The citizen who has grown to maturity in a foreign country, will do well if he is able to master the English lan- guage sufficiently to help himself in all his business transactions. Unless he be thrown into daily and intimate contact with English- speaking neighbors, he will never feel quite at home in the newly ac- quired language. Religious expression being so voluntary and spon- taneous, he can be least expected to get satisfaction from a service con- ducted in what is to him a foreign language. Yet there can be little doubt that often the foreign language is used in these services for a much longer period than may be necessary. The older people are sometimes slow to appreciate the needs of a younger generation, and often the young respect the wishes of their elders too much to demand a change. There are cases in the county where the refusal on the part of a congregation to adopt the language of the country has re- sulted in internal controversies, or even in the establishment of rival churches. The Need for a State-Wide Consciousness — It is almost inconceiv- able how these various problems which are so challenging can be met without some inter-group understanding. The logical starting point in the solution would seem to be the state as a whole. A smaller unit would hardly have sufficient information. A single denominaiton or a number of local churches can do little. For many rural churches the situation will be hopeless until the time comes when a state-wide policy of reparishing and the redistribution of pastors can be adopted. At present many church groups work as though they were entirely Rural Religious Organization 63 ignorant of one another’s existence. Leaders with vision narrow and provincial, fail to see their task as a whole, and in its proper relation to other churches and to other rural agencies in the local community itself, and in the state as a whole. The question of organic union is not involved here. No one familiar with the historic background de- tailed in this study would advocate any idealistic scheme of amalgama- tion of widely divergent types. It is simply a matter of rising above provincialism to a plan of Christian statesmanship in this most im- portant sphere of rural life. Contents Part 1 — Present Outlines of Religious Organization Page Complex nature of religious organization.. 1 Religious bodies enumerated and the major groupings 2 Distribution and location of churches, membership, and parishes : 3-11 Part II — Social History of the Religious Groups The streams of early settlement 12 The Lutheran stream 20 The Roman Catholic stream 27 The “Reformed” stream 29 The streams flow together 34 Part III — Tendencies and Problems of Readjustment Organization tendencies 40 Problems in readjustment 55 Part IV — The Future in Rural Religious Organization Strategic location of churches 60 “Sufficient volume of business” principle 60 Overchurched and underchurched territory 60 Absentee and migratory pastors 61 The language question 62 The need for state-wide consciousness 62 Research Bulletin 61 January, 1925 A Study of the Principal Changes Which Take Place in the Making of Silage , nn i n l U *} AHH W. H. PETERSON, E. G. HASTINGS and E. B. FRED Agricultural Experiment Station of the University of Wisconsin Madison Contents The nature of silage making 1 General plan of the experimental work 2 Filling the silo 2 Temperature determination 3 Samples for analysis 3 Observations on the samples 4 Methods of analysis 4 Moisture 4 Volatile and non-volatile acids 4 Alcohol . 4 Carbohydrates .... 5 Forms of nitrogen 5 Sampling and analysis of silo gases 5 Number and kinds of bacteria 5 Chemical and physical changes in the silo 12 Gases 12 Temperature changes 14 Fermentation products 15 Non-volatile acids other than lactic 17 Forms of nitrogen in corn silage 20 Loss of dry matter 22 Destruction of starch 22 Destruction of pentosans 23 Effect of inoculation on the production of silage 24 Experimental silage made in milk bottles 25 Experimental silage in large containers 25 Effect of inoculation under conditions existing in the silo 27 Summary 29 Literature cited 31 A Study of the Principal Changes Which Take Place in the Making of Silage* From the Departments of Agricultural Chemistry and Agricultural Bacteriology S OME IDEA of the importance of silage for Wisconsin may be gained from a survey of the number of farms and the number of silos. In 1923 there were somewhat less than 200,000 farms in Wisconsin and more than 100,000 silos. No other state in the Union compares with Wisconsin in this respect. New York comes second with about 55,000 silos. There are said to be 500,000 silos in the United States and more than one-fifth of these are within the boundaries of this state. A farm practice that is used by such a body of farmers is certainly deserving of scientific stud}'. The Nature of Silage Making When green plant tissue is placed in a closed contained, the plant cells continue to respire and produce carbon dioxide and other prod- ucts. As a result of this respiratory process The oxygen is exhausted in a few hours. During the time that oxygen is abundant, carbon dioxide is practically the sole product that is eliminated from the cells, but when the oxygen supply becomes much reduced, other compounds such as ethyl alcohol, acetic and lactic acids may be formed in small quantities./ Coincident with the diminishing functions of the plant cells comes an increase in the activities of lower forms of life, such as bacteria, yeast, and molds. Due to the absence of air the molds probably func- tion for only a few hours. The yeasts may continue to grow for a longer time, but plate counts indicate that they usually disappear in a few T days. The chief forms of micro-organisms that remain are bac- teria. When the green material is ensiled, a great variety of bacteria is present, but many of the forms on the growing plant cannot sur- vive under the conditions existing in a silo. Other forms present on the growing plant find conditions extremely favorable for their repro- duction and increase to enormous numbers. The exudation of the sap from the plant tissue is one of the important steps in the process of *The writers were assisted in the analytical work by J. A. Anderson, L. A. Burkey, Audrey Davenport, B. P. Domogalla and E. G.' Schmidt. ? Wisconsin Research Bulletin 61 making silage. Owing to its content offrsugars, proteins and salts, the cell sap furnishes an excellent food for the growth of various kinds of micro-organisms, especially the lactic acid bacteria. The conditions in the silo are usually favorable for the growth of this group of bacteria and hence it is not surprising that they multiply at an enormous rate. These myriads of bacteria continue for a considerable time and then slowly decrease. Changes in food supply and accumulation of fer- mentation products finally result in unfavorable conditions and a con- sequent reduction in the number of organisms. The bacteria produce profound chemical changes in the plant tissue. Many of the carbohydrates are converted into carbon dioxide, ethyl alcohol and organic acids ; the proteins are partly hydrolyzed into proteoses, peptones, amino acids and ammonia. These chemical changes are most rapid in the early stages of the fermentation. This is also the time of greatest increase in temperature and the maximum number of bacteria. To follow and correlate all these changes simultaneously would ob- viously require the cooperation of many workers, particularly in the early stages of the fermentation when frequent samples must be taken. It is for this reason that most of the published articles on silage mak- ing have been concerned with a study of but one or two of the sev- eral factors involved. To draw conclusions from bacteriological work done on one sample of silage and chemical analyses on another is, to say the least, unsatisfactory if not impossible. It was the purpose of this investigation to take a number of samples during the course of the fermentation and to make chemical and bacteriological examina- tions on the same sample at the same time. It was hoped that in this way the data would give a more unified and accurate picture of the process. This has been accomplished to a certain degree. No review of the voluminous literature on silage other than that incident to a discussion of the experimental data will be given, as numerous bibliographies have already been published (5, 10, 13, 15, 16, 23). \ General Plan of the Experimental Work Filling of Silo — A stave silo 10 feet in diameter and 30 feet high was filled on September 16, 1922, with corn, Golden Glow, Wisconsin No. 12, selected from a large field of about 50 acres. The corn from this portion of the field was of even growth, prac- tically free from weeds, and in the early dent stage. The leaves were still green and the ears were well filled. As far as could be judged it was in the best possible condition for silage purposes. It was brought to the silo within 1 to 2 hours after being cut and run through a Blizzard cutter and shredder set to cut one-inch pieces. During the filling processs three men were kept in the silo packing the cut corn. When about 10 feet of corn had been placed in the silo the cutter was stopped and about 200-300 lbs. of the cut corn thoroughly mixed with a fork. A sample of about 30 lbs. was taken for chemical and Changes In The Making of Silage 3 bacteriological analyses. At the same time, three bags were filled with 40 lbs. each of corn, closed in such a way as to prevent seepage in and out of the bag and placed in the silo equidistant from one another about half way from the wall to the center. The bags were made of heavy canvas which had been rendered water proof in order to prevent any change in their contents due to the downward move- ment of silage juices. That the water-proofing material did not in- hibit the growth of bacteria was determined by a preliminary fermenta- tion of glucose yeast-water medium in one of the bags. When the silage was removed, several months later, the bags were taken out, the contents weighed, the dry matter determined, and from these data the loss of dry matter was calculated. A second set of bags was placed in the silo about 10 feet above the first set. These two levels bounded a zone of silage from which all samples were taken during the fermentation by boring through the wall and re- moving the silage. To determine the settling of the silage and the position of the experimental zone of silage, a piece of 2"x4" timber 9 feet long was placed in the silo at the same level as the lower set of bags. Wires attached to its ends led to the top of the silo where markers were placed on each wire. As the silage settled the marker moved down the wall of the silo and the distance from the final posi- tion to that at the time of filling showed how much the silage had set- tled at the location of the scantling. The settling was most rapid during the first 24 hours when the marker moved down 3 feet from its original position. After 5 days the distance was 4 feet; after 30 days, 5 feet and when the silo was opened Sy 2 feet. During this time the silage at the top had settled 10 feet so that the total silage now occupied only about two-thirds of the volume of the silo. Temperature Determination — An iron pipe one inch in diameter 20 feet long was placed in the middle of the silo with the lower end resting on the timber. After the silo had been filled thermometers were suspended in the pipe at various distances from the top. To prevent any appreciable change in temperature while the thermome- ters were being withdrawn from the pipe and- read, the lower part of the thermometer was inserted through a cork into a slender test tube filled with water. With this arrangement no visible change in temperature took place for several minutes after the thermometers were brought to the surface. This simple device was entirely satisfactory and eliminated all the possibilities of error which accompany electrical measurements of temperature. Samples for Analysis — At the time of filling each set of bags, about 30 pounds of the cut corn were taken to the laboratory for analysis. After mixing thoroughly, three samples of two kilos each were taken for moisture determination. Successive portions of about 10 kilos weight were pressed in a powerful tin-plated press and about 2 liters of juice collected. This juice was used for chemical and bacterio- logical analyses. In making the analysis on this liquid it is assumed that the plant juice expressed is representative of the entire contents 4 Wisconsin Research Bulletin 61 of the cells. While this is not entirely accurate the method yields comparative results and is probably as free from error as any exist- ing rapid method that could be used to furnish samples for both chemi- cal and bacteriological purposes. The drum and plate of the press were carefully cleaned and sterilized each time before using to free them from any bacteria. The samples taken during the fermentation were obtained by boring a three-inch hole through the wall of the silo and then by means of an iron hook removing about 10 kilos of the silage. The hole was then closed with a wooden plug. To avoid any effect of air intro- duced in taking the previous sample a new hole was bored at least five feet from the old one each time a sample was taken. Ten samples were taken through the wall, two from the bags, two from the silage adjacent to the bags and two of the green corn, making 16 samples in all. Observations on the Samples — The first sample taken 24 hours after the silo was filled still had the characteristic appearance and odor of green corn forage. After 2 days, the forage first began to have a slight odor of silage. This odor was distinct after 3 days and very decided at the end of 5 daj^s. At* the end of 16 days the aroma, color, and taste were typical of a first class silage. Methods of Analysis Moisture — One kilo samples were dried rapidly at 65° C. and then placed in a steam oven and dried at 98° C. to constant weight. Volatile and Non-Volatile Acids- — The silage juice was acidified with sulfuric acid and distilled with steam until 100 cc. of distillate con- tained less than 1 cc. of 0.1 N acid. The residue from the steam dis- tillation was concentrated on a steam bath to about 50 cc. placed in a Kutscher-Steudel extraction apparatus, and extracted with ether for 48-60 hours. The ether was evaporated off and the extracted acids titrated with 0.1 N barium hydroxide. The lactic acid in the barium salts was determined by oxidizing a portion with permanganate by the Von-Furth-Charnass (9) method. As a check on this procedure, in a few cases the barium salts were evaporated to 10 to 20 cc. and then diluted with absolute alcohol to a volume equivalent to 90 per cent alcohol. The barium lactate is kept in solution while the other barium salts of organic acids such as malic, etc. are precipitated. A portion of the alcoholic solution was evaporated to dryness in a platinum dish, acidified with 35 per cent sulfuric acid, and ignited to barium sulfate. From the weight of barium sulfate the equivalent weight of lactic acid was calculated. Alcohol — The silage juice was neutralized to phenolphthalein, sat- urated with sodium chloride, and the alcohol removed by distillation. The distillate was oxidized with potassium dichromate and sulfuric acid and the resulting acids distilled and titrated. From the titration figure the corresponding quantity of alcohol was calculated as ethyl alcohol. Changes In The Making of Silage 5 Carbohydrates — Total reducing sugars were determined after hydrol- ysis and clarification by the Shaffer Hartmann (27) method. Pento- sans were determined by the Krober (14) phloroglucin method. Starch was determined by extracting the samples dried at 65° C. with alcohol, digesting with saliva and determining the reducing sugars formed. Forms of Nitrogen — Total nitrogen was determined by the Gunning modification of the Kjeldahl method, ammonia by Folin’s (11) method, and amino nitrogen by Van Slyke’s (29) method. Soluble proteins and intermediate digestion products were precipitated by Folin and Wu’s (7) tungstic acid method and the nitrogen calculated by difference from the total soluble nitrogen and the nitrogen in the filtrate. Sampling and Anjalysis of Silo Gases — Samples of the silo gases were obtained by means of an iron tube and aspirator bottles contain- ing mercury. The tube was 'closed at one end, sharpened to a point and had three holes bored at an angle through the wall. The first or sampling bottle was connected to the second or reservoir bottle by means of a rubber tube attached to a side tube at the bottom of the bottle. The mouth was closed by a two-hole stopper containing glass tubes which could be closed by stop cocks. To obtain a sample of gas, the iron tube was driven into the silage and then attached to the sampling bottle. By opening the right stop cock and lowering the reservoir, gas was drawn into the first bottle. When full of gas the stop cock was closed, the second tube opened and the gas forced out into the air by raising the reservoir. After washing out the apparatus three or four times with the silage gases, the last sample was retained in the bottle and taken to the laboratory for analysis. The gas was transferred to a Burrell gas apparatus and analyzed for carbon dioxide, oxygen, hydrogen and hydrocarbons. Number and Kinds of Bacteria Counts of the number of bacteria in silage juice were made by the dilution method, plate method and direct microscopic method. Aside from the total number of bacteria, dilution counts were carried out with various kinds of media. At each analysis 1 per cent concentrations of glusose, succose, lactose, xylose, mannitol and sodium lactate in both yeast water and in beef peptone media were inoculated with different dilutions of the juice. In order to get an estimate of the total number of different sugar- fermenting organisms in silage. 6 to 7 dilutions were made for each of the media. The reaction of the media was adjusted to pH 7.0. The various dilutions of the expressed silage juice used to inoculate the various sugar media were obtained as follows : 1 cc. of the juice was added to a 99 cc. water blank for the 1/100 dilution, but for succeed- ing dilutions 10 cc. of the 1/100 or 1/1000 etc. dilutions was added to a 90 cc. water blank. Tubes of the sugar media were inoculated with 1 cc. of each dilution, so that growth in a tube thus inoculated would mean that the cc. of silage juice from which the dilution was made 6 Wisconsin Research Bulletin 61 contained at least as many bacteria as was represented by the dilution. On the basis that 1 cc. of the solution contained but one viable organism, the estimated total number of bacteria per cc. of juice is probably much lower than is the actual number. All inoculated sugar media were incubated at 37° C. Observations as to growth, its appearance, formation of gas etc. were made after 24, 48 and 96 hours. At the end of this time, incubation was con- tinued at room temperature for 2 weeks when titrations were made of the acid production. Glucose-yeast-water agar was used for all plate counts. The direct counts were made as follows : by means of a standardized platinum loop, 1/100 of a cubic centimeter of the juice was spread over 4 square centimeters, dried, fixed with heat and stained with Loeffler’s methylene blue. The growth and action of silage bacteria in litmus milk was de- termined by inoculating tubes of milk with the various dilutions. One set of inoculated milk tubes, covered with a half inch layer of melted vaseline was heated for 10 minutes at 80° C. after inoculation. Methy- lene blue reduction in milk was carried out at each analysis with a 1/200,000 dilution of the stain. Table I. — Number of Micro-organisms in the Juice of Corn Silage Plate counts In 1 cc. of juice No. Age Yeasts Bacteria 1 Days 0 500,000 2,400,000 2 1 500,000 1 ,340,000,000 3 2 33,000 880,000,000 4 3 700,000,000 1 ,640,000,000 5 5 6,700 6 11 2,100 1 ,575,000,000 7 30 2,500 856,000,000 8 44 260,000,000 9,500,000 9 85 3,300 10 132 1 ,800 2,000,000 The results of the bacteriological analysis by means of plate counts are given in Table I. In harmony with the data from various investi- gators (6, 12) it is plain that the number of bacteria in the ensilage undergoes an enormous multiplication during the first few days after the fodder is put in the silo. Almost immediately after the cut corn is ensiled there is a rapid multiplication, especially of the lactic-forming bacteria, from less than 3 million to more than 1 billion per cc. The significance of this change in the flora of silage and its effect on the composition cannot be questioned. In spite of fluctuations, this high number of living bacteria was maintained for 11 days and not until more than 30 days had elapsed was there any marked decrease in numbers. It was almost four months after the corn had been put into the silo before the number of bacteria decreased to that found on Changes In The Making of Silage 7 the original sample. The greater part of both thermal and chemical changes occurs during the period of active growth of the bacteria. Owing to the fact that cell death soon overbalances cell reproduction the number of living bacteria decrease. Because of the good insulating property of the silage there is usually a gradual gain in temperature although the number of living bacteria is not so great. It is reason- able to expect there will be no definite relationship or parallelism between the number of living bacteria and temperature except during the early stages of the process of silage formation. The products of decomposition also tend to accumulate although the number of living bacteria is decreasing; hence there is no parallelism here except dur- ing the initial stages of silage production. On the plates seeded with diluted juice of the fresh corn a great variety of colonies appeared, spreaders, chromogenic forms, small pin point colonies. Plates of samples taken 24 hours later were entirely different. Instead of the varied flora, only one or two types of colonies appeared. The spreaders and the chromogens disappeared and in turn were replaced by great numbers of the small pin point colony bacteria which in the proper medium form large amounts of lactic acid. (28) To gain some idea of the general characteristics of the silage flora, trans- Chart I. — The Relation Between Number of Bacteria. Acidity, Soluble Nitrogen and Temperature 8 Wisconsin Research Bulletin 61 fers of representative colonies were made from each sample and these cultures studied. Almost without exception it was found that the organisms belonged to the general group of lactic acid-forming bacteria. In Chart I are given the curves of total number of bacteria, acid formed, increase in temperature and also increase in soluble nitrogen. As might be expected the number of yeasts normally present on the corn fail to show any decided increase in the silage. Two days after the corn is ensiled, there is a decided falling off in the number of yeast cells; this decrease is especially noticeable after the first five days. The formation of alcohol in silage cannot be accounted for as a product of yeast development. It is no doubt produced by certain groups of the lactic acid bacteria. Counts made by the direct microscopic method showed the presence of so much debris that it was found impossible to make an accurate determination of the number of bacteria. Examination of the mounts proved^valuable as a means of detecting the change in kinds of organ- ism as well as number. Corn juice at the beginning showed many kinds of organisms while the samples after 1 and 2 days until the end of the test showed only a few types. Attempts were made to classify the bac- teria in silage by the use of various sugar media. The large number of organisms found in silage and the difficulty of adequately describing them is well known. It was hoped that the fermentation of various sugars might prove useful in the separation of certain selected groups. Unfortunately the dilutions carried out on the various media were not always high enough to give a measure of the maximum number present in the samples of juice. The results of the dilution tests are given in Tables II and II. These data show the role of the lactic acid bac- teria in corn silage fermentation and will be discussed in detail. During the first 24 hours after ensiling, the number of bacteria in- creased 100 times, while, in the succeeding 24 hours the increase was 1000 fold, reaching the huge number of 100 billion per cubic centimeter of juice. After 7 days the numbers decreased, although even after 132 days there remained 10 million active bacteria per cubic centi- meter. It may be fairly assumed that the majority of all bacteria contained in silage will grow in glucose-yeast-water media. Since the number of bacteria found by means of the xvlose-yeast-water equalled that mea- sured by the glucose medium, it may likewise be concluded that the pentose fermenters comprise the vast majority of bacteria contained in silage (20). That the bacteria counted were actually fermenting the xylose and not merely growing in the yeast water is proved by com-' paring the acid production on yeast water alone with that produced in yeast water plus xylose. The complete data showing the acid pro- duction in yeast water alone have been omitted. The inoculated yeast water rarely gave more than 3.0 cc. and in high dilutions was below 2.0 cc. of O.l N acid for 10 cc. of medium. In Table III it may be seen that the acidity developed in the highest dilution for the xylose medium Table II.— Number of Micro-organisms in Silage Juice at Various Ages Changes In The Making of Silage 9 C. O S7. ocooooooooo ocooooooooo ooooooooooo — COOOCOOOO— cccccoooo COCCCCO— - Xylose yeast water v ooooooooooo ocooooooooo ooooooooooo — CCOOOOOOOO OCOOOOOOO— OOOOOOOO— — OOC— — — — ooc ooooooooooo ooooooooooo ooooooooooo — OOOOOOOOOC +j occccooo— oooooo- — 2 £ oooo— — 3 — coo O ° W) 03 03 S.O -MKin-xci'fflN 'Q — — ro Tt 00 00 or,— Q c/: L-at^OOOO' 10 Wisconsin Research Bulletin 61 Table III. — Production of Acid by Silage Bacteria in Xylose and Glucose Media O. 1 N. Acid in 10 cc. of media Glucose Xylose Sample Yeast Beef Yeast Beef No. water peptone water peptone Low High Low High Low High Low ! High dil. dil. dil. dil. dil. dil. an. dil. cc. cc. cc. cc. cc. cc. cc. cc. . 1 4.0 3.7 0.8 0.6 6.0 3.0 2.6 0.8 2 4.0 2.6 1.2 0.6 6.0 4.2 1.6 1.2 3 4.4 3.4 4.2 5.6 4 5.6 4.8 3.6 2.0 7.2 5.0 6.8 5.2 5 8.0 6.8 4.0 4.0 8.0 8.8 6.4 3.2 6 7.0 6.0 4.2 2.2 7.4 3.4 5.6 3.4 7 6.8 9.0 2.0 3.6 10.0 8.8 6.0 3.2 8 7.2 6.8 3.6 3.6 7.8 7.2 4.6 3.2 9 8.8 5.8 6.0 4.0 7.0 10.6 9.0 3.6 10 11.0 8.4 5.2 4.8 9.8 8.6 6.0 3.2 11 6.8 8.2 2.0 3.0 4.0 8.4 4.0 4.0 12 8.4 8.8 2.0 4.0 5.0 8.2 7.2 3.8 13. . 7.0 4.3 10.9 6.0 Average 6.85 5.22 3.09 2.95 7.18 6.48 5.36 3.16 fell below 4.0 cc. in only two cases and averaged 6.48 cc. It is evi- dent that xylose must have been fermented to produce this quatitity of acid. Determinations of sugar on the fermented cultures, also showed a large part of the xylose to have been destroyed. The titration figures, Table III, show a generally higher acid pro- duction from xylose than from glucose. This result is in harmony with quantitative data previously obtained from the fermentation of glucose by certain pentose-fermenting bacteria. All of the pentose fermenters studied in the previous papers, (8, 21) L. pentoaceticus, L. pentosus and L. arabinosus were found to ferment a larger per- centage of xylose and arabinose than of glucose or other hexoses. A second factor operating to give a low acidity from glucose is the pro- duction of neutral bodies such as ethyl alcohol and carbon dioxide by certain of the pentose fermenters. It is to be observed that higher acidities are developed by the bac- teria found in old silage than by those found in the early stages of the fermentation. In old silage low acid producers are eliminated and a flora is developed which both produces and tolerates high concentrations of acid. A comparison of the figures from low and high dilutions indicates that for both xylose and glucose the amount of silage juice in the ino- culum influences the final acidity reached. This may be due either to a different mixture of bacteria in varying dilutions or to the absence of the high acid producers from the last dilutions. It is not improbable that the production of acid is a symbiotic relationship depending upon the presence of certain mixtures of bacteria. An alteration in this mutual relationship markedly influences the degree of acidity attained. Table IV. — Acid Production in Milk by Silage Bacteria Changes In The Making of Silage II ^ 0(00000 O SHOOOO OThOOOcOrFiMcOOl TfOOOOJOOOOOoO cooooo50t^oi>in oooi©oocooocO'*£o fflxoofi - HnMr . 010 0CO COO c^co ao sz -H M 00 r? in CO l> oo a o The control has been subtracted from each 10 cc. of culture. 12 Wisconsin Research Bulletin 61 The number of acid-forming bacteria in milk is shown in Table IV. The evidence indicates that a high percentage of the organisms commonly found in silage ferment milk. Acid production from milk is especially true during the first week, later the numbers decrease but are present in high numbers until the end of four weeks. Chemical and Physical Changes in the Silo Gases — One of the first chemical changes in the silo is the con- sumption of oxygen and production of carbon dioxide. This change begins immediately after the tissue is placed in the silo and within a few hours the oxygen has disappeared. From Table V and Chart 2, it can be seen that within an hour after ensiling, the gas in the in- terstices of the forage contained 4 per cent of carbon dioxide. In 5 hours the percentage had risen to 17.9 and the oxygen had almost completely disappeared. From this time on, the carbon dioxide in- creased rapidly to a maximum of 68 per cent, 46 hours after the ma- terial was placed in the silo. It then began to decrease and con- tinued to decrease for 131 days. Table V. — Composition of Gas Contained in Silos Sample No. Location in silo Age ot silage Carbon dioxide Oxygen Silo No. 1 1 Bottom Bottom 1 hr. Per cent 4.0 Per cent 13.7 2 5 hrs. 17.9 0.6 3 Bottom 10 hrs. 25.1 0.0 4 Bottom 23 hrs. 45.8 0.0 5 Top Bottom 23 hrs. 35.3 0.0 [ 6 46 hrs. 68.0 0.0 7 Top Bottom 46 hrs. 47.1 0.0 8 3 days 3 days 5 days 5 days 7 days 7 days 11 days 16 days 16 days 23 days 23 days 37 days 37 days 67 days 67 days 131 days 59.0 0.0 9 Top Bottom 40.6 0.0 10 47.4 0.0 11 Top Bottom 32.3 0.0 12 42.2 0.0 13 Top Bottom 27.9 0.0 14 36.4 0.7 15 Bottom 30.0 0.0 16 . Top Bottom 22.9 0.0 17 27.1 0.0 18 Top Bottom 21.7 0.0 19 25.3 0.4 20 Top Bottom 20.4 0.6 21 19.7 0.6 22 Top Bottom 18.6 0.4 23 18.0 0.8 Silo No. I > 1 9 ft. from surface 2 days 2 days 7 days 7 days 15 days 15 days 23 days 23 days 55.6 0.0 2 16 ft. from surface 57.3 0.0 3 5 ft. from surface 28.0 0.5 4 16 ft. from surface 43.2 0.0 5 5 ft. from surface 24.6 1.6 6 13ft. from surface 31 . 1 1 . 1 7 4 ft. from surface 23.7 0.4 8 10 ft. from surface 27.4 0.2 Changes In The Making of Silage 13 A similar increase followed by a decrease in the carbon dioxide con- tent of silage gases has been noted by Neidig (17). In one instance he found a decrease from 87 per cent of carbon dioxide on the third day to 40.5 per cent on the twenty-first day. TablejVI. — The Composition of the Gas at Different Plages in the Silo Sample No. Location in silo Age of silage Carbon dioxide Oxygen Nitrogen 1 2 ft. from surface Days Per cent 20. 1 Per cent 1.3 Per cent 78.6 2 4 ft. from surlace 7 20.1 1.2 78.5 3 10 It. from surface 7 27.9 0.0 72.1 4 20 ft. from surface 7 42.2 0.0 57.8 5 0.5 ft. in, 20 ft. from surface 30 25.0 0.5 74.5 6 4 it. in, 20 ft. from surface 30 26.7 0.3 73.0 7 0.5 ft. in, 10 ft. from surface 37 19.7 0.8 79.5 8 4 ft. in, 10 ft. from surface 37 20.4 0.6 79.0 9 0.5 ft. in, 20 ft. from surface 37 22.7 0.6 76 . 7 10 4 ft. in, 20 ft. from surface 37 25.3 0.4 74.3 11 0.5 ft. in, 10 ft. from surface 44 17.8 1.6 80.6 12 4 ft. in, 10 ft. from surface 44 18.6 0.4 81.0 13 0.5 ft. in, 20 ft. from surface 44 21.9 0.8 77.3 14 4 ft. in, 20 ft. from surface 44 23.4 0.0 i 76.6 As can be seen from Table VI, the percentage of carbon dioxide de- creases at the top and sides of the silo; a condition which shows that the carbon dioxide diffuses outward and upward through the silage. The atmosphere surrounding the silo may be regarded as a vacuum with respect to the carbon dioxide, while to the oxygen, the interior of the silo is a place of greatly reduced pressure. The oxygen does not pene- 14 Wisconsin Research Bulletin 61 trate very far into the silage due to the absorption of this gas by molds and other aerobic forms of life. This oxidation of organic matter explains the high temperatures found near the surface of the silage. Attention is called to the fact that for the first 5 hours oxygen is absorbed faster than carbon dioxide is given off. The sum of the carbon dioxide and oxygen is less than 18 per cent, while the per- centage of oxygen by volume in the atmosphere is about 20.9 per cent. This must mean that some of the oxygen is used for other purposes than oxidation of carbon to carbon dioxide. The result is a negative pressure in the silo. That this negative pressure exists can be easily demonstrated by packing silage into a bottle and closing it with a stopper and tube dipping into mercury. Within a few moments the mercury will rise 1 or 2 inches in the tube due to the absorption of the oxygen. Later it will fall and gas can be seen bubbling through the mercury. Although the silage gases were examined many times for hydrogen and hydrocarbons none was ever found. Cellulose ferment- ing organisms are therefore believed to be absent or inactive in silage fermentations. Temperature Changes. — Next to the changes in the composition of the silo gases, the most immediate effect of ensiling corn forage is an increase in the temperature. A steady rise in temperature took place for 15 days equal to 7°C. near the bottom of the silo and 20° C four feet from the top. For the intermediate depths the increase was be- tween the top and bottom figures. The rise in temperature appears to be inversely proportional to the depth. A difference of almost 10 de- Table VII. — Temperature Changes at Different Depths in the Silo Dept! ti from the surface in feet No. Age 4 7 13 19 j Beginning 24 hours °C 18 0 °c 18.0 oC 18.0 oC 18.0 2 20 5 20 5 20.0 20.0 3 48 hours 23 2 22 0 22 1 21.0 4 60 hours 25.0 24.8 23.0 21.6 5 72 hours 26 0 25.5 23.5 22.0 6 96 hours 28 0 25.0 24.5 22.8 7 5 days 6 days 7 days 8 days 9 days 10 days 11 days 14 days 16 days 20 days 25 days 30 days 40 days 50 days 60 days 65 days 85 days 130 days 30 0 25 5 25.0 23.2 8 31 3 26 0 25.6 23.7 9 33.0 26.5 26.0 24.0 10 34 0 27.0 26.5 24.2 11 35 0 27 2 26.5 24.2 12 35 5 27 2 26.5 24.6 13 35 8 28 0 26 6 24.8 14 36.8 28.8 26.8 25.0 15 37.5 29.0 26.8 25.0 16 37 6 29.2 26.6 24.8 17 37 5 29 2 26.0 24.6 18 37.5 29.0 25.6 24.4 19. . . . 36 6 27 8 24.0 22.8 20 36 4 26.0 20.0 20.5 21 36 3 25. 1 20. 1 19.5 22 36 0 24.5 19.1 18.3 23. . . 35.8 33.4 22 0 15 0 14 0 24 18.0 5.0 7.0 Changes In The Making of Silage 15 grees exists between the 4-foot level and the 7-foot level. It is prob- able that the higher temperature in the first 4 feet is due to molds and aerobic bacteria. Little or no oxygen penetrates beyond this depth. The results of these measurements are given in Table VII and Chart 3. Fermentation Products — Simultaneously with the bacterial examina- tion, determinations were made of the chief fermentation products of the bacteria. A decrease in sugar with the production of alcohol, car- bon dioxide, volatile and non-volatile acids are the chief changes taking place in fermenting silage. In Table VIII are given the data for these determinations together with those for certain other physical and chemical constants. As has been frequently observed, the moisture content increases with the age of the silage. This is due to the respiration of plant cells and to the action of micro-organisms. The rate of increase is greatest for the first two days but continues to rise slowly throughout the fermen- tation period. The specific gravity of the juice from each sample was measured by means of the Westphal balance. At the beginning the specific gravity was 1.037, after 3 days, 1.042 and from then until the end of 140 days there was very little change. Because of the slight variation in the figures the complete data are omitted from the table. Little or no increase in the acids takes place for the first 24 hours. During this time many of the bacteria found on the green tissue are destroyed while the acid producers multiply in great numbers. It re- quires from 24 to 48 hours for a typical silage flora to appear and to begin the formation of fermentation products. The increase in bacteria 16 Wisconsin Research Bulletin 61 go; ® * z « T> O O W > « .Th'O K 3 CC ° 0.0 HZ T. iO *-t OS N © Ol •“■I" ^olOH>LOCOOOOO'-i©© •oicst>mooeoOcOi-ooco gO-HinOrHOrHCOMMOOSO ! >3(MC' ; lCOCOiOiOcOcOcOI>r''r^OO iC^-cOfOOO©OJ00010Tl<0© ■roocoTjt^©in©LO« ^OOOOOOOOhhhh'h COIO-t^^HOCOS-tfOCCO •©OiooiTfcafOiM© gLO©io-^inoomoco 'OOdcOTf^iOcDcO > 00 © CO M 1-H . coo y— 1 O 05 rf k _ H 05iO©'co©t^r^>-i©oo©r-i ^«^iniooor} CL, ^H01C0Tf 4 or Ag. found Percent- age found Ba. or Ag. Theory for malic acid 1 Immature marsh corn Barium Gm 0. 1232 Gm 0.1068 Per cent 51.0 Per cent 51.0 2 Immature marsh corn Silver 0.0984 0.0610 62.0 62.0 3 Immature marsh corn Silver 0.1085 0.0674 62.4 62.0 4 Mature marsh corn Barium 0.5386 0.4670 51.0 51.0 5 Mature marsh corn Silver 0.0580 0.0358 61.7 62.0 6 Mature marsh corn Silver 0.1013 0.0632 62.4 62.0 7 Silage — 3 days old Barium 0.3484 0.3018 50.6 51.0 8 Silage — 3 days old Silver 0.1587 0.0996 62.8 62.0 9 Silage — 3 days old Silver 0.1244 0.0771 62.8 62.0 10 Silage — 3 days old Silver 0.1800 0.1132 62.9 62.0 1 1 Silage — 130 days old Silver 0.1000 0.0599 59.9 62.0 12 Immature upland corn Barium 0.1276 0.1110 51.2 51.2 13 Immature upland corn Silver 0.0720 0.0444 61.7 62.0 Table XI. — Distribution of the Different Forms of Nitrogen in Corn Silage Changes In The Making of Silage c/: • , See cct^t^oct^cr-cCoccCeCTj-' T 3 • D.Z oco;^-ocir;i>in?Dino:occ 5 oc^-ccccoccco • I'' OiiOCCC'JOCr^CTiC'.CCiC'JC'J oc • * 5 CJ t> 00 L 0 « 0 < 0 ^HC 0 C 0 0 C 0 COI> >< x c^ocr^r^i>i>r^iommi-o — o< O t- 00 a o • ”Z .t>t>00Oc0^OCCt^iDT0Cr}‘i> ^(NCOTfincOOr-iOOOXOlM'^CCfO GCc: TiT-iL 005 I>« 005 S 0 :<£!Or} , cD , ^ , T}' — ^(McCCMCNCNKMClrCTf" D . — '.Dcoin^^ocOi-ifOOiNLn^o COt^tdXCC^^CiCOTxt^OiOOi-iCO uo LOOCOOCCC'-^Cai^tC’-rOC — ^i-iCOt^-COt^OCiOO^rfuOOt^t^ '^c^c^deoco^coioicioioiocoiCLC t^-fO'!tcO'^ , cOCi'-Hl>CT:oO’-il>C 5 I> oc -^ 1 •^rotc-Ha>TjTj'iooc^cco;co 5 (M^occMCiOO^^o'-t^cc«cr.^- <.riO 050 >I>t> 00 I>I>C 0 C 0 « 2 i 0 OI> -C'-*-T-ic£>coO'sfoociO'-'C v JcoT}'ia 19 20 Wisconsin Research Bulletin 61 Chart 4. — Relation Between the Destruction of Sugars and the Formation of Acids and Alcohol The results obtained from the corn forage and silage raised the question as to the presence of malic acid in corn during the early stages of growth. On July 22, 1924 samples of young corn grown on upland and marsh soils were taken and analyzed. The marsh corn was about 30 inches high and the upland samples about 40 inches. Malic acid was found to be present in both of these samples. These results indi- cate that malic acid is probably present in corn throughout its period of growth and is not due to the type of soil on which the corn is grown. Forms of Nitrogen in Corn Silage — The changes in the forms of nitrogen during ensiling have been studied ( 1 , 3, 18 , 30) far less ex- tensively than the fermentation of the sugars and consequent produc- tion of acids. The bacteria which are found in silage have a nitrogen as well as a carbon metabolism. Hydrolysis of proteins to proteoses, peptones and amino acids, followed by deaminization of the amino acids and the production of ammonia are biochemical reactions brought about by the silage micro-organisms. The data in Table XI show a progressive increase of soluble nitrogen with the age of the silage. At the beginning the soluble, nitrogen comprises about 15 per cent of the total nitrogen while at the end of the fermentation it amounts to about 45 per cent. The soluble nitrogen consists of about one-half protein nitrogen and one-half peptide, amino, and ammonia nitrogen. Of these three the free amino nitrogen is the principal form and only minute quantities of ammonia and amide nitrogen are present. The Changes In The Making of Silage 21 percentage of distribution of the nitrogen at the beginning and after 145 days of fermentation is as follows : Corn forage Silage 145 days old per cent per cent Insoluble nitrogen 84.0 55.7 Soluble protein nitrogen* 7.6 21.0 Free amino nitrogen 1.9 18.4 Peptide nitrogen 6.2 4.4 Balance 0.3 0.5 100.0 100.0 The total nitrogen of the forage per hundred grams of dry matter was 1.34 gm. and after 145 days in the - silo it decreased to 1.27 gm. showing an apparent loss of 5.2 per cent. If the loss of dry matter, 10 per cent, is considered in the calculation, the actual loss is 14.7 per cent. Losses reported by previous investigators (1, 19, 23, 26) have been of approxi- mately the same order. The data of Table XI are represented graphically in Chart 5. Milligrams Nitrogen in loo Gm. Dry Silage 1200 1100 1000 900 600 700 6 00 500 400 300 200 100 k \ \ 5 O/ul. ■>le Hi a ?r? -- > 7 otaL Sotub >le Ni troge , - -> # # A # able / Y°slL °rptei. n Nit roge> y * Sol 7... — '' > ^ -> Amir 70 Nl iroge n in 5 olubi le — -- non- Protein Nitrogen — l J- 1 -l-l 6 10 12 14 16 16 20 22 24 AGE IN DAY 5 1 40 Chart 5. — Forms cf Nitrogen in Silage at Various Times During Fermentation 2 ? Wisconsin Research Bulletin 61 Loss of Dry Matter — In order to determine the loss of nitrogen and the destruction of pentosans and starch it is necessary to know the loss of dry matter. If this loss is not known it may appear that there is no change or there may be even an apparent gain in the pentosans, starch and other constituents. The water proof bags which had been filled with a known weight of forage and buried in the silo at the time of filling served for this purpose. Feeding of the silage began early in January and the first set of bags were reached January 27, 1923, 132 days after the silo was filled. The bags were taken to the labora- tory, weighed, the contents removed, and the empty bags weighed. It was found that the bag itself had practically the same weight as when it was filled. Since it had not taken up any moisture, it is felt that if there was any movement of silage juices these did not penetrate the bag. The contents of the bag therefore represented the original corn forage free from any additions or removals due to possible downward movement of silage liquid. As a check on the weight of silage obtained by difference the weight of silage was determined by direct weighing. The two sets of figures checked within one ounce. The data are given in Table XII and show an average loss of 10.1 per cent for the first set of bags and 9.8 per cent for the second set. Destruction of Starch — The effect of silage making on the starch content of corn has not been investigated to the same extent as the Table XII. — Loss of Dry Matter in the Formation of Corn Silage No. of bag. Wt. of corn fodder Moisture in fodder Wt. of dry fodder Wt. of silage Moisture in silage Wt. of dry silage Loss of dry matter Series I.— -8 ft. from top Lbs. Per cent Lbs. Lbs. Per cent Lbs. Lbs. Per cent 1 40 40.0 66.2 13.52 2 40 62.4 39.9 66.4 13.41 3 40 62.0 39.2 66.6 13.09 Average 40 62.2 15.12 39.7 66.4 13.34 1.78 10. 1* Series II. — 5 ft. from bottom i 1 40 63.0 39.6 66.0 13.47 2 40 63.3 39.5 66.5 13.23 3 40 62.7 39.7 67.0 13.10 Average 40 63.0 14.80 39.6 66.5 13.27 1.53 9.8* *Corrected for alcohol and acetic acid content of silage which is lost in drying. fermentation of the sugars. Dox and Yoder (4) reported that the starch content remains constant throughout the fermentation process. Their data however, show a slight decrease in the percentage of starch found in samples after fermentation was well under way as compared to the percentage found in the samples taken during the first two days of fermentation. They also neglect the loss of dry matter, which must have been at least 5 per cent. If their data are examined with these Changes In The Making of Silage 23 facts in mind, it is possible to calculate a loss of about 10 per cent of starch. Shaw and Norton (25) found that silage contained a slightly higher percentage of starch than the corn forage from which it was made. This percentage increase could be brought about by a loss of sugars which would enrich the residue with respect to its starch con- Table XIII. — Destruction of Starch in the Formation of Corn Silage Calculated for 100 gm. dry matter Sample Starch in Loss of starch Forage Silage Silage from 100 gm. dry forage Gm. Gm. Gm. Gm. Per cent 8 ft. from top 28.0 22.9 8 ft. from top 27.3 22.8 8 ft. from top 27.1 22.3 Average 27.5 22.7 20.4 7.1 25.9 5 ft. from bottom 27.1 21.5 5 ft. frdm bottom 27.7 21.6 5 it. from bottom 26.6 20.7 Average 27.1 21.3 19.2 7.9 29.2 tent. They give no figures on the loss of dry matter so it is impossible to decide whether there was an actual destruction of starch or not. The data given in Table XIII show both a percentage and an actual decrease in the starch content of the silage over the corn forage. The average decrease for the top and bottom series of bags is 20 per cent, by direct comparison of the forage and silage, and 27 per cent if the loss of dry matter is considered in making the Calculation. The data here given are on the composite sample for the three bags at each level. Three sub-samples were taken of each composite sample and were analyzed in duplicate so that the average figure of 27 per cent loss represents 24 different determinations. The loss is surprisingly high and should be accepted only tentatively until sufficient data have been accumulated to determine whether such large losses are usual or ex- ceptional. Destruction of Pentosans — Annett and Russell (2) reported a loss of 32 per cent of furfural-yielding substances in ensiling rather immature corn. Shaw and Norton’s data show a decrease in the furfural-yielding substance in the silage as compared with the corn forage. Assuming a loss of 10 per cent dry matter in their fermentation, the average loss of furfural for two years experiments is 17 per cent. In a previous pa- per from this station it was shown that from 15 to 20 per cent of the pentosans in corn forage are destroyed by ensiling. The results obtained in the present investigation are even higher, ranging from 24 Wisconsin Research Bulletin 61 20 to 30 per cent. The higher figures may be due to a more extensive fermentation and a longer period of ensiling. The data are given in Table XIV and show a decrease in pentosans from 19.6 per cent in the corn forage to 15 to 16 per cent in the silage. The loss in dry matter in these samples ranged around 10 per cent. Table XIV. — Destruction of Pentosans in the Formation of Corn Silage Calculated for 100 gms. of dry matter Sample Age of silage Pento- sans in forage Pento- sans in silage Pentosans in Silage from 100 gm. dry forage* Loss of Pentosans Series of 1921 Days Gm. Gm. Gm. Gm. Per cent Top silage, uninoculated . 38 19.6 16.3 14.7** 4.9 25.0 Top silage, inoculated . . . 38 19.6 16.2 14.6** 5.0 25.5 Top silage, inoculated . . . Bottom silage, uninocu- 38 19.6 15.2 13.7** 5.9 30.1 lated 139 19.6 15.3 14.6 5.0 25.5 Bottom silage, inoculated. 139 19.6 15.7 15.3 4.3 21.9 Bottom silage, inoculated. Series of 1922 139 19.6 16.4 15.7 3.9 20.0 Top silage, uninoculated . Bottom silage, uninocu- 132 19.5 16.6 14.6 4.9 '25.1 lated 146 19.5 | 16.2 1 14.5 5.0 25.6 ^Obtained from column three by correcting for loss of dry matter. **In calculating these figures a 10 per cent loss of dry matter was assumed. Allowing for this additional loss the actual loss of pentosans as given in column five is found to be from 4 to 5 gm. per hundred grams of dry corn fodder. The above figures are from two years experiments and probably represent the usual loss in silage making. It is probable that alcohol and acid originate from the fermentation of pentosans as well as from sugars. Effect of Inoculation on the Production of Silage The possibility of reducing the losses and improving the quality of. silage by inoculating the corn forage has been considered by a number of investigators. The results obtained can not be said to prove definitely that inoculation influence's the quality of the silage or the kind of the fermentation products formed. To prove that the added bacteria func- tion in the process is a difficult matter owing to the fact that their pre- sence may be masked by the abundant flora which is normally found in corn silage. If the added bacteria play any considerable part in the fermentatton, their presence should be indicated both by their numbers and by their characteristic fermentation products. The latter is probably the more reliable index and should be particularly evident in the early stages of the fermentation. Changes In The Maki'ng of Silage 25 Experimental Silage Made in Milk Bottles — The first work on the effect of inoculation was done in the fall of 1919, when selected cultures of bacteria which had been isolated from corn silage, were used to inoculate both sterilized and unsterilized corn forage. The forage was inoculated with the culture and then packed into milk bottles closed with a one-hole rubber stopper through which passed a bent glass tube. The free end of the glass tube was sealed by insert- ing in a test tube containing 2 or 3 inches of mercury. The bottles were incubated for 10 days at 27° C. and during this time a strong evolution of gas was noted. When opened all except the sterilized, un- inoculated controls had about the same odor and taste as that of nor- mal silage. The contents of each bottle was analyzed and the data ob- tained are given in Table XV. A comparison of the data from the un- sterilized silage with and without inoculation shows an increase of from 20 to 120 per cent in the fermentation products as a result of inoculating the forage. The sterilized corn was changed into a silage which contained large quantities of the fermentation products normally found in silage. It is evident that added bacteria can produce silage without the aid of plant enzymes, and even when plant enzymes and the natural bacterial flora are present, can markedly influence the quantity of products formed. Experimental Silage in Large Containers — In 1920 the effect of in- oculation was tried on a larger scale. In one series 50 gallon barrels were used and in another series two metal tanks 8 feet high and 4 feet in diameter were filled with inoculated corn. These experimental silos were filled at the same time as one of the large silos was being filled. The corn was' in excellent condition for silage making; the ears were well matured but the leaves and stalks were still green. Samples of silage were removed for analysis at three different times during the fermentation. Numbers of bacteria and fermentation prod- ucts were determined each time. The complete data have already been published and only the last analyses are given in Table XV. In general the inoculated silage showed a more vigorous and sustained fermentation than the uninoculated controls. The differences in the numbers of bacteria were most marked after 12 days, showing a much greater persistence of bacteria in the inoculated samples. The increase in bacterial count is greater in the silage inoculated with a mixed culture than in that inoculated with a single strain of bacteria. The large number of bacteria in the inoculated silage is associated with an increase in fermentation products. These products are char- acteristic of the added bacteria. The silage inoculated with L. pento- accticus shows a pronounced increase in acetic acid and ethyl alcohol. These products are the compounds characteristic of the added bacteria as has been shown by fermentation of various sugars. Judging from the ap- pearance of the plates and from the fermentation tests with lactose and xylose, it was concluded that the S. lactis type persisted and acted Table XV. — Comparison of Silage Made from Corn With and Without Inoculation Wisconsin Research Bulletin 61 O « a £ 2 1 I "CCt' OMfi OtOrH l>-*fCM i}<00eC ' CM CO T}< I> •— i o t^- ©COOm OOOiO rfcOOC • MCOC to cm co ■ csi>m ■ 00O5CC c.£ s O o « 2 ” CO 05 00 lO lOOS 05 CM CO 05 in I >00 05 05 occomo 05-^0 0 COI'- 05 05 m ©men© inmt> I I rf u o > o c, O 03 4) C «-> hHOOOC! hJ t /3 CO CO ^ u CO CO 8 l Joo si si TJn: 05 03 03 3 ' ‘ 03 03 p p . I hi O 03 C C c o o ° c c c u 3 3 CO O 03 9)00 C 3 O O O O fflfflCGCQ 03 03 03 03 _V S- l- U l- O t* t* >_ t- c c * od' co cg a c 3 a ^cqpqcqcq hh ©©t^occs O’ Changes In The Making of Silage 27 only during the first days of the fermentation and that the L. pentoace- ticus type is the predominant type during the last stages of the fermenta- tion. Effect of Inoculation under Conditions Existing in the Silos — In 1921, the effect of inoculation was studied under silo conditions by inoculating corn forage, packing this in^ water proof bags, and burying these in the silo at' the time of filling. The inoculum consisted of a 1 per cent yeast-water solution in which the bacteria had been grown for three days. 200 cc. of this culture was sprinkled on the corn forage as it was placed in the bag and the contents of each bag thoroughly mixed after each sprinkling. An equal quantity of water was added to the control bags. The same water tight bags used in the 1922 work were employed and 40 lbs. of cut corn placed in each bag. Three sets of bags were placed in the silo; one set 5 ft. from the bottom, a second set near the middle and a third set 5 ft. from the top. The corn was in excellent condition for ensiling when the first set of bags was placed in the silo but rain that night and for several days thereafter made it impossible to haul from the low land where the first lot of corn was obtained. After 10 days it was decided to fill the rest of the silo with corn from higher land. This corn was extremely dry and water had to be added during the cutting. The quantity of water necessary to bring the moisture content of the silage up to 69 per cent was measured by means of a water meter connected in the hose line. Moisture tests on the wetted forage showed that it had not absorbed sufficient water to bring the moisture content up to the desired percentage. Matured and partly dried plant tissue does not readily absorb water. It is highly probable that much poor silage re- suh.s from the erroneous idea that addition of water to dry corn gives as good results as ensiling corn of the proper maturity. The first or uninoculated bag in each series served as a control. The second bag was inoculated with a culture of L. pentosus No. 124-2. T liis organism produces lactic acid almost entirely from the hexoses, and acetic and lactic acids from the pentoses. The third bag was inoculated with a mixture of L. arabinosus, No. 102 and L. pentoaceti- cus No. 41-11. The first culture is much like 124-2 but does not fer- ment xylose ; the second forms alcohol and carbon dioxide from the al- dohexoses. and like 124-2 and 102 acetic and lactic acids from pentoses. If the added bacteria dominated the fermentation, lactic acid should be high and alcohol and acetic acid low in bag No. 2. In bag No. 3 the conditions would In- the same as in No. 2 if culture 102 predominated during the fermentation, and high in alcohol if No. 41-11 were the determining factor. The data are given in Table XVI and show a somewhat higher acidity for the inoculated than for the uninoculated silage, in the case of six out of seven bags. The influence of cultures 142-2 and 102 is dis- tinctly shown in the four cases of the top and middle series. 28 Wisconsin Research Bulletin 61 CM 05 < W X I W O < c 7i % ec o U fe- o C/3 h u u Q O w K C-l £ O 2 O < ►4 D u o fe< o H r 1 w fe- te, W i X w j CQ < Changes In The Making of Silage 29 The data for alcohol show no outstanding effect of inoculation in the first series, but in the second series this product is low, as would be expected from cultures 124-2 and 102. The third bag in the bottom group contained a silage somewhat un- usual in composition. The acetic acid and the ethyl alcohol were unus- ually high, and the lactic acid particularly low. These data suggest a secondary fermentation of lactic acid, with the production of acetic acid and possibly ethyl alcohol. Culture 41-11 has been found to pro- duce such a change in old fermentations. After all the sugar has been destroyed, it attacks the lactic acid and forms acetic acid and carbon dioxide. The evidence strongly indicates that this had taken place. This silage was also examined for esters and aldehydes. The acetic acid in the form of an ester, per 100 gm. of dry silage, amounted to 0.120, 0.230 and 0.320. gm. for bags Nos. 1, 2 and 3 respectively. A distinct color test for aldehyde was obtained from all three silages by the Schiff and Lewin tests. Since the latter test is almost specific for acetaldehyde, being given by only a few aldehydes other than acetalde- hyde, it may be assumed that this compound was present. A quantita- tive determination by the Ripper von-Fiirth method gave from 5 to 10 mg. per 100 gm. of dry silage. Iodoform was formed when the distillate was tested by Goodwin’s modification of the Messinger method for acetone. Since acetaldehyde forms iodoform under these conditions, it is not possible to say whether acetone was present or not but the presence of an abundance of starch in silage would be conducive to the growth of acetone-producing bac- teria. The color, odor and flavor of the different samples were judged by a number of persons familiar with silage, but no decided difference in quality was ‘found. The first series of samples was too dry to be called good silage. The second series was normal in all respects, with the sample inoculated with cultures 41-11 and 102 slightly superior to the other two. In the third series the control and the 124-2 silage were both good but the silage inoculated with the mixed culture had a sweet- ish ethereal odor not found in good silage. As already mentioned the analysis showed the presence of considerable quantities of ester. Summary 1. — The first noticeable change in the making of silage is in the com- position of the «ilo gases. The oxygen disappears in 4 to 5 hours. The carbon dioxide increases rapidly for about 48 hours when it comprises from 60 to 70 per cent of the silo gases. After this time it begins to fall and continues to decrease for several months. No trace of hydro- gen, methane or other hydrocarbons is found in the silage gases. 2. — The second most immediate effect is an increase in temperature amounting to 7°C. near the bottom and 20° C, 4 ft, from the top of the silo. The temperature continues to increase for 15 days and then falls off, but continues at a high level for 60 to 70 days. 30 Wisconsin Research Bulletin 61 3. — Fermentation products appear in marked quantities simultaneous- ly with the appearance of large numbers of bacteria ; from 24 to 48 hours after ensiling. It is therefore probable that the bacteria are the chief agents involved in the production of ethyl alcohol, acetic and lactic acids. These compounds originate mainly from the sugars, but a marked destruction of pentosans and starch indicates that the latter substances are also in part converted into fermentation products. 4. — Almost coincident with the increase in acid, there is a decided change in the bacterial flora. Within 12 to 24 hours after the corn is cut and placed in the silo there is a vast gain in the total number of bacteria, chiefly the lactic acid organisms. The forms commonly pre- valent on green plant tissue, the chromogenic bacteria, the slow growing non-acid forming rods and cocci soon disappear and in turn are replaced by the high acid-formers. The conditions for the growth of the lactic acid bacteria in the silo are excellent, hence they reach almost unbelievable numbers — frequently hundreds or thousands of millions in one cubic centimeter of the juice. This huge mass of micro-organ- isms exists for only a short time. Owing to unfavorable conditions, such as the accumulation of products of growth, they begin to die off. Although the evidence is not conclusive, it is probable that the great mass of bacteria in silage is a result of the successive growth of dif- ferent groups of bacteria rather than the persistance of one group over a period of several weeks. Of the various groups of aciduric bacteria present in fermenting silage none is present in greater numbers and is more active than the pentose-fermenters. Aside from the breaking down of pentoses, the majority of these organisms play an important role in the formation of alcohol. These so-called mannitol-forming bacteria are no doubt the active agents in the production of ethyl alcohol. At the time of filling there were approximately 500,000 yeast cells per cc., two days later this number decreased to less than 35,000 and remained at less than 10,000 throughout the entire experiment. 5. — Approximately 10 per cent of the dry matter, 25 per cent of the pentosans and 25 per cent of the starch contained in the corn forage are destroyed as a result of ensiling for 4 months. 6. — Silage can be made without the aid of the living plant cell. Sterilized corn inoculated with L. pentoaceticus bacteria produced a silage similar in composition to that of normal silage. 7. — Inoculation of corn forage with certain bacteria produces a more vigorous and sustained fermentation. The effect of inoculation is seen in an increase in the number of bacteria and in the quantity of fer- mentation products. These products are the same as those produced from various sugars by the added bacteria. Under field conditions, the quality of naturally fermented corn is usually so good that the cost and extra time required for inoculation hardly seems desirable. Under conditions where it is not possible to secure corn of even ripeness, it may be found advantageous to use artificial inoculation. Additional work on this phase of the silage study should be carried out before a definite statement can be made. LITERATURE CITED (1) Amos, A., and Woodman, H. E. 1922 An Investigation into the Changes Which Occur During the Ensilage of Oats and Tares. In Jour. Agr. Sci. v. 12, Part 4, p. 337-362. (2) Annett, H. E., and Russell, E. J. 1908 The Composition of Green Maize and of the Silage Pro- duced Therefrom. In Jour. Agr. Sci., v. 2, Part 4, p. 382-391. (3) Barthel, C. 1921 Vid ensilageberedning forekommande jasningar. In Kgl. Land, Akad. Handl. Tidskrift., v. 60 p. 92. (4) Dox, A. W., and Yoder, L. 1920 Influence of Fermentation on the Starch Content of Ex- perimental Silage. In Jour, Agr. Research, v. 19, p. 173-179. (5) Edin., H. and Sandberg, E. 1922 Om Ensilering och Ensilage. Orienterande Litteraturstudier och Forsok. In Meddelande No. 221 fron Centralanstalten forsoksvasendet pa jorbruksomradet Husdjursavdenlingen No. 33. Bakteriologiska avdelningen, No. 26, 90 pages. (6) Esten, W. M., and Mason, C. J. 1912 Silage Fermentation. In Conn. Agr. Exp. Sta. Bui. 70, 40 p. (7) Folin, O., and Wu, H. 1919 A System of Blood Analysis. In Jour. Biol. Chem., v. 38, p. 81-110. (8) Fred, E. B., Peterson. W. H. and Anderson, J. A. 1921 The Characteristics of Certain Pentose-Destroying Bacteria. Especially as Concerns Their Action on Arabinose and Xylose In Jour. Biol. Chem. v. 48. p. 385. (9) Fiirth, O., and Charnass, D. 1910 tiber die Quantitative Bestimmung der Milchsaure durch Ermittlung der Abspaltharen Aldehydmenge. In Biochem. Z., v. 26. p. 199-230. (10) Gorini, C. 1919 Studi sui silo Lattici in Base Alla Fisiologia Microbica. In Reale Institute Lombardo di Scienze E. Lettre., v. 53, p. 192-205. (11) Hawk, P. B. 1921 Practical/ Physiological Chemistry. P. Blakiston & Co. Philadelphia. 7th Edition p. 519. (12) Hunter, C. A. 1921 Bacteriological and Chemical Studies of Different Kinds of Silage. In Jour. Agr. Research, v. 21, 10, p. 767-789. (13) Hunter, O. W. 1917 Microorganisms and Heat Production in Silage Fermenta- tion. In Journ. Agr. Research, v. 10, No. 2, p. 75-83. (14) Krober, E. 1901 Untersuchungen iiber die Pentosanbestimmungen Miftelst der Salzsaure Phloroglucinmethode Nebst Einige Anwen- dungen. In Jour. Landw., v. 48, Heft 4, p. 357-384. (15) Kuchler, L. F. 1923 Electro Silage in Germany. In. Int. Rev. Sci. and Practice of Agr., New Series, v. 1, No. 4, p. 857-876. (16) Lamb, A. R. 1917. The Relative Influence of Microorganisms and Plant En- zymes on Corn Silage Fermentations. In Iowa Agr. Exp. Sta, Bui. 40, 20 p. (17) Neidig, P. E. 1914 Chemical Changes During Silage Formation. In Iowa Agr. Sta., Research Bui., No. 16, 22 p. (18) Neidig, R. E., and Snyder, R. S. 1921 The Application of the Van Slyke Method to Hydrolyzed Protein Extracts of Silage Crops. In Jour. Amer. Cherrt. Soc., v. p. 951-959. (19) Perkins, A. E. 1923 Losses 'and Exchanges of Materials during the Storage of Corn as Silage. In Ohio, Agr. Exp. Sta. Bui. 370, 16 p. (20) Peterson, W. H., and Fred, E. B. 1920 The Role of Pentose-Fermenting Bacteria in the Production of Corn Silage. In Jour. Biol. Chem., v. 41, p. 181-186. (21) Peterson, W. H., Fred, E. B., and Verhulst, J. H. 1921 The Destruction of Pentosans in the Formation of Silage. In Jour. Biol. Chem., v. 46, p. 329-338. (22) Peterson, W. H., and Fred, E. B. 1920 The Fermentation of Glucose, Galactose and Mannose by Lactobacillus Pentoaceticus, N. Sp. In Jour. Biol. Chem., v. 42, p. 273. (23) Ragsdale, A. C., and Turner, C. W. 1924 Silage Investigations. Loss of Nutrients in the Silo and During the Field Curing of Corn. In Mo. Agr. Exp. Sta., Re- search Bui. 65, 10 p. (24) Russell, E. J. 1908 The Chemical Changes Taking Place During the Ensilage of Maize. In Jour. Agr. Sci., v. 2, Part 4, p. 392-410. (25) Shaw, R. H., and Norton, R. P. ,. 1920 A Comparative Study of Corn Silage in Concrete and Stave Silos. In Jour. Diary Sci. v. 3, p. 300-307. (26) Shaw, R. H., Wright, P. A., and Deysher, E. F. 1921 Nitrogen and Other Losses During the Ensiling of Corn. In U. S. Dept. Agr. Bur. Animal Indus., Bui. 953, 16 p. (27) Shaffer, P. A., Hartmann, A. F. 1921 The Iodometric Determination of Copper and Its Use in Sugar Analysis. In Jour. Biol. Chem., 45, p. 349-390. (28) Sherman, J. M. 1916 A Contribution to the Bacteriology of Silage. In Jour. Bact. v. 1, p. 445-452. (29) Van Slyke, D. D. 1912 The Quantitative Determination of Aliphatic Amino Groups. In. Jour. Biol. Chem., v. 12, p. 275-284. ( j 0J Woodman, H. E., and Ambs, A. 1924 Further Investigations into the Changes Which Occur Dur- ing the Ensiling of a Green Crop. In Jour. Agr. Sci., v. 14 Part 1, p. 99-113. Research Bulletin 62 August, 1925 Experiments on the Control of Wildfire of Tobacco OCT JAMES JOHNSON and HERBERT F. MURWIN Agricultural Experiment Station of the University of Wisconsin Madison CONTENTS Introduction . 1 Summary of Earlier Work ~ 1 Overwintering Studies 3 Dissemination Studies 8 Spread of Wildfire in the Field 8 Seed Bed Infestation 10 Dusting and Spraying Experiments 11 Seed Disinfection 14 Loss of Virulence 17 The Wildfire Toxin 19 Practical Considerations 20 Summary 23 Literature Cited „ 35 Experiments on the Control of Wildfire of Tobacco' T HE CONTROL of the wildfire disease of tobacco caused by Bacterium tabacum (Wolf and Foster) has been the subject of considerable investigation since the outbreak of the disease in North Carolina in 1917 (14). The outstanding observation, bearing on control, has been the fact that the disease originates in the seed bed and that practically all cases of field infections are traceable to this source. The prevention of seed bed infection is, therefore, the most logical aim of all methods of control. This naturally involves : first, the determination of how or on what materials the causal organism lives over winter or from one crop to the next; and second, methods of pre- venting such infected material from being introduced into the seed beds. Once seed bed infection occurs and is discovered, the grower must choose between discarding the infected plant beds entirely or taking a risk in using some or all of the plants, relying on unfavorable weather conditions to prevent further serious spread of the disease. This latter method is economically hazardous, as it is likely that the disease may prove disastrous to a crop if proper weather conditions for the dis- semination and the development of the disease occur. Precautions to prevent dissemination in the field are of doubtful value as a means of control; their effectiveness is at least very limited, and probably more often they are effective only under relatively unfavorable weather con- ditions for the development of disease. The investigations reported in this bulletin are consequently mainly concerned with a study of the factors which may account for seed bed infection, together with methods of preventing such infection. The practical conclusions arrived at are also to a considerable extent in- fluenced by several years of observational studies made during field surveys. Summary of Earlier Work The control of tobacco wildfire has received some experimental at- tention in most of the tobacco districts in which it has occurred. While some difference of opinion exists as to the relative importance of the methods of overwintering of the causal organism, practically all in- vestigators agree that the causal organisms may survive from one crop to the next on infected and cured tobacco leaf, except that in flue- cured tobacco sufficient heat may be used to kill the organism. The subsequent dissemination of this infective material to the seed beds Cooperative experiments with Office of Tobacco Investigations, Bureau of Plant Industry, I’nited States Department of Agriculture. 2 Wisconsin Research Bulletin 62 may naturally occur in several ways, the most unusual of which has been announced by Valleau and Hubbard (13) who claim that the wildfire organism is commonly transmitted through the spitting of tobacco juice into the seed beds. Wolf (15) and Fromme and Wingard (5) were first to point out the possibility of overwintering on seed and introduced the formalin and corrosive sublimate seed treatments respectively as control measures for tobacco wildfire. The importance of overwintering on seed in the Connecticut Valley has been questioned by Anderson and Chapman (1) and Clinton and McCormick (3). Similarly, overwintering in soil has been suggested by the earlier workers, but this again has been questioned by more recent observa- tions and experiments. Information concerning the possibility of overwintering of the wild- fire organism on seed bed covers (cloth and sash) and frames is espe- cially meager. The possibility has been recognized, however, and re- ported in some cases as occurring (15). Tobacco stems (leaf-midribs) both in commercial fertilizers and as untreated fertilizer material have been held responsible, by observa- tion, for some cases of overwintering. This seems least likely in the case of the manufactured fertilizers containing stems where heat treat- ment is used (15). Untreated stems and stalks, since they usually carry leaf fragment^ which may naturally be infected, are probable overwin- tering carriers as pointed out by Anderson and Chapman (1). Experimental evidence on the actual dissemination of the wildfire organism is small and fragmentary. Observational evidence is abun- dant but rarely convincing. Since almost any material which has been exposed so as to carry the causal organism physically may conceivably carry it from place to place, this subject is not a very fruitful one for satisfactory speculation or experimentation. It has been suggested by various workers that long distance dissem- ination may occur most often through transportation of infected seed, plants, or commercial tobaccos and by dry winds. With respect to transmission of the disease from plant to plant, in seed beds and in the fields, all investigators agree on the effectiveness of rain, especially when accompanied by strong wind. Heavy storms and hail which in- jure the leaf surface are especially favorable to subsequent heavy in- fections as well as for dissemination . The control of wildfire in the seed beds by dusting or spraying fre- quently with copper-lime dusts or Bordeaux mixture has been rec- ommended by workers in the Connecticut Valley. When properly ap- plied it is claimed to be an effective control measure. This method has not been generally adopted outside of New England and some question as to its value has already been raised in our work (8). Dust- ing and spraying in the field has received some attention by other workers (12) with negative results. Since the work reported in this paper had been practically com- pleted, Anderson (2) has published his results on overwintering of Control of Wildfire of Tobacco 3 tobacco wildfire in New England. His results indicate that the bac- teria winter most successfully in situations where they are not sub- jected to keen competition from the growth of other organisms — prin- cipally in fairly dry situations — and that they winter least successfully under conditions moist enough for competing organisms to grow. He concludes the wildfire organism may overwinter on cured leaves in the barn, plants standing in the field, on boards, sash, and dry frag- ments of seed pods, but that overwintering in leaves exposed to decay or in the soil is least likely. Overwintering Studies The overwintering experiments were designed to determine how long and under what conditions the wildfire organism is most likely to sur- vive the period during which its host plants normally can not be the source of its propagation. The main tests have been made with artifi- cially infested materials which are most likely to be concerned with overwintering and seed bed infection. These have been stored under different conditions in most instances, and tested from time to time as to their ability to yield infection when placed in contact with young tobacco plants. It has been assumed that the application of infected material to a unit area, in many cases hundreds of times greater in quantity than that which would occur under natural conditions, reduces the errors which might result from working with only a relatively small amount' of material. Conditions for infection have been made as ideal as possible both by wounding the plants and by maintaining favorable environmental conditions. Considerable variation in this condition is evident, however, from the results. Tests were made soon after the materials concerned were infested and before storing away in all cases to make certain that the causal organism was pathogenic at the start of the test. The results are, therefore, believed to be reliable from the experimental standpoint. From a practical standpoint we have also tested out materials supposedly infected naturally, and made a considerable number of field observations, and these factors are also taken into consideration in drawing final conclusions. In the 1922 experiments artificial applications to seed, boards, cloth, soil, etc. were made! with both pure cultures of the organism and with the juice extracted from badly infected green leaves. Two different sets of applications were made known as Series I and Series II. These materials were divided up into separate portions each suitable for one test. It was planned to store one-half of this material out-of-doors in the winter months, but this was not done in some instances because the organisms were apparently dead on those materials most commonly out-of-doors in winter, before the winter months arrived. The cured leaf material was cured under normal conditions in the shed, and the buried leaves were, of course, outside all winter. The 1923 materials were inoculated artificially with dried crushed leaves, for the reason, that this would seem from our 1922 experiments 4 Wisconsin Research Bulletin 62 to offer the best opportunity for the persistence of the causal organism over winter. Part of this material was stored at room temperature and part in a weather-instrument chamber out of doors where the material was protected from rainfall. Inoculations with these materials have been tried in several ways. Frequently they were made by scrubbing or washing of the materials in a small amount of water and making fifty wound inoculations on individual plants in pots with the washings. In other cases platings on agar were made from the materials and wildfire-like colonies used for inoculation. More reliable results are obtained by placing the materials directly upon young vigorously growing seedlings in “flats” after wounding them. The flats were then well watered and kept covered with paper for one or two days, keeping the plants and paper moist in the meantime. The infested materials were removed from the flats three or four days after the inoculation was made. Our exper- iments have led us to question any conclusions based on negative re- sults from inoculating individual plants with material in which the causal organism is not abundant and is in a latent state, even if such plants are vigorous, wounded and placed under good environmental con- ditions. Seedling inoculations in the greenhouse in which at least 100 plants are involved seems the most reliable test. Inoculations in out- of-door sections of seed beds are not apparently as reliable on account of the danger of dissemination of the organism from section to section, and less certainty in the control of the environmental conditions. Prac- tically all of our results are based on greenhouse tests. The first series of experiments were started in the midsummei of 1922, for the purpose of comparing the survival of the wildfire organism on or in seed, soil, cloth, boards, and dried, naturally infected leaves, cured naturally infected leaves, and green leaves buried about 4 inches in the soil, without direct contact with the soil and with mixtures of soil in proportions of 1 to 5 and 1 to 10. In addition the watery ex- tract from green leaves and the pure culture suspension used for in- festing the seed, soil, etc. was saved for comparative tests, as was the dried green leaf pulp from which the green leaf extract was made. An attempt has been made to present the data from inoculations made with these infested materials in condensed form. The percentages of infection given are not comparative throughout for the reason that different methods of inoculation were used in some cases and because of the variable conditions for infection which cannot be avoided. It is also to be expected that the dilution of the suspension of organisms recovered from the infested materials naturally varies greatly. Within certain limits, however, the percentages are believed significant and to these attention will be called. The principal value in the results, how- ever, lies in the outcome as to whether infection was or was not ob- tained after repeated trials. In this respect the results are believed to be significant to a high degree. Attention has already been called to the fact that the number of organisms involved in these tests are prob- ably infinitely greater than would be likely to occur under normal con- Control of Wildfire of Tobacco ditions, so that the small amount of material used as units (10 grams, seed, 1 square foot of cloth, 16 square inches of boards, etc.) are compara to a much larger quantity of these materials under practical condit : From Table I it seems evident that the wildfire organism c vive but a comparatively short period in liquids exposed t contamination, and that its limit of survival on such materi soil and dried green-leaf-plup, cloth, and boards is only two months, under the conditions of this experiment, hand, on tobacco seed and on dried and cured leaves still alive after nine months. It is interesting to note the comparative survi There is some indication that seed tends to influence on the wildfire organism and that deleterious action. This suggestion is bas behavior in the two cases between the and the green-leaf inoculum. In subse relationships, however, we failed to A second series of tests with Series I was started in the ear eously throughout the fall a respects similar. The data II. The results agree garding overwintering pure-cultures or extr iective powers, w liquid extract f after only a tained infec included * were n befor 6 Wisconsin Research Bulletin 62 leaves crushed or powdered. These were applied in the fall of 1923 V dipping the various materials into a water suspension of the infected ie after which they were rapidly dried and stored under the desired ions. previous season’s experience indicated that the most reliable jld be obtained by direct inoculation to flats (about 22"xl4") everal hundred young and vigorously growing plants, prac- lts for the winter of 1923-24 are based upon this method, the experiment, 100 plants were pulled at random from tal number of lesions on these counted. Flats showing .ounts were carefully searched for any single lesion n most cases the results given are averages of with these materials, as shown in Table III, ganism survived most successfully on dried soil, or on dry stalks ; apparently not so ' J e poorly on cloth and in cured leaves, ’st soil and rotting leaves. The var- oubt in part due to differences in se, although effort was made to some quantitative significance, e causal organism survives ider similar environmental at the organism sur- temperatures than Considerable vari- he condition of that dried ter either such as moist lort Control of Wildfire of Tobacco 7 vational and experimental evidence is offered to keep the matter in doubt. The experiments here have repeatedly indicated that the wildfire bac- teria cannot survive more than a month in ordinary moist loam soil. Some difference may exist in different soils in this respect, but over- wintering of any organisms in soil is very doubtful except as dry in- fected leaf tissue is lodged in dry soil, and does not become intimately associated with it at any time for even short periods. This condition may occur in tobacco sheds or other protected places where the soil remains dry. Table V shows how well the bacteria survive in air-dry soil or in sterile soil whether wet, moist or dry, as compared with soil kept moist, or remaining moist for only a sufficient time after infesta- tion to permit drying. The readiness with which the wildfire organ- ism overwinters in sterile soil or dry soil as compared with unsterilized moist soil seems to be a good basis for the assumption that over- wintering is largely dependent upon competition with other organisms as already has been suggested by Anderson (2). To be sure, this ex- planation seems to account for the comparatively rapid deterioration of the wildfire organism in contaminated liquids, rotting leaves and in ordinary moist soil as compared with otherwise sterile or dry soil. On the other hand, when comparing the persistence of the organism on seed and dried or cured leaves with its persistence on cloth and boards under similar moisture conditions, it does not seem to satisfy the re- quirements wholly. Although this whole matter requires further veri- fication, we are inclined to include in the overwintering requirements the protective action of certain materials, generally host tissue, and perhaps the absence of injurious substances not commonly considered as such. On the other hand, as will be shown later, the wildfire or- ganism may deteriorate more rapidly in pure culture than in the dor- mant condition on seed, in which case competing organisms do not ex- plain their death or loss of pathogenicity. The effectiveness of decay in destroying the wildfire organism is more clearly shown in Table VI, but, on the other hand, conditions favorable for continued decay do not seem to be required for ultimate destruction of the parasite. Various miscellaneous experiments have been conducted with over- wintering which will not be presented in detail except to say that thus far infection has been obtained from air-dried leaves after eighteen months, from cured leaves after fifteen months and from artificially in- fected seed after twenty months, although in some cases the period of longevity of the organism on these materials has apparently been considerably less. It seemed likely to us that if other plants were subject to infection by the wildfire organism these might also prove to be an overwintering agent. To test out this probability a considerable number of other plants (common garden and field crops and common weeds) were arti- ficially inoculated. Most plants tried were found to be subject to the disease when succulent young plants were inoculated under favorable 8 Wisconsin Research Bulletin 62 environmental conditions. The results of this phase of the work have already been published (10). No evidence, however, has been secured that sufficient infection occurs in nature on other hosts than tobacco to warrant the belief that they ever play a part in overwintering, nevertheless it seems worth while for investigators of this disease to be on the look-out for evidence of such cases. Dissemination Studies The question of dissemination of this disease involves a considerable number of problems, of which some are now open only to speculation while others are apparently more likely to be solved by observational than by experimental data. The problems involved are in some respects distinct, since they in- volve long distance dispersal, dispersal to adjoining districts, dispersal from farm to farm, spread from plant to plant in the field, and in the seed bed, as well as the original source of infection of the seed bed. Wildfire apparently spread from North Carolina to thirteen widely sep- arated tobacco growing states east of the Mississippi River in five years. At present there can be speculation only as to the agency of dispersal, since many might conceivably be involved. Although pre- liminary experiments indicate that active fermentation may destroy the causal organism, all portions of the tobacco leaf do not ferment actively. Some experiments indicate, for instance, that dry infected leaves could withstand a temperature of 100° F. for five days, although moist cured leaves could carry the organism only one to two days at this temper- ature. It seems likely, therefore, that commercial tobaccos of certain kinds may be a common long distance dispersal agent, since the or- ganism may quite likely survive two years in tobacco leaf tissue. Dry wind storms may readily carry infested material for long distances and infected seed and plants may be involved in special cases. All of these, excepting wind dispersal, seem to have been excluded in certain cases of epiphytotics which have been observed. The spread from farm to farm within a given area is still a sub- ject of speculation. The more or less localization of the disease in districts, as in Wisconsin in 1922, seems to indicate local spread which cannot be attributed to infection from commercial tobacco, or even the use of home grown tobacco by the workers as suggested by Valleau and Hubbard (13), a practice which is quite uncommon in the north. A careful survey definitely excluded dissemination by seed or plants as a possibility in that year. Dissemination of infested material by wind, especially dry wind, within the district seemed the most logical ex- plantation, although in isolated cases other means accounted for the spread from farm to farm. Spread of Wildfire in tke Field. The importance of rainstorms with wind as a dispersal agent in the field and in uncovered or cloth covered seed beds is recognized by all workers on wildfire. The actual distance and amount of dissemination Control of Wildfire of Tobacco 9 following rainstorms can only be assumed, however, from the area and number of new infections occurring, which are brought out by condi- tions favoring infection. The causal agent may have been spread in many cases before the storm. Unless the wind is especially severe it is not generally believed that rain storms carry the disease over wide areas. In the spring of 1924, some seed bed experiments were laid out to test this subject by placing flats of plants with bare ground be- tween them at varying distances, up to twenty feet, from a central source of infection, but, unfortunately, the results were not convincing owing to the small amount of infection obtained. To test the possibility of man carrying the disease about in any one field or distributing it to other fields, two experiments were conducted, An artificially infected pad of cloth was used with which leaves in the field were brushed sufficiently to break the plant-hairs in one case and touched lightly in another case. Infection occurred in both cases, but was more marked in the former. One experiment was conducted dur- ing a moist period of weather and the other during a dry period. The relative results were apparently the same in both cases. Where wet infected cloth was applied to wet leaves the best infection was secured, although good infection was also obtained with wet cloths applied on dry plants. Some infection was also secured from the dry infected cloth on wet plants, when the contact was sufficient to break the plant-hairs. When dry cloth was used on dry plants no infection resulted. These tests seem to indicate that the disease may be readily spread in an infected field by man brushing against the plants when the leaves or the clothes or both are wet, but not when these are both dry. An important question relative to dissemination relates to the in- fluence of the amount of infection that can be permitted to enter the field on the seedlings when transplanted, and the extent to which infection can be kept down by the removal of diseased plants or diseased leaves. On June 28, 1923, an isolated piece of ground was selected and di- vided into eight plots, each 30 feet by 36 feet. One hundred and twenty plants, three feet apart each way, were set in each plot. These plants were selected according to the amount of disease present on them, the “badly diseased” ones showing lesions on all the leaves and the “slightly diseased” ones showing no actual lesions at all, although they came from a section of a seed bed which had been inoculated about two weeks earlier, but upon which no infection had occurred, owing apparently to unfavorable conditions for infection. The “considerably diseased” plants showed a few lesions on the lower leaves. The different lots were pulled and transplanted by different individuals to prevent con- tamination in handling. The season was unfavorable for wildfire, and at times no signs of the disease were visible in the field. Following light rains, a slight upward spread on the infected plants was noted, but no general spread occurred until following a short rain storm with strong wind about the middle of September, when the plants were full grown. Following this a heavy infection developed. On September 25, the 10 W isconsin Research Bulletin 62 number of leaves infected on each plant was estimated by two different individuals. The average infection per plant is shown in Table VII. The data show mainly that the “slightly diseased” plants eventually gave almost as much disease as the “heavily diseased” plants. The spread of the disease into healthy plots was evident, more to the eastward than to the westward, and consequently the infection in Plot 1 is believed to be due largely to the organisms originally present. Sim- ilar plots conducted in 1924 corroborated the conclusions from the pre- vious season. The experience with the careful removal of diseased leaves at short intervals from a small center of infection in plots in 1923 and 1924 was of such a nature as to indicate little or no value resulting from this practice if favorable conditions for the disease develop later in the season. While this work has been done on a large scale in Wis- consin, in the control work in 1922 in cooperation with the State De- partment of Agriculture the subsequent unfavorable conditions for the development of the disease did not give a true measure of its value. Under favorable weather conditions for infection and consequently reproduction of the parasite and for its dissemination, a very small percentage of disease in the field may rapidly develop into a large one, which may subsequently be very injurious to the crop. Transplanting of even a very small percentage of infected plants or of only slightly diseased plants is, therefore, not believed to be war- ranted in view of the damage which may result. It should be stated, however, that observational evidence on a large number of farms under apparently similar weather conditions indicates that there is no close correlation between the amount of infection in the seed bed, or the original infection in the field, and that subsequently occurring. The actual condition of the plants themselves, as a result of local field conditions, seems to play a large role. Seed Bed Infestation. All matters considered, the transfer of infested material into the seed bed is the most important problem to be taken into consideration in connection with the control of wildfire. It has been shown that it prac- tically can be taken for granted that overwintering will not occur in soil lying out-of-doors. It has been shown at this station that the wildfire organism can over- winter on artificially infested seed, and that artificially infested seed sown in seed beds may result in infection of seedings (Plate 2, bottom) A number of trials have been made however, in which infection has not been secured as a result of sowing infested seed, although conditions apparently ideal for infection to occur have been maintained. Although ir is not generally believed that seed under field conditions is infested, experience here indicates that it is a factor which must be reckoned with. It is a wise precaution, therefore, not to save seed from in- fected fields, but if seed must be taken from such fields, it should be thoroughly disinfected before sowing. The subject of seed disinfection is discussed on page 14. Control of Wildfire of Tobacco 11 Overwintering- experiments here as well as those of others, have shown that the wildfire organism readily survives the winter in dry or cured infected tobacco leaves. In the tobacco shed and stripping rooms a very considerable amount of refuse containing the living causal organism must exist following work on an infected crop. Here, apparently, lies the most important factor for seed bed infection. The dissemination of this material to the seed beds may occur in a number of different ways, unless precautions are taken to prevent it. The loose refuse should be burned or buried, followed by the precaution of placing the seed beds at a considerable distance from the tobacco shed. The use of lumber, cloth, or any other material on the beds which has been stored in the shed should also be guarded against, since this involves overwintering not necessarily on these materials them- selves but on pieces of infected leaves which may be attached to these materials and carried to the seed bed. The wildfire survey (Plate VII) in Wisconsin in 1923 brought out the following interesting observation bearing on dissemination from sheds. Out of about ninety cases in 1922, 60 growers placed their seed beds near their sheds and 27 developed wildfire in 1923. Twenty-three growers placed their seedbeds a considerable distance from the sheds, and only one developed wildfire in 1923. Out of nine new cases of wildfire in 1923 seven developed in beds placed near the sheds. This evidence seems to point towards the general importance of dissemination of infected material from the curing sheds to the seed beds. The survey in 1924 did not indicate such close correlation between location of beds and infection, but the season was unsual in many respects and other complicating factors may have played a part in infection. On farms where wildfire has previously occurred, it is an excellent precaution to keep the seed beds a considerable distance from the building which may harbor the parasite in order that wind, animals, or man may not readily transfer even small bits of infected material to the seed beds. Furthermore, seed bed boards or frames, cloth or sash, should not be stored in sheds. If they are so stored or have been on an infected bed the previous season, they should be cleaned and disinfected if again used for seed beds. According to our results, infested lumber piled out of doors in such a way that it all becomes wet will not harbor the organism from season to season. Cloth covers are not likely to carry infected material unless stored under infected tobacco. These can be readily sterilized by boiling or steaming when desirable. Dusting and Spraying Experiments In the Connecticut Valley, efforts have been made to control wildfire in the seed beds by dusting with copper-lime dust and spraying with Bordeaux mixture (1, 3). Their experimental results in the green house have shown very marked reduction in the amount of. infection on seed- lings following these treatments, and a high degree of benefit was like- wise obtained undrr out-of-door seed bed conditions. Neither in the 12 Wisconsin Research Bulletin 62 green-house nor in the field is absolute control claimed, however, by these investigators. Experiments along a similar line were started in connection with our work in the fall of 1922 (8). Most of the work has been carried out in the green-house with seedlings in flats (about 14 inches x 22 inches) which usually contained 300 or more seedlings. One set of experi- ments was also conducted in 3 foot x 3 foot seed bed areas out of doors. The first experiment was planned to show the difference in control obtained in wounded as compared with unwounded plants, inoculating artificially before and after dusting or spraying, together with a rela- tive comparison of the effectiveness of spraying and dusting and their frequencies of applications. The flats were inoculated three times (in a few cases two times) with a water suspension of the wildfire or- ganism from cultures. The data secured are shown in Table VIII. The percentage of infection obtained is relatively high, and it may be objected that this experiment was not a fair comparison as to the value of dusting and spraying on account of the number of inocula- tions and the amount of inoculum used. In the absence of any method for duplicating natural dissemination only the inoculated controls can be relied upon for comparison. While these indicate infection approxi- mately twice as great as that of the sprayed and dusted flats they were not as badly diseased as may be frequently noted in plant beds under conditions of natural infection. Table VIII indicates that only about 20 per cent more of the plants were infected, and only about two to four more infections per plant occurred in the wounded as compared with the unwounded seedlings. (It is estimated that each plant received on the average ten or more wounds.) In both the wounded and unwounded series, plants dusted or sprayed after the inoculation showed considerably more infection than plants dusted or sprayed before inoculation. No important con- sistent difference between spraying with Bordeaux and dusting could be noted in this experiment, as there was more variation between the "brand” of dust or spray used than between the methods of applica- tion. The Bordeaux paste spray used from appearance was apparently of inferior grade. “Fungi-Bordo” gave better results than "Nu-Rexo.” Increasing the number of applications of “Nu-Rexo” reduced the per- centage of plants infected and the number of infections per plant. A second experiment showed that corrosive sublimate sprayed on the plants one-half hour before inoculation reduced the percentage of plants infected from 96 to 49, and the average number of infections from 5.79 to 1.44. Leaf injury was produced by the corrosive subli- mate which could be reduced, however, by adding lime without mater- ially influencing its effectiveness. Following this trial lime alone was tried in comparison with copper-lime dust. This test seemed to indi- cate that lime alone was as effective as the commercial copper-lime dust. An experiment was then conducted in out-of-door seed beds in the spring of 1923, running duplicates in 3 foot x 3 foot seed bed areas. Air- Control of Wildfire of Tobacco 13 f slaked lime, “Limate”, “Niagara D-6,” “Nu-Rexo”, “Corona Bordeaux,” k “Sanders Dust,” “Fungi-Bordo,” “Corona Sulphur,” Bordeaux (4-4-50) spray and calicum caseinate (“Kayso”) were compared. Six applications of the chemicals were made, two being applied before one light artificial inoculation of the wildfire organism made on June 14. On July 11 an examination of the beds seemed to indicate that the “Limate” and “Kay- i so” plots were as free from wildfire as' the uninoculated controls. Slight infection was found in the others and considerable infection in the in- oculated controls. “Niagara D-6” and “Sanders Dust” gave some leaf injury but not enough to seriously affect the plants. On October 13, 1923, young seedlings in flats were dusted with “Kayso” and Limate” in comparison with “Sanders Dust”, “Fungi-Bor- do”, and dry soil. The percentage of plants' showing infection are shown in Table IX. “Kayso” alone apparently gave the best results, due probably in part to its adhesiveness. “Limate” was approximately as good as the copper-lime dusts. Soil dust for some reason increased infection above that of the inoculated controls. These results with spraying and dusting are believed to have some bearing upon the theory and practice of this method of control for wildfire, although corroboration of the results and conclusions may be necessary to bear them out. Copper, the toxic constituent in Bor- deaux spray and copper-lime dusts, has never been regarded as a good germicide and its use as al, spray to prevent bacterial infection is quite unusual in the history of plant disease control, although its value in preventing fungus invasion is universally recognized. The experiments indicate further that copper is not the effective, agent in the case of wildfire control. It seems more likely that the effectiveness of spraying and dusting is due in part at least to its physical rather than to its chemical action. While “Limate” or “Kayso” is not recommended for the practical control of wildfire, yet the latter could probably be used to advantage on account of its adhesiveness. Spraying and dusting, with any material used in our tests however, do apparently not pre- vent the occurrence of more or less wildfire in the seed beds when conditions favorable to the dissemination and development of the causal organism occur. The experiments on dissemination have shown that a very slight amount of seed bed infection, in fact an infection so small as to be undeterminable at the time of planting, may result in heavy field infection, providing conditions favorable for the dissemina- tion and development of the disease occur in the field. If, therefore, spraying and dusting do not wholly control the dis- ease, the question may be raised as to the actual value of this practice. If wildfire becomes annually a common and serious seed bed trouble in any given district, spraying and dusting, or some better method of control, may need to be resorted to. Under conditions where only a low percentage of the seed beds are infested in a district, it will probably be safer in the long run for the grower to discard infested seed beds entirely in preference to taking the risk of placing even a slight amount of infection in the field, such as may occur in infested 14 Wisconsin Research Bulletin 62 sprayed or dusted beds. The actual value of this practice however, must finally be determined largely by the results which the growers obtain from its use rather than from experimental trials of the kinds described. Seed Disinfection It was pointed out earlier that while seed was not apparently a com- mon source of wildfire infection, it is regarded as an unsafe practice to sow seed grown one or two years previously in an infected field with- out thorough disinfection. Formaldehyde solution (1-16) was first used for seed disinfection against wildfire. Earlier experience here with this disinfectant indicated that it was injurious to germination in some cases and this observation has also been reported from other stations, particularly from Virginia. Corrosive sublimate (1-1000) treatment was recommended as a substitute by Fromme and Wingard (6), their experiments having shown that no injury to germination of the seed occurred under their conditions. This treatment was recom- mended and used soon after in the northern sections where wildfire was on the increase. Experience here with the corrosive sublimate treatment like that of others (I, II) proved disastrous, for the reason that while the treated seed germinated in subsequent seed germination tests on filter paper, (Plate 3, bottom) it almost universally failed to germ- inate for the farmers. The injurious action of the corrosive sublimate : treatment (Plate 3, top) was found to occur only when the seed ' was sprouted before sowing (either as' mixed with rotten wood or as pure seed) as is a common practice in northern tobacco-growing dis- . tr icts. When sown directly in soil, the treated seed sprouts normally ; : and this method of sowing is the common practice in Virginia and other southern districts. Corrosive sublimate treated seed also practically fails to germinate on potato agar. The failure of corrosive sublimate treated seed to germinate is believed to be due to the toxic action of the corrosive sublimate absorbed and retained by the seed, which in contact ; with filter paper or soil passes from the seed, but in contact with other seeds, in decayed wood or on agar, is not absorbed from the seed. We cannot agree with Anderson and Chapman’s (1) explanation of hardening , of the seed coat in this respect nor that treatment with water alone may i result in a similar injur}% although there have been cases where seed treated with water alone and lying in cloth bags in contact with other bags, treated with corrosive sublimate, fail to germinate, apparently due to the diffusion of the toxic property from one bulk to the other. , Experiments were accordingly started with the purpose of finding some satisfactory method of disinfecting tobacco seed for tobacco districts where seed is normally sprouted before sowing. A large number of tests on modifications of the corrosive sublimate and form- alin treatments were first tried. Later calcium hypochlorite, “Bacilli-KiT (B. K.), cupra-ammonium carbonate, electrically generated ozone (with possibly nitrous oxide), heat in vacuo, Seed-o-San, Semesan, Uspulun. Control of Wildfire of Tobacco 15 Bayer’s Compound and other commercial compounds, and silver nitrate were tried. Following these treatments the rate and percentage of germination of seeds on filter paper and in bulk were determined, as well as the disinfection secured by sowing the treated seed on potato- dextrose agar; plates. The data on this subject are too voluminous to present in detail here so the principal results only are given. Seed stored moist for as long as forty-eight hours, /and then dried, showed no injurious effect on germination either on filter paper or in “bulk”. Seed treated with corrosive sublimate, kept moist for eight hours or longer after treatment, retarded germination on filter paper markedly, and no germination occurred in bulk. Corrosive sublimate treatment at various temperatures from 0° C. to 30° C. did not appre- ciably influence the usual result (i. e. germination on filter paper but no germination in bulk). Two to eight washings after treatment with the sublimate did not measurably alter its normal behavior. Poor drying after treatment retarded germination only slightly as compared with moderate to good drying. Soaking seed in water up to two hours before treatment with the sublimate had no influence on the result. A twenty minute treatment with corrosive sublimate, 1 to 1000, re- tarded germination appreciably on filter paper, as compared with shorter treatments. Five, ten and fifteen minute treatments with corrosive sublimate, 1-1000 ,gave good but not perfect disinfection of seed so far as wildfire was concerned, but did not permit germination in bulk. Corrosive sublimate (1-500) for fifteen minutes retarded germination somewhat more on filter paper than did the standard treatment. Cor- rosive sublimate (1-2000) was not effective as a disinfecting agent. Soaking seed in water after treatment with corrosive sublimate up to thirty hours did not favor its germination on filter paper or in bulk but rather added to the injurious action secured. These, or other modifications of the corrosive sublimate treatment which have been tried, including those recommended by Anderson and Chapman (I), do not permit the germination of the seed in bulk, or in decaying wood with anything like sufficient certainty to warrant its recommendation in districts where sprouting before sowing is prac- ticed. The formaldehyde treatments did not prove particularly injurious to the particular lots of seed used in these experiments, either on filter paper or in bulk up to about 2 per cent formaldehyde with 15 minute treatments. The objections to the formaldehyde treatment lie primarily in the fact that its disinfecting properties are not so reliable as cor- rosive sublimate up to strengths which are not likely to be injurious to the germination of some lots of seeds. Our experience and that of others also has been that formaldehyde (5) is much more injurious to some lots of seed than to others for reasons not fully understood, and it is, therefore, not regarded as a promising tobacco seed disinfectant. Calcium hypochlorite (about 2 per cent Cl. water) retarded germin- ation about 10 per cent only in treatments from two to twenty-four hours, but the seed was markedly bleached. 16 Wisconsin Research Bulletin 62 ‘Bacilli-Kil” (B. K.), about 3.38 per cent NaClo up to four hours, followed by washing, did not injure seed appreciably, but seed was bleached and it was not effective as a seed disinfectant in treatments of less than four hours duration. Cupra-ammonium carbonate (spray formula) did not injure seed ger- mination up to 1 hour treatment, but it did not give sufficient disin- fecting action to warrant further trials. Five hours treatment killed seed but did not satisfactorily disinfect it. Ozone (with perhaps nitrous oxide) generated electrically did not injure dry seed up to eight hours, but wet seed was killed in about four hours. Neither treatment was sufficiently effective as a disin- fecting agent in our tests. Heat treatments, even under reduced pressure, did not give satis- factory disinfection up to temperatures that killed the seed in these limited trials. A number of commercial seed disinfectants, mostly of the organic mercury compounds, have been tested both as dust and liquid treat- ments. These have included principally Seed-o-San, Dupont Semesan, Dupont Dust Disinfectant No. 12, Bayer’s Dust, Bayer’s Compound and Uspulun. None of these met all the requirements for disinfection of tobacco seed. The dust treatments as a rule do not permit germin- ation in bulk, and the liquid treatments retard the germination in bulk to such an extent as to render them unsafe to recommend in practice. The disinfecting value of these compounds against the wildfire organism on tobacco seed proved in all cases to be so low at any of the strengths recommended (and in some cases with increased concentrations and long treatments) that they cannot be recommended for this purpose. In the first experiments with silver nitrate as a disinfecting agent for tobacco seed a solution (about .33 per cent) was used in treatments varying from two to thirty minutes. Germination was not appreciably- injured either when tested on filter paper or in bulk, and good disin- fection was secured in all cases. In a second preliminary test silver nitrate was used in strengths varying from 0.1 per cent up to 0.8 per cent for fifteen minutes and again germination was not injured appre- ciably even at the higher strength, and excellent disinfection was se- cured at all concentrations. A number of trials subsequently made with silver nitrate indicated that it is the least harmful, of any disinfecting agent tried on tobacco seed, and that its disinfecting properties are as good if not better than that of corrosive sublimate (Plate IV). Ac- cordingly^, it was suggested that silver nitrate 1-1000 treating for 15 minutes be substituted for corrosive sublimate treatment of tobacco seed, especially in districts where seed is commonly sprouted before sowing. During the spring and summer of 1924, a decided outbreak of wild- fire occurred in Wisconsin, owing to very favorable weather conditions for its occurrence. In a few cases seed was suspected of being the agency of introduction into the seed bed, although the seed had been treated with silver nitrate. This led to further investigations on the Control of Wildfire of Tobacco 17 subject of seed sterilization, comparing particularly silver nitrate and corrosive sublimate treatments. For this purpose seed heavily inoculated by artificial means was used, which following treatment was plated out on potato agar, often using as many as forty dishes with around two hundred seeds included in each dish as a test for each treatment. As a result of these tests it appeared that occasionally a wildfire or- ganism escaped the recommended treatments, sometimes one in five or ten thousand seeds. ‘The results again indicated that silver nitrate was somewhat more effective than corrosive sublimate as a disinfecting agent. It was also especially noticeable that the former was much more effective against fungus saprohytes than the latter. In this connection, it must be remembered however, that the seeds used in these experiments were infested with at least a hundred and probably with a thousand times more of the wildfire bacteria than commonly occur on seed under natural conditions, so that it is doubtful if more than one seed in several hundred thousand escapes disinfection in practice. Since there are, however, three to four hundred thousand seeds in an ounce, the possibility remains that occasionally wildfire may escape the present methods of seed disinfection. In order to reduce this possibility to a minimum or eliminate it en- tirely, a double seed treatment with silver nitrate was resorted to, permitting the seed to dry one or more days between treatments. At the same time the length of the treatment has been reduced since early experiments showed that even a two minute treatment with silver nitrate was very effective. With the double treatment, each treatment lasting ten or even five minutes, it has been possible to disinfect the seed so completely that no wildfire organism has been recovered from seed so treated after extensive trial. Therefore, it seems that the double treatment should be used in preference to the single treatment in cases where the most reliable disinfection is required. The germination of the seed is apparently more retarded by two treatments than where only a single treatment is given, but this has not been found to be of more than one or two days duration and, con- sequently, is not to be regarded as a serious objection. The advantage of two five-minute treatments lies largely in the fact that germination is retarded somewhat less than with two ten-minute treatments. It has also been noted here that it is not advisable to sprout the seed in the same cloth in which it was treated. Loss of Virulence The experience of most workers with the wildfire organism has been that it may relatively rapidly lose much or all of its virulence in culture. This phenomenon is common with many bacterial parasites and is said to occur in nature also. A simple though fairly extensive experiment was conducted with 8. tabacum for the purpose of determining in the first place the best 18 Wisconsin Research Bulletin 62 cultural method for this organism in order to retain its virulence, and secondly to form a possible basis of reasoning in regard to overwinter- ing of the organism. Three virulent strains (isolated from different sources) were selected and transferred to three different media, potato-dextrose agar, beef- peptone agar and bouillon. These media were made up in sufficient quantity to last throughout the experiment and sealed in tubes with paraffin. In all, about 650 cultures were involved. Several original transfers were kept, and transfers were then made serially weekly and monthly, fifty-two weekly and thirteen monthly transfers being made in the experiment. The cultures were kept in the refrigerator at about 8-10° C. throughout the experiment. At intervals of about one month the non-transfers and the last weekly and monthly cultures were tested for their virulence by making 50 wound inoculations with a water sus- pension from each culture on the leaves of young tobacco plants in pots supplying suitable conditions for infection. The results are recorded as percentages of infection. Considerable difference occurred in the rate of infection and the size and appearance of the lesions, but these cannot be gone into detail here. The results on the percentage basis were on the whole quite variable when compared from month to month, due undoubtedly to variation in environmental conditions af- fecting infection. Studied in detail, the results also show occasional contradiction, i. e., a culture would at one time give a higher, and at another time a lower percentage of infection when compared to an- other. Taken as a whole, however, the following conclusions seem warranted from the data. The degree of virulence showed a general downward tendency on all media with increasing age, most marked on beef-peptone agar and least marked on potato dextrose agar (Plate V). In the case of potato agar the greatest loss of virulence occurred when no transfers were made (Table 10) and the least when weekly transfers were made. In the case of beef-peptone agar the greatest loss of virulence occurred in the weekly transferred series and the least in the no-transfer series. At the end of 15 months the loss of virulence was complete on beef- peptone agar in all three strains used. In bouillon not much effect of the transplanting itself was noted, the evidence being somewhat in favor of monthly or weekly transplants above no transplanting as regards retention of virulence. Some difference existed in the three strains used in regards to their ability to retain their virulence under any one condition, strain 3, for instance, was over twice as virulent as strain 2 on untransferred beef- peptone agar. When transferred back into potato-agar, the cultures in all cases seemed to be approximately equal in vigor of growth, but virulence was not materially altered. None of the cultures, as a rule, gave as high percentage or as good infection as freshly isolated cultures from new lesions. Aside from this the best culture medium for B. tabacum seems to be potato-dextrose agar, with transfers at intervals of somewhat less than one Plate I Top — Typical symptoms of wildfire on portion of tobacco leaf. The chlorotic area or “halo” surrounding a whitish or brownish central necrotic area is characteristic of this disease. Bottom — Wildfire infection in a seed bed often kills young plants. Plate II Top — barly and Late symptoms of blackflrc of tobacco. This disease differs quite strikingly from wildfire and is due to a different bacterium. The control methods, however, are much the same so far as known so that it is believed the recommendations presented in this bulletin for wildfire control apply to blackfire also. Bottom- Wildfire may overwinter on the seed. (A) Control plot sown with uninfested seed. ( B) Wildfire resulting from sowing artificially infested seed. Plate HI Top — Corrosive sublimate prevents sprouting of seed in bulk. Silver nitrate treatment does not prevent sprouting by this method. Bottom — The germination of tobacco seed after different treatments using the ordinary method for testing the seed. (Note that corrosive sublimate does not prevent germination by this method, nor does it prevent germination when seed is sown dry in the soil. C. — Control no treatment. S.N. — Disinfected with silver nitrate. C.S. — Disinfected with corrosive sublimate. E. — Disinfected with formalin. Plate V Illustrating the comparative loss of virulence of the wildfire organism grown in pure cultures for one year on different media and with different frequencies of transfers. A— Potato-dextrose agar. 1 — Transferred weekly. B — Bouillon 2 — Transferred monthly. C — Beef peptone agar. 3 — Xo transfer. Plate VI Wildfire symptoms produced by toxin only. A. — Leaf inoculated with a sterile filtrate from wildfire cultures. Dish above shows results of plating out from such spots. B — Inoculated with bacteria and toxin from wildfire culture. Dish above shows organisms present on plating out. Inoculation with bacteria alone re- quires longer to produce symptoms. EAST HALF OF DANE COUNTY Plate VII Spread of tobacco wildfire in Dane County, Wisconsin. The cases of in- fection on farms in 1922, 1923, and 1924 are shown. It will be noted that on some farms the disease re-occurred three years in succession while on other farms no infection occurred after 1922. A marked spread of the disease can be noted in 1924. The survey was made in detail in only the townships of Burke and Sun Prairie in 1921. The survey from which this map was made was supported by the Wisconsin State Department of Agriculture. Control of Wildfire of Tobacco 19 month. The wildfire organism may lose all or part of its virulence under relatively favorable conditions for the growth and storage of the organism. The fact that it has favorable conditions for multiplica- tion outside of the host is not necessarily conducive to continued patho- genicity. That the organism can and does retain life and pathogenicity upon such materials as seed or leaves in the dry and latent state is evi- dent from the overwintering tests. Under other conditions, whether the material which harbors it is dry or sufficiently moist to favor growth, it may rapidly die out or lose its virulence as evidenced by the results with soil, cloth and wood. It seems evident, however, that whether the wildfire organism is in all cases really killed or merely loses its virulence has not been actually determined in overwintering ex- periments conducted thus far. Studies along this line may explain some of the peculiar cases of behavior in overwintering studies. The Wildfire Toxin The common and characteristic halo surrounding the ordinary wild- fire lesion, the less common chlorosis of bud leaves which sometimes occurs, with few, if any, organisms in the chlorotic area is evidence that a toxic substance is produced by the wildfire organism which is apparently soluble. To obtain further information on this point, cultures of the wildfire organism on potato agar were suspended in water and filtered through a small Berkfeld filter. The filtrate proved sterile on plating and was used for inoculations on tobacco in the ordinary way. Typical halos were produced by the sterile filtrate in one to two days. Platings from these spots showed that they were sterile (Plate VI). These tests, with other modifications, were repeated five times with similar results. The wildfire organism produces toxin which, though greatly diluted, is very effective in rapidly producing chlorosis in plants. The wildfire bacteria when washed free of this toxin required several days longer to produce typical lesions than did the toxin alone. This observation is of considerable significance in work with the wildfire organism and led us to reconsider some of the previous experimental work here as well as that of others, and explained some observations previously not un- derstood. It was always noted, for instance, that the wildfire organism required several days to produce any signs of infection from some of the overwintered materials, whereas 24-48 hours sufficed for cultures. It is evident that the presence or absence of already formed toxin had much to do with this result. Most of the determinations on overwintering and influence of other environmental conditions on the pathogenicity of the wildfire organism have been based on inoculations of living plants. It is possible, there- fore, that these results apply more particularly in some cases to the effect of the toxin than on the organism itself. Anderson, for instance, found that alternate freezing and thawing did not kill the wildfire or- ganism, since inoculations with the exposed cultures produced typical 20 Wisconsin Research Bulletin 62 infection. In experiments here, cultures exposed to alternate freezing and thawing for short periods also gave infection, but these same cul- tures failed to give growth on other media when transfers were made from them some time after the exposure. It seems evident that the or- ganism was largely, if not wholly, destroyed, but that the toxin was not injured. The loss of virulence or pathogenicity, as noted in the previous chap- ter, is no doubt related at least in part to a tendency of the organism to continue to produce toxin in culture. Apparently the observation that old cultures readily produce symptoms on plants, as compared with subsequent transplants, is not the result of any greater virulence of the parasite, but is rather a consequence of the transplanting of the toxin previously produced. Similarly, the question may be raised as to whether the symptoms produced on a plant, as a result of introducing the toxin through a wound, justify including the plant in the host range of the parasite. In work with the host plants of B. tabacum (10) no distinction was made as between the toxin and the organism, in all cases probably inoculating with both. Some of these trials have been repeated sufficiently, how- ever, to justify the belief that in most cases, at least, the organism was actually parasitic, although the initial symptoms may have been pro- duced by the toxin introduced from cultures. The records show, how- ever, that in practically all cases infection was obtained by spraying the inoculum on unwounded plants, as well as by wound inoculations. Practical Considerations With the appearance of a new disease of economic importance, the quick demand for control measures often requires the dissemination of such information as may be rapidly gained from limited experimental data, together with deductions from what is known about similar dis- eases. When the problem is subsequently more thoroughly investigated, it is natural that the relative importance of the earlier recommendations will be altered, possibly some eliminated and others added. This es- sentially has been the history of the development of wildfire control measures. Control measures can best be applied by adequately under- standing a disease and selecting and using control measures that apply best to the case at hand, rather than by blindly following directions. The results secured from the investigations described in this bulletin do not fundamentally change the principles of control which have been previously recommended by this Station and by workers in other states. They corroborate previous results based on meager data, justify or eliminate certain recommendations which were in doubt, in addition to altering some of the methods. The fundamental consideration which should be kept in mind in controlling the wildfire disease is that it is of an infectious nature and that measures for its control are, therefore, based largely* on efforts to prevent its introduction in the first place into the seed beds, and failing Control of Wildfire of Tobacco 21 in this the necessary precautions should be taken to prevent its intro- duction into the field. When wildfire has occurred on a farm in the preceding season, it is evident from our experiments that any material which may harbor even extremely small pieces of infected plant tissue may be a possible source of infection to the new plant beds if permitted to reach them, and favorable weather conditions for the infection follow. It is believed, however, that the actual dissemination of infected material from the curing sheds, where the material has remained dry, to the seed beds is one of the most common sources of infection. Seed bed frames or covering, or any other material coming in contact with the seed beds should not be stored in the curing sheds, where infested tobacco hangs, without being thoroughly cleaned and disinfected before using. It is not believed that wood or cloth readily harbors the wildfire organism except as it carries pieces of infected leaf tissue. To further insure sanitary seed bed conditions it is believed advisable to locate the seed beds a considerable distance from the farm buildings or, at least, from the curing shed, since various agencies may easily carry infected material which may be harbored in or about buildings to the plant beds when they lie close at hand. The refuse from the preceding year's crop should be burned or buried to reduce danger of its dissemination. It is an unsafe practice to sow seed grown in fields infected with wildfire, without adequate disinfection. Corrosive sublimate cannot be used for disinfecting seed when seed is to be sprouted before sowing. Silver nitrate, one part to one thousand parts of water, is a satisfactory disinfecting agent, but two treatments of 5 or 10 minutes each, allow- ing the seed to dry between treatments, is believed necessary to insure complete disinfection. Some retardation to germination usually occurs. Seed should preferably not be sprouted in the same cloth used in dis- infection. The wildfire organism dies out in a comparatively short time in moist soil. Planting in previously infested fields is believed to be a safe practice, especially if the “stubble” from the preceding crop is plowed under in the fall. If infection occurs in the seed beds it is hazardous to use even ap- parently healthy plants from such beds. It is believed to be a good plan to construct small beds, separated by paths, rather than to use large, continuous beds in which infection can spread more readily. In this manner the seed bed areas not infected can be used with greater assurance of safet}-. At the first signs of infection in the seed beds, effort should be made to destroy the infected plants together with the immediate surrounding area. After trying out various methods, the conclusion has been reached that the most convenient and cheapest way to do this is simply to cover these areas with three or four inches of soil. In case a heavy early infection develops in the field, plowing under and replanting with healthy plants should be given serious consideration. If plowing is not done, all infected plants should be removed before re- 22 Wisconsin Research Bulletin 62 planting. Removing the infected plants only or picking off infected leaves is of doubtful value in checking the disease if the disease is scattered throughout the field, even on only a small percentage of the plants. Working in a wildfire infected field when the plants are wet is con- ducive to spreading the disease, if the plants are of such a size that the leaves are touched consecutively. It is believed that the methods of control worked out and suggested for wildfire apply equally well for the similar tobacco disease known as blackfire (Plate II, top). Control of Wildfire of Tobacco 23 SUMMARY 1. — Practically every case of wildfire infection in the field can be traced to seed bed infection. This is borne out by three years of observation of the disease in Wisconsin, as well as being sup- ported by reports from various other states. The control of wild- fire is. therefore, almost entirely a matter of preventing seed bed infection by the wildfire organism. 2. — Locating the materials which are the common sources of carrying the wildfire organism to the seed beds is, therefore, an important phase of the development of control methods. This involves determining on which of the various materials likely to come in contact with seed beds the bacteria causing the disease are most likely to overwinter. 3. — Other things being equal, the wildfire organism lives over winter most readily in the dry and dormant condition. 4. — The wildfire bacteria readily overwinter on infected to- bacco leaves which are cured or dried and which remain dry be- tween growing crops. Small amounts of infected tobacco trash may accidentally reach tobacco seed beds in a number of differ- ent ways and this material is believed to he a common source of infection. 5. — To reduce chances of infection from overwintered material in the curing sheds, it is advisable to locate the plant beds a con- siderable distance from the curing sheds or other places which may have harbored infected dry tobacco over winter. 6. — -The wildfire organism can readily live over winter on the seed. In fact, the experiments indicate that it can live in the dormant stage as long as two years on seed. Seed is not, however, apparently a common source of infection, although it may become so. It is not to be regarded as good practice to sow seed from infected fields, unless they have been adequately disinfected. 7. — The wildfire organism does not seem to remain alive as readily on wood or cloth, even when kept dry, as on leaf tissue or seeds, but since these materials may readily harbor infected leaf fragments, especially if stored in the curing sheds, the cleaning and disinfection of seed bed frames and covers may frequently he advisable. 8. — As far as can be experimentally determined the bacteria do not over-winter in moist soil, consequently there is no danger, as far as known, in using land for tobacco which has grown a previously infected crop, especially if the refuse and stubble from the preceding crop are thoroughly plowed under. 9. — Plants from infested beds preferably should not be used for* transplanting. If weather conditions are favorable for the disease an extremely small amount of seed bed infection may re- sult in heavy field infection. In a small percentage of cases this seed bed infection may be so small as to escape even careful in- spection. 24 Wisconsin Research Bulletin 62 10. — Spread of the disease in the field is almost entirely de- pendent upon rainfall, especially with strong winds. No satis- factory measures to prevent the spread of the disease in the field are known. 1 1 . — While dusting with copper lime dust or spraying with Bor- deaux mixture in the seed beds, reduce the amount of infection, these procedures are not believed to prevent seed bed infection sufficiently to materially reduce the amount of subsequent field infection if conditions for the spread of the disease become favor- able. 12. — Seed disinfection with corrosive sublimate cannot be used where seed is to be sprouted before sowing. After trying out a number of seed disinfecting agents, it was found that a solution of silver nitrate (1-1000) gives the best results. It is believed, how- ever, that two 5- or 10-minute treatments (drying the seed be- tween treatments) is required to adequately disinfect infected seed for wildfire control. 13. — The wildfire organism may often lose its virulence in cul- ture in a relatively short time. Many factors seem to be con- cerned in this loss of virulence, among them being the nature of the culture medium, the frequency of transfer and “strain” differ- ences in the organism. 14. — The wildfire bacteria produce a toxin in host tissue and in cultures which is responsible for the chlorosis produced in plant tissue. This toxin is readily separable from the bacteria by filtra- tion and will in itself produce typical symptoms by inoculation. This fact should be taken into consideration in further studies on such subjects as host relations, overwintering, loss of virulence, and culture studies. Table 1. — Comparison of Different Materials as Overwintering Sources. Inoculated With Pure Cultures and Extract from Green Infected Leaves. (Series I. — 1922.) Control of Wildfire of Tobacco 25 05 c o o 05 c o Months o j + ; 1 • 0 : • : i • ! 1 : 1 : 1 : 1 : 1 ■ 1 : • 1 ' 1 : 1 : i + + o 05 + o o o 1° |o 1 lo 1 • |+ O' o + o “ + + o 1 o 1 ol 1 o 1 o 1 o lo 1 1 : j ; 1 : + + o + o o o 1 o ! o 1 t o I : 1 : 1 : 1 : 78 70 1 o 1 06 L * 1 CM 1 00 40 o o ° o 1 1 o i 1 1 o • 80 09 o 00 00 1 o 1 °° 00 o o ol 1 o 1 o 1 ! o 1 : 1 : o 05 20 o o o ! o ! ° o 1 o o o 1 lo 1" 84 1 O 1 98 CM 1 CM CM CD o o o 1 o i o | o 0C o !° 92 1 1 o 1° lo 1 CD o* Q) tJC CtJ c a? ! cm 00 00 62 o o C- 28 o ° 1 o Tf o 92 a a i ) Pure culture Green leaves Infested material a a a V. ‘c E ’c ■y. / ) c !/ T n c CC J Dried green leaf pulp Extract from green leaves used for 1, 2, 3, 4. . . Water suspension used for 1, 2, 3, 4 Dried leaves f 1 Cured leaves Control Control Buried leaves with and without soil contact . . No. 1 cm re * 1 m | 1 I 50 1 j 1 j 1 00 05 o - CM ir. — • ^ o 5 oE- o 2 *£ c £ £ to a CC !>■ c 111 a? c o £ + OD O G CHT3 o - M5 1C T3-^ O ■si: l a o 6*3 qj QB_rc £— i to y Table II. — Comparison of Different Materials as Overwintering Sources. Inoculated With Pure Cultures and Extract from Green Infected Leaves. (Series II.— 1922 .) Percentage 1 of infection after Months W] o [SC( + DN + SIN o 1 Researce o ! o 1 o 1 o 1 1 1 1 ! i E i o i 1 >ULLETII< o o 1 o 1 ! 1 1 r 6 o! 1 2 ol 1 lo ! 1 i+ l + + o o o o ! 1 O o 1 1 + + o 1 o o- o l* + + o o o 1 1 o I j ! o 1 id « + + o o o o 1 I o- : j : 1 ! o + Days CM 24 42 o 00 1 o- ; 1 ; | o + CO 32 ■<* cm o o 1 1 1 ol 1 o ! o! 1 ° o o! o 1 1 o 1 1 + i CM CO 40 CO CM o o o o 1 ° o o- | O. o o 1 o 1 | + CO co CM o cm| CO 1 1 O i Tf 1 o CO 1 - 1 | o 1 o o 1 1° + CO 22 00 CO CM 00 Ol CO 00 CO 1 o o X |oi + “ CM CM * <>■ ! cc ! I 34 0- ^ I 0 01 o ° u Ol ! o +. Ol CO n 52 o CM u 1 CO | 00 CM CM Ol i>- |o + Source of inoculum Pure culture Green leaves Pure culture Green leaves Pure culture Green leaves Pure culture Green leaves Pure culture Green leaves Pure culture Green leaves Green leaves : Pure culture Infested material c/ 3 D < 0 n 5 3 0 3 r. 5 n < 3 r. : 5 \t i 5 3 L> r. 3 3 J Green leaf pulp Control : No. - CM CO i * 1 uO 1 o 1 i 00 I 55 •Where percentages are given they are based largely on fifty wound inoculations on single plants. The 4, 7 and 9 month inoculations were on seedlings in Hats. + = infection but percentage was not determined. Table III. — Comparison of Different Materials as Possible Overwintering Agents When Inoculated With Dried Infected Leaves. ( 1923 .) Control of Wildfire of Tobacco 27 ||3 iO 1 o j CO I 1 o 1 O 1 t> CM in 1 C" 1 SO 1 !> i> 1 1 SO 72 Ut Cfl Oj i) — ' 1 " I 1 o’ Ja a 1 Z 1 1 1 1 72 1 1 72 .“JO s X ^ 1 SS X S) O t" 1 CD SO cO t" 1 l> SO SO < H 1 1 go* so 00 00 o a> SO .2^ c 05 u « CM 00 o J a a z „ r 72 V sr £ an o so so m in so iO m C £ Co j2 SO so Tf O o 00 00 co .ii s in 00 CO CM 72 ^ cs 3—1 »■* 6 Z _ 72 CO a; H ■ 1 CO o o CM 00 CO i 1 t" r- ( £ | o z 1 ! c; r, C -C 03 C QC O * SR SR SR r 1 CO CO CM CM CM CO CO < e £§« 03 CO o o .2~ c I> 1> C3 o 72 j_ 03 a; ^ — o* J a a z _ 03 o H CjC i> c QC O SR < s 1 : 03 o "o C 43 o on S • - 03 ■a c 1 — £ 'go >3 >> >3 t- >> >3 03 "o 03 "o >3 -a 1 •- G — - Q Q Q <5 Q I 3 IS 73 — 03 a sO C > CS S3 72 > 72 a; > aj - _■ 72 o cj o 33 to cs 03 CB QD C -2 a) T3 a; % 03 o X £ Seet ‘o C/2 "o C/2 o PC Q p U Table IV. — Influence of Storage Conditions on Overwintering on Various Materials. (1923.) Wisconsin Research Bulletin 62 in o * Lesions on 100 plants O CO 70 oc 174 187 128 144 167 186 <2 H Age in months I> t> e- CO CO CO cO CO 1 CO 6 Z Lesions on 100 plants n 26 37 88 OS 74 OS 00 79 98 <£ h Age in months CO cO CO CO lO £ n in m iO t No. 3 Lesions on 100 plants CD 05 56 50 Tf CO 06 108 128 c n V H Age in months m in in 1C •<* Stored Inside Outside Inside Outside | Inside Outside Inside Outside | Inside Outside Inoculated material X c V T Cl c CC ! c l V c a r-i "a c c c c "c h 3 > 3 V c a £ T a X u % c T C r 3 ) 3 > 7 2 3 ) 3 Table Y. — The Influence of Soil Conditions on the Wildfire Organism. ( 1923 .) Control of Wildfire of Tobacco m 6 2 Lesions on 100 plants o © 174 222 291 112 167 c n © ►J i—i d-2 o’a 2 « O'- ; O O. 'to© 32 >=X ac o . • U, - T 3 ■ X ■ T 3 "St) +-> ai CO33 3 £< o a o « c „ 1 « > _ CO o.2 OO U < Table VII. — Dissemination of Wildfire in Field. Plots in Row from West to East. One Hundred and Twenty Plants per Plot Equally Spaced. Control of Wildfire of Tobacco u 02 £ Most of the infection in this plot believed due to infestation which was not visible at time of planting Infected from neighboring plots Infection largely from the 5 infected plants in the plot All about equally diseased All about equally diseased All about equally diseased Amount due to handling questionable due to plot lying in direction of general spread Infection all due to spread from other plots Average number of leaves in- fected per plant Sept. 25 10.6 3.9 SOI 14.5 14.2 13.3 14.5 13.5 Planting (June 28) All “slightly diseased” All healthy All healthy except 5 Alternate rows of healthy and slightly diseased . . Successive rows of healthy, slightly, considerably and badly diseased. Rows north to south Same as 5 but rows east and west All healthy plants after handling diseased plants All healthy Plot - (N CO in (D 00 3le VIII. — Percentage of Plants Infected and Average Number of Infections per Plant Following Dusting and Spraying for Wildfire in Green-House Flats. 32 Wisconsin Research Bulletin 62 r -i * | 1 r [ o' G C ^,0 3 f— ( o 00 © CO CM 05 05 00 i oc 1 X o o © 1— » X w3 X CM 1 o o | . J CM -? -t OO CM 00 m CO 00 CM T-H CM •3* Tf X > ; u I T3 m m m m o m m in m m to o o io to m © © m 1C in o > o • o H-> co © © in CM o 00 m m (N CO m t" co oo 00 © X X o 00 00 00 00 i> 00 05 05 CO m o iO m r>. G G s to CO CM 05 00 CM 00 05 o "Cf 00 o © © CO © OO X I> l> i> 00 00 CO 00 00 CO 05 05 05 CO in m m X m o c? O C/3 0) QC cd G 0) 00 m . o m CM 00 o Tj* CM © X © CO 1 n © © o T i> CO oo 05 l> 00 l> t" 00 05 <35 CO m Tt< m CD 1 rj* © a* « 1 -a G O [o '5c G 00 CO 00 oo CO CM OO OO oo o O o OO OO oo o CM oo i 1 |X ■ X © o o 3 1 h E, 1 a cd o c /2 cd oo 00 00 00 OO CM oo 00 oo CO OO o 00 00 oo OO OO CM X X X X j ' C o U | £ cd 1 >> >> >5 1 v ' 3 3 1 1 3 3 Ih y, Ih Ih i a a 2 _ a a 2 ; - "C C/3 x cd o C/3 C/3 3 1 3 V « 0) G c 5 c a •j3 O •r$ a cd *.£} c 3 •J a J E 3 3 cd a ) o 3 c : 3 t: J u c c o £ o c. > 'qd a ) •-< a ) s-. c ) 0) 111 a > a) ff J < cc J < < cc V > 3 45 T3 G O T3 V -a a 3 o £ G Z Tarle X. — Average Percentage of Infection With Bacterium Tabacum Carried on Different Culture Media for Control of Wildfire of Tobacco 34 Wisconsin Research Bulletin 62 Table IX. — Percentage of Plants Infected Following Dusting With Different Materials for the Control of Wildfire in Green-House Flats. Material used Perce ntage oi ini Fection Series 1 Series II Average Sander’s Dust 75 63 69.0 Limate 58 79 68.5 None (inoculated control) 84 79 81.5 Fungi-Bordo Dust 64 84 74.0 Calcium caseinate (“Kayso”) 49 65 57.0 Soil Dust 96 90 93.0 None (uninoculated control) 0 0 0 Literature Cited (1) Anderson, P. J., and Chapman, G. H. 1923 Tobacco wildfire in 1922. Mass. Agr. Exp. Sta. Bui. 213 : 1-27. (2) Anderson, P. J. 1924 Overwintering of tobacco wildfire bacteria in New England. Phytopath. 14: 132-139. (3) Clinton, G. P., and McCormick, F. A. 1922 Wildfire of tobacco in Connecticut. Conn. Agr. Exp. Sta. Bui. 239 : 365-422. (4) Fromme, F. D., and Wingard, S. A. 1921 Treatment of tobacco seed and suggested program for control of wildfire and angular leaf spot. Phytopath, (abstracts) 1920: 21. (5) Fromme, F. D., and Wingard, S. A. 1922 Blackfire or angular-leaf spot of tobacco. Va. Agr. Exp. Sta. Tech. Bui. 25 ; 4-42. (6) Fromme, F. D., and Wingard, S. A. 1922 Blackfire and wildfire of tobacco and their control. Va. Agr. Sta. Bui 228: 1-19. (7) Jenkins, E. H. and Chapman, G. H. 1923 Wildfire of tobacco in 1922. Conn. Agr. Exp. Sta. Tobacco Exp. Sta. Bui. 2: 7-38. (8) Johnson J. 1924 Experiments with dusting and spraying for the control of tobacco wildfire in seed beds. Phytopath (abstracts) 14: 28. (9) Johnson J. and Murwin, H. F. 1924 Disinfection of tobacco seed against wildfire. Phytopath. (ab- stract) 14: 28. (10) Johnson J., Slagg, C. M. and Murwin, H. F. 1924 Host plants of Bacterium tabacum. Phytopath. 14: 175-180. (11) Thomas, H. E. 1924 Tobacco wildfire and tobacco seed treatment. Phytopath. 14: 181-187. (12) Tisdale, W. B. 1923 Report of Tobacco Experiment Station. Fla. Agr. Exp. Sta. Ann. Rpt. 1923 pp. 125 R-140 R. (13) Valleau, W. D., and Hubbard, C. 1924 Angular leaf spot and wildfire infection of tobacco plants by spitting. Phytopath. (abstract) 14: 29. (14) Wolf, F. A., and Foster, A. C. 1918 Tobacco wildfire. Jour. Agr. Res. 12: 449-458. (15) Wolf, F. A. 1922 Wildfire of tobacco. N. C. Agr. Exp. Sta. Bui. 246 : 4-26. September, 1925 i $0 -1 yj 7 Research Bulletin 63 Transmission of Viruses From Apparently Healthy Potatoes v HF U8RARV Of I'Ht OCT 26 ya* r* y*r * ^iVERSITy of iLLlr JAMES JOHNSON Agricultural Experiment Station of the University of Wisconsin Madison CONTENTS Introduction - Experimental Methods Inoculations with Diseased Potato Foliage Inoculations with Apparently Healthy Potato Foliage... Inoculations from Tubers and Other Organs of Potato.... Symptoms of the Diseases The Infectious Nature and Increasing Virulence of the Viruses . — The Properties of the Viruses Trials with Potato Seedlings and Other Healthy Plants. Other Host Plants Transmission Back to Potato A Combination Disease Discussion of Results Summary 1 1 2 2 4 4 5 7 7 8 8 10 11 12 Transmission of Viruses From Apparently Healthy Potatoes' D URING the course of cross-inoculation studies on certain virus diseases of solanaceous plants, it was nuted that symptoms were secured on tobacco from potatoes selected as healthy controls, and that these symptoms did not materially differ from those secured when various virus diseases of the potato were used as a source of inoculum' An investigation of this matter, therefore, was undertaken and it became increasingly evident that extracts from potato plants which are healthy, within the ordinary meaning of this term, are capable of inducing symptoms • of disease in tobacco and other solanaceous plants. Furthermore, the ability of inducing this disease is apparently universally present within most, if not all, of the standard varieties of potatoes. Three distinct symptoms have been secured which are associated with at least two and probably three distinct viruses. These viruses behave like those of most virus diseases of plants as regards transmissibility, and have been not only transferred repeatedly through several generations of tobacco, but have also been used to infect a wide range of other solanaceous plants. In fact, one of these virus diseases when inoculated back into the potato, after having existed in tobacco for one or more generations, produces under the proper environmental conditions a most malignant disease. Many problems have naturally developed during the course of these in- vestigations, centering around an explanation of these results. As far as can be judged at present only two theories appear to be logical. Either potatoes are almost universally infested with viruses or they are capable of initiating virus diseases in other plants. Whether actual proof of one or the other of these theories can be established remains to be determin- ed by further experimentation. In the meantime it has been thought advisable to present the data secured up to this time in summarized form, with brief discussion of some of the more important features of the problem. Experimental Methods The potatoes used in practically all cases have been grown in a low temperature (17-22° C.) green-house suitable for a good development of the potato. In connection with other experiments similar plants have been subjected to a wide variety of environmental conditions, including high temperatures, and in no case have selected healthy potatoes exhibited any symptoms of a disease comparable to those to be described as a result of these changes in environment. The Triumph variety of potatoes was used in all experiments unless otherwise mentioned. Cooperative experiments with Office of Tobacco Investigations, Bureau of Plant Industry, United States Department of Agriculture. 2 Johnson, James. A virus from potato transmissable to tobacco. Phytopath, (abstract) 15 : 46 - 47 , 1925 . ? Wisconsin Research Bulletin 63 The tobacco plants and other host plants used were grown in a high temperature green-house (27-32° C.). The plants were grown in very fertile soil, transplanted to 4-inch pots and usually inoculated when very young, i. e., with only two to four leaves large enough to be inoculated. The potato or other foliage used for inoculum was crushed in a small sterile mortar, the juice strained through cheese-cloth, and inoculated by means of sterile needles wrapped at the end with a small wad of absorbent cotton to more readily carry drops of the inoculum. Twenty to forty punctures per plant were usually made in the leaf blade and midribs, al- though experiments showed that a fair percentage of infection could be secured by a much smaller number of punctures. In the early experiments, ten plants were usually used in each series of inoculations. In later experiments only five plants were used. This number gives equally reliable information except in cases of negative results, in which case, however, the experiment has always been repeated with the same source of inoculum. Infection was sometimes evident in six days, although ten to fourteen days were usually required for final counts. The strain of tobacco used in practically all the experiments was the common commercial variety grown in Wisconsin, Connecticut Havana No. 38. Several other of the more distinct varieties of Nicotians tabacum have been tried sufficiently to warrant the belief that probably no important varietal differences exist as regards susceptibility to these diseases. Inoculations with Diseased Potato Foliage In connection with the earlier work, and in later work where mosaic potatoes were used as controls, inoculations have been made from 28 different Triumph potato plants with mosaic symptoms. This involved inoculation to 210 tobacco plants, infection being secured on 154 or 73 per cent of the plants inoculated. The potatoes used were mostly Wisconsin grown, coming, howevqr, from several different farms. Inoculations have been made from fourteen other diseased potatoes show- ing yellow-dwarf, spindle tuber, leaf roll, rugose mosaic, and other symptoms not clearly defined. These plants come from tubers grown in such widely separated sections as Maine, New York, Wisconsin, and Oregon, as well as being of different varieties. One hundred and five inoculations were involved, infection being secured on 71 plants or 68 per cent. No consistent differences were noted between the symptoms secured from the different diseased potatoes used as sources of inoculum. Inoculations with Apparently Healthy Potato Foliage 1 he potato plants used in this group of inoculations have been largely of the Triumph variety grown in Wisconsin. In connection with a tuber indexing project single eyes have been grown from over 12,000 different tubers, coming from twenty different farms during the past winter. These potatoes grown at a low temperature (17-22° C.) in the green-house in fertile soil have given an exceptional opportunity for the selection of Transmission of Viruses 3 normal plants as the source of inoculum. From this stock, however, ap- proximately only 170 healthy plants have been used, over fifty of these coming from stock indexed for mosaic the previous season and grown in isolated plots. In addition healthy plants have been used from tubers of different varieties, coming from Maine, Michigan, Idaho, Oregon, Colorado and Florida. These potatoes have included in addition to the Triumph such varieties as the Green Mountain, Irish Cobbler, Rural New Yorker, Early Ohio, Burbank, People’s Russet, Peach Blow, King and Brown Beauty. The plants were usually selected for inoculum purposes when very young, showing usually only four or five leaves, the plants being only three to six inches high. In the case of about 150 of these plants only two or three of the basal leaves were used as the source of inoculum. The plants were then transplanted to 6-inch pots and allowed to grow for an addi- tional two months (Plate IV). During the course of this time not a single one of these plants selected as healthy showed any symptoms of potato mosaic or any other disease. Inoculations have been made from a total of 170 healthy Triumph po- tato plants, involving 965 inoculations to tobacco. Infection was secured on 681 plants or 70 per cent of the plants inoculated. In about 2 per cent of the cases the potato plants used failed to give any symptoms in the first trial, but whenever this occurred a second inoculation was made and in every such case positive results were then secured. Consequently, all potatoes tested in this series regardless of the source or variety have yielded the symptoms, in question on tobacco. The different experiments naturally have not been equally successful in the percentage of infection obtained. In some experiments four or five infected plants out of five inoculated were the rule, in others only one to three were the rule. These differences are believed to be due in part to environmental differ- ences or to variations in predisposition of the different lots of plants upon which the inoculations were made. In experiments in which different varieties of potato were used, 20 plants (two of each) were tested. One hundred and twenty plants were inocu- lated, infection was secured with 66 plants, that is with 55 per cent of the plants inoculated. As a rule, no significiant or consistent differences in expression of symptoms were secured from different varieties. Some varieties, however, seemed to give a higher percentage of infection than others. The Rural variety, especially, was found to be unlikely to give positive results, unless conditions were especially favorable. It should be stated in this connection that we have secured at least two or possibly three distinct types of symptoms from healthy potatoes, although they are not apparently limited to particular varieties. The Rural variety is distinctly different from the other varieties studied in that it is least like- ly to yield the “mottle” type of disease and most likely to yield the “ring-spot” type. The difference between these symptoms will be de- scribed later. 4 Wisconsin Research Bulletin 63 Inoculations from Tubers and Other Organs of Potato A number of inoculations were made to tobacco directly from the potato tubers themselves, rather than from plants grown from such tubers. For convenience, no distinction was made in this group of inoculations be- tween the use of healthy and mosiac tubers, since in about, one-half of the cases the condition was not definitely known, the distinction being in any case apparently immaterial to the problem under consideration. The tubers used in these inoculations came from the various sources previ- ously referred to and included all the varieties previously mentioned. The inoculations' were usually made with extract secured by scraping the flesh of a freshly cut tuber and straining through cheese-cloth. Al- together 63 different potato tubers have been used, involving inoculations to 365 plants with a total of 130 plants (38 per cent) infected. The per- centage of infection is considerably lower than that secured from potato foliage. Apparently, the infective principle exists in a less' virulent or a more diluted form in the tuber than in the foliage, although this con- clusion need not necessarily follow from the foregoing results. In two sets of experiments a considerable number of tubers failed to give infection. It was decided, therefore, that it is simpler for present purposes to use the growing plants as sources of the inoculum, the plant in any case having to be grown to determine whether the tuber used was healthy or diseased. Our data, therefore, show a considerable number of negative results with tubers, but in every such case a plant was grown from the same tuber and positive results secured from this. The infective agent does not appear to be localized in the tuber, interior areas of tuber flesh giving infection as well as the cortical zone. Young white sprouts gave good infection and extract from the stem or root tissues gave as good infection as did that from the green leaves. Symptoms of the Diseases The symptoms usually secured on tobacco on inoculation directly from potato is a regular pattern of mottling so obscure or mild that it readily escapes casual observation. On close observation in comparison with healthy controls, however, there is usually no difficulty in recognizing the symptoms, although in some cases it is difficult for even the trained eye to distinguish the diseased from the healthy leaf. On the other hand, in many cases the mottling is especially marked, (Plate 1A), and may in some cases be accompanied by necrosis, (Plate 3E). The pattern of the mottling has no resemblance to that in ordinary tobacco mosaic. The latter is usually most conspicuous on the youngest bud leaves, whereas in the case of the disease from the potato the bud leaves are normal and symptoms appear only on the older leaves of the young plants (Plate 3B, F). General chlorosis, leaf distortion and stunt- ing are usually absent in the first transfer on tobacco, although in sub- sequent transfers to tobacco marked necrosis and stunting may occur. The pattern of the mottling and necrosis in these subsequent transfers’ is quite characteristic though not uniformly so (Plate 1C). Transmission of Viruses 5 As previously indicated, in inoculations from Rural New Yorker potatoes, it was noted that the symptoms of mottling secured from other varieties were rare or entirely lacking. In a small percentage of the inoculated plants, however, a distinctive necrotic symptom developed (Plate 2B) which resembled a disease which had been occasionally noted in field tobacco, and which has been referred to by some observers as “ring-spot". This symptom has been noted subsequently, however, to occur in inoculations from other varieties than the Rural, and it is apparently associated at times with a preliminary mottling. That the “ring-spqt” disease is physiologically distinct from that associated with the “mottling” and “spot- necrosis” types of symptoms secured from potatoes appears evident in sub- sequent transfers to tobacco and to other differential hosts. This would seem, therefore, to indicate the existence of at least two distinct types of infectious agents in apparently healthy potatoes. During the course of most of the experiments the necrotic type of symptom com- monly associated with mottling on tobacco has been regarded as merely a more virulent form of the latter. Whether this is true or whether this necrotic type indicates still another virus, i. e., a possible third type, in combination with the mottling type, has not been satisfactorily determined. Certain recent experiments lend probability to the idea of three distinct virus types. While the ring-spot symptom, usually starting with ring-like chlorotic areas, leads to necrotic ring-spots, this development is distinctly different from the necrosis commonly associated with the mottling symptom. The latter form of necrosis frequently bears a relation to the direction of the principal veins in the early stages, (Plate IB), and the entire leaf may subsequently collapse. At other times only small necrotic areas are formed which may subsequently become sufficiently numerous to cause a gradual death of the entire leaf. This is the form of disease which will later be shown to cause a striking disease when inoculated back to potato. For the purposes of the present discussion, therefore, reference to this is made as “spot-necrosis” in contrast with the other two symptom types, “mottle” and “ring-spot”, although further studies may show the first two of these to result from the same virus. The Infectious Nature and Increasing Virulence of the Viruses It was at first supposed that the symptoms secured on tobacco from healthy potato might be the result of a toxin or other irritable substances, such as are well known in animal pathology, and consequently not be- longing in the category of the virus diseases. Trials soon showed, how- ever, that the transmission of the disease from tobacco to tobacco is more readily accomplished than is the original transfer from potato. In such subsequent tobacco transfers a higher percentage of infection is' secured, (Table I), the incubation period is shorter and the symptoms are more marked. If it is considered that the “spot-necrosis” type of symptoms belongs with the “mottle” type, then very striking increase in virulence occurs as a result of passing the virus through one or more 6 Wisconsin Research Bulletin 63 generations of tobacco. Occasionally, however, the “spot-necrosis” form has been obtained on tobacco directly from healthy potatoes, but more often it develops following the transfer of the “mottle” form through one or more generations' of tobacco. Increased intensity of the “mottle” and the “ring-spot” form have also been noted, and increased Table I. — Summary of Results of Inoculations to Tobacco With Viruses of the Type Secured From Apparently Healthy Potatoes. Source ot inoculum Number used as inocu- lum Number plants inocu- lated Number plants infected Per cent infection Potato (Triumph) mosaic foliage 28 210 154 73.3 Potato (3 var.) 7 other virus diseases 14 105 71 67.7 Potato (Triumph) healthy foliage 170 965 681 70.5 Potato (10 var.) healthy foliage 20 120 66 55.0 Potato (10 var.) healthy tubers 20 120 30 25.0 Potato (Triumph) healthy and mosaic tubers 43 245 100 40.8 Potato (seedlings) healthv 110 5(?) 4 . 5(?) Tobacco — infected with viruses from healthy potatoes 695 556 80.0 Tobacco — tobacco mosaic + virus from healthy potatoes 270 220 81 .5 Other plants (18 species) healthv foliage.. 180 0 0 Controls — no inoculation 360 3(?) 0.8 virulence up to a certain limit is apparently established whether “spot- necrosis” is considered as a separate virus or not. This fact is of special interest, since as far as known' such increase of virulence has not been definitely noted in other virus diseases of plants, although it is commonly observed with animal diseases and is also supposed to occur with some bacterial diseases of plants. In the course of experiments conducted to throw some light on questions of infectivity, increased virulence and other properties of the virus, 695 plants have been inoculated and infection has been secured on 556 or in 80 per cent of the plants inoculated. This percentage of infection compares favorably with that ordinarily secured in experiments with other virus diseases. The virus has been passed through ten successive generations of tobacco, with but little if any changes in its behavior after the second or third generation. The change in virulence may be merely a result of changes in concentration of the virus. It also may develop that the number of transfers through tobacco is of little importance and that similar changes may occur as a result of the continued passage of the virus up- ward into the new leaves of the host plant, when allowed to develop to larger size than commonly used in our experiments. Neither of these ideas is substantiated by preliminary observation, however. o H . . u \ > 1 « I Cv3 rH So, « w « h 0 o c/2 (X 73 £ n O So WCC £ w < £ 1 Z C — p W < PS ^ j ^AB9q | — -7 I-.Q ^ o ipaiu | iqgif 3 g I AAB9q | . . w QO [ Uinipgui Wi | ^AB 9 q I in uimpgui iqgri cx,g L I : AAB 9 q | .-iCM • umrpgui | cm. | AAB9q | uinipgui j Tj ■ ■ ._>~ a 2 "o : _ ~~ ..-vcsti^ _ oh; 5- o ^ « a u ^ ^ oo • is £ >>(- 03 .* 55 •-’S-© • %TO 5^3 e © • “ c g C . o a a 55 « K. 73 4 ) 03 C !-s.«Sg-s= Sts O^c^oat^a c'k O wj’tc'aj Pea Disease Survey in Wisconsin 13 in present practice may best be illustrated by comparing two adjoining factory districts on the same soil types and comparable in every respect. One of these districts operating in its thirteenth year was established by a company which had previous experience with the disease, and which had avoided planting peas repeatedly. In this district only three fields showed rootrot, and that in small amount. Two of these three fields were grow- ing their third crop of peas, and no field in the district was growing more than its third crop. The loss was negligible. In the second district, operating in its twelfth year, repeated planting of peas on the same ground had not been avoided. In this district 18 fields were found with rootrot, half of which were severely infested with heavy losses in five. All of these five fields were said to have grown many crops of peas be- fore, one having produced at least seven crops in ten years. Relation of Rootrot to Soil Type and Drainage Earlier observations. — In a previous paper (10) some scattered observa- tions have been recorded showing that in different localities where peas have been grown intensively in a similar manner, there is great difference in the time which has elapsed before the disease has made its appearance in fields, and also in the rate of its spread and increase in destructiveness. These differences appeared to be associated with differences in the capacity of soils to hold water, or with drainage and sub-irrigation. For instance, in Wisconsin the Superior red clay appeared to be more subject to severe injury from rootrot than contiguous loams. Some sandy soils in Maryland underlaid by impervious clays seemed remarkably favorable for the development of disease — an observation which seems to be supported by more recent observations by Drechsler (3). In irrigated districts of the Rocky Mountain States, peas on soils of low moisture-holding capacity rarely suffer from disease unless subirrigated, though on some of them occasional diseased plants can be found. There are several ways in which the water relations of soil might affect the development of disease in peas. The most apparent of these is the favorable environment which abundant water in the soil might provide for the semi-aquatic parasite causing the disease. If, as has been assumed tentatively in this paper, the fungus is widely distributed in soils, it may be originally much more frequent in wet soils than in those which do not retain water. Studies in 1924. — Whatever cause or causes give rise to the observed variation in the behavior of rootrot, it was clearly of great importance in this survey to determine to what extent the several pea growing soils of Wisconsin do affect the behavior of this disease. Any differences which might be found would not only affect the cropping systems which must be used on the several soils to avoid disease, but might affect the direction of expansion of the industry. The method of classifying soils found most suitable for this study is that provided by the Wisconsin Soil Survey. To a considerable extent, soil types as distinguished by the soil survey are representative of a certain degree of drainage. Some entire series are characteristically well drained. Others are poorly drained. Within each soil type there are, however, many relatively minor, but still, from the point of view of this study, important differences in drainage. Fields of uniform soil type. — In the course of the survey, peas were examined on 27 distinct soil types besides seven groups of incompletely classified soils, making in all 34 groups into which the total of 688 fields are divided. There are, therefore, too few fields in many of these groups to afford a satisfactory basis for comparison. A complete summary of the data obtained is given in Table II. From this complete table, one important conclusion can be drawn. None of the soil types encountered shows any promise of furnishing an environ- 14 Wisconsin Research Bulletin 64 ment where peas may be grown without danger from rootrot. Diseased fields are recorded on all but eight of the types. The eight exceptional types are represented by so few fields that the absence of disease in the location where they were found can not be taken as evidence that they are naturally less liable to disease than others. For instance, of the fifteen fields on Superior silt loam, only three were growing their second crop, and these were in a relatively new canning district in situations where disease would hardly be expected. The 13 fields on Fox silt loam were in a district producing its fifth crop of peas but none of these were growing- more than its third crop. Under these circumstances, it may be said that this soil type has a more promising record indicating freedom from disease than any other. In contrast with the Fox silt loam, the soil type which shows the most unfavorable record with reference to rootrot is the Colby silt loam. Since only 20 fields were encountered on this soil, its present record with reference to rootrot should not be regarded as convicting it of being the most favor- able soil for disease in the state. The table shows, however, that of the nine fields found which had grown one or more crops of peas previously, all were diseased to a greater or less extent. This soil is characteristically compact with poor internal drainage and in many cases with faulty sur- face drainage. In view of the tendency of the canning industry to expand on to this soil, a more comprehensive examination of the behavior of the crop on this soil should be made. The two leading soil types. — The two soil types upon which sufficient numbers of fields were found for adequate comparison are Miami silt loam and Carrington silt loam. These show no important difference in behavior (Table III). The Carrington silt loam shows a larger percentage of infested fields, but a lower percentage of fields extensively invaded. When these two soils are compared with the total number of clay loams and clays summarized in the same table, it appears that the heavier soils show both greater percentages of total fields infested and of fields thoroughly infested. Since the average cropping histories of the fields on the heavy soils is not markedly different from that on the silt loam, the comparison appears to demonstrate a greater tendency for rootrot to become trouble- some on the heavy soils. Table III. — Comparison of Carrington Silt Loam and Miami Silt Loam, the Two Chief Pea-Growing Soils of Wisconsin, With the Total Clay Loams and Total Clays as to the Percentage of Fields Found Infested With Rootrot and the Percentage of Fields Showing Light, Medium, or Heavy Infestation. Soil Total fields Number of fields with .rootrot Per cent fields with rootrot Percentage of total fields (I Light nfestation Medium ) Heavy Carrington silt loam . . . 180 57 32 16 6 9 Miami silt loam 146 37 25 8 3 12 Total clav loams 76 33 43 17 12 14 Total clays 51 21 41 16 8 18 Comparison of soil classes. — In an attempt to condense Table II in significant manner preserving the summarized cropping histories of fields, all sands and sandy loams were placed in one group, all loams and silt loams in a second group, and all clays and clay loams in a third. Table IV. prepared in this way, reveals differences in behavior between the medium light and the heavy soils. In the group of clay loams and clays it appears that rootrot makes its entry more promptly and spreads through the field Pea Disease Survey in Wisconsin 15 more rapidly than in the silt loams and loams. All fields on clays and clay loams growing their fourth crop were diseased, while 29 per cent of such fields on loams and silt loams, and 40 per cent on sands and sandy loams were still rootrot free. An attempt has been made to condense Table II by grouping together all fields in the same soil series. The groups thus formed appear to be too small in most cases for satisfactory comparison. Rootrot in uneven fields. — The comparison of the behavior of soil types presented in the foregoing tables does not emphasize differences so much as examination of individual fields extending over two or more soil types, or fields not uniform in drainage. Fields of one soil type, with poorly drained portions, or of two or more types differing in tendency toward wetness were found in almost every district. In such fields, as a general rule, whatever the cause of the wet spots, whether lack of drainage, or seepage, or the texture of the soil enabling it to hold water, rootrot appeared to have entered the field first in these wet spots. Many of the apparent exceptions were probably due to accidental introduction of the rootrot fungus. In determining the presence of disease on very wet ground, microscopic examination was frequently used to distinguish rootrot from drowning of roots from standing water. These general field observa- tions emphasize more than do the tables the differences between soils types, and at the same time strengthen the suggestion that the observed differences between heavy and light soils may be associated fundamentally with the natural wetness of such soils. Relation of Soil Reaction and Fertility to Rootrot No special study of the relation between soil acidity or fertility and the occurrence and destructiveness of rootrot was made in this survey. Neither in previous field experience, nor in that gained in this survey has it been obvious that any important correlation exists between these field condi- tions and the disease. There are, to be sure, a number of instances cited among growers in which high acidity and low fertility have been correlated with destructive occurrence of disease; but when these have been examined they have not provided convincing evidence that correlation with either of these conditions was the essential factor causing loss. On the other hand, certain canning companies have attempted to render diseased land suitable for pea growing by carefully conducted liming and manuring experiments, but have failed completely. Rootrot in First Crop of Peas As indicated in figures 5 and 6 a small percentage of fields was found infested in what appeared, from incomplete cropping records, to be the first crop of peas. Many of these had probably grown some unrecorded crops earlier. There are, however, eleven fields in which adequate records indicate that no peas have been grown before, and in which rootrot infesta- tion was found ranging from a mere trace up to 100 per cent of the field. A number of these fields clearly owed their infestation to inoculation from neighboring diseased fields in the following manner : two from diseased fields on the same farm where the diseases had long been established ; one from surface drainage from an adjoining higher field ; one from an old barn yard included in the field ; one from manuring with uncured vines from the outside of a silage stack ; and one with root- rot along the roadside from passing loads of pea vines. The other five fields which contained rootrot in their first peas were all either poorly drained, wet soils, or contained the disease only in wet pockets. In addition to these there were a number of fields observed diseased in Table IV. — Relation of Rootrot to Cropping on Light, Medium, and Heavy Soils. Numbers of Fields in Roman Type Percentages in Italics. -aS Si AABaq uintpaui iqStl AAB3q uinipaiu iq§n I AAeaq uimpaui iqgil rsi 1*3 I I !■ Aa eaq «j Y I uinipaui I I ON I -HOQ I COIN I CO °0 I MIN I OS'* iqglj II N I ^CM , - I -M AAeaq M In I I CO | SO’ lutupaui M IN | MM MIC | CO > iqsyi MCC I I>OC 00 C> IN'!' I ^=0 I 'd'tN r-.co r^co — > i oo 'o AABaq uimpaui iq§H M=C 0=0 « M c ^ £ co CO O o~ «o hCI rfCO C^Oj M s CO — >> Light infestation, 1-25%; medium infestation, 26-75%: heavy infestation, 76-100%, oi the area of the field. Pea Disease Survey in Wisconsin 17 their second crop which were said to have been diseased in their first. In these, as in the preceeding cases, the disease was mostly in wet soil. If Aphanomyces euteiches, occuring native in Wisconsin, was the source of infestation in these fields, it appears that it was originally restricted to wet locations. Persistence of the Parasite in the Soil Once peas have failed from rootrot and their decaying roots have released into the soil myriads of thick walled oospores, the parasite is able to persist for a remarkably long time. In a previous paper (10) instances were cited in which peas had failed from rootrot when planted on a field in which peas had failed six years earlier. Certain canners have reported experiences which indicate a survival of the parasite for a still longer period. During 1924 a number of well attested cases were encountered in which peas were growing on fields in which the last previous crop had “blighted” presumably from rootrot from one to ten or more years earlier. Assuming that in all of these cases rootrot was the cause of the earlier blight — an assumption which is undoubtedly true in nearly all of the cases — these fields have been classified in Table V. on the basis of the number of years that have elapsed between the previous blighted crop and 1924. The fifteen fields on which peas had blighted within ten years were found to contain more or less rootrot, the majority of them being still heavily infested. Of nine fields blighted more than ten years ago, only three were still thoroughly infested, while five were, as nearly as could be ascertained by careful search in the field, entirely free from the disease. This indicates that rootrot infestation does actually tend to diminish gradually with time, but the length of time required to free contaminated soil is discouragingly long. It appears unsafe to reliant peas on infested soil within a decade. In this connection it may be added that field observation seems to indicate that the disease persists longer in heavy wet soils than in soils less favorable for the advent of the disease in the first place. This rela- tion is not obvious from the table, however, and requires more careful records for its confirmation. Resistant Varieties and Date of Planting in Relation to Injury from Rootrot Although it has been shown in experimental trials that no variety of peas is completely immune to rootrot, and that among the usual commercial varieties there is little difference in resistance, as measured in experimental trials, nevertheless search was made for evidence indicating that the slight differences between varieties is of any importance in averting loss on infested soil. The most resistant varieties that have been found, the Horal and Rice’s No. 330, did not occur in surveyed fields. It may be added, however, that in a trial conducted by the Columbus Canning Company in cooperation with the U. S. Department of Agriculture during the summer these two varieties showed far greater resistance than has been shown by any commercial variety in general use. A study of the resistance in commercial plantings is greatly complicated by the fact that the date of planting is a factor to be considered when comparing the damage sustained by different fields. For instance, the record shows that there was a slightly smaller percentage of fields of Alaskas and Winners diseased and that they seemed on the wliole to suffer smaller crop losses than other varieties. However, the peak of the planting season of these two early varieties was from threie ; to 18 Wisconsin Research Bulletin 64 four weeks in advance of that of the sweet varieties, a fact which undoubtedly accounts in large part, if not completely, for their lighter damage. Thus a comparison of a few selected fields furnishes more reliable evidence of varietal resistance than a comparison of the records of varieties as a whole. From a comparison of suitable fields it appears that the Green Admiral alone showed appreciable resistance. One of the older companies operating in an infested district plants Admirals on all fields suspected of harboring disease. On one farm in this district, a uniform 18 acre field of thoroughly Table V. — -Persistence of the Rootrot Fungus in the Soil. Fields Which Are Known to Have Grown “Blighted” Peas Arranged According to the Period of Years Since Peas Were Grown Last and According to the Degree of Infestation Found in 1924. Interval since “blighted” peas • 1- P c o Z 2i 43 OC Medium | p Heavy | “ 3- o O o Z Light | ^ yea s 3 ■3 p £ Heavy | “ 5- p G O z Light 1 Medium | £ rs >> > a p E None | © or 43 CC J Medium | 3 >re >> > a V Carrington silt loam 1 1 Miami silt loam 3 3 2 Colby silt loam 1 Wabash silt loam .2 Unclassified loam 1 Red clay loam 1 T Clay loam . Superior fine sandy loam 1 1 Sandy loam 2 U nclassified 2 2 Total 1 3 2 5 2 2 5 1 3 infested Miami silt loam was planted to Alaskas and Admirals on the same day. When first observed on June 30 both varieties were thoroughly infested to the extent of about 90 per cent of all of the plants. The Alaskas were beginning to die, but the Admirals showed no evidence of injury above ground. On July 22, the Admirals were ready to harvest. The vines were short, pods poorly filled, and leaves dead on the lower half of the vines. However, the six acres of Alaskas yielded 166 pounds of peas per acre, while the Admirals produced 2,111 pounds per acre, though quality was not of the best. Here, the Admiral seemed to demonstrate marked resistance. In another similar field suitable for comparison the same varieties were planted on different dates — the Alaskas on April 11, and the Admirals 11 days later. On June 30 the root destruction had advanced far in both cases. The Alaskas were filling pods, though about 25 per cent of the plants were almost dead, while the Admirals were but 12 inches tall appearing perfectly healthy. Alaskas yielded 1,800 pounds per acre while Admirals yielded 1,035 pounds. The low yield of the Admirals in this case appears to be due to the later date at which they were planted, permitting the disease to attack them at an earlier stage of development. A number of other less closely comparable instances add evidence in favor of the view that under certain conditions the Green Admiral pea has some degree of resistance, enabling it to produce a fair yield under conditions which damage other varieties much more severely. Until the Pea Disease Survey in Wisconsin 19 newer resistant varieties become available the Green Admiral appears to be the only pea showing a sufficient degree of resistance to warrant its use on infested soil, although this variety may fail utterly under severe condi- tions of disease. The Control of Rootrot The findings of this survey suggest very clearly the control measure which must be employed to avoid rootrot— control measures which have been for the most part stated previously. In districts where pea culture on a large scale has been introduced recently and where there are few diseased fields already established, increase in disease can be avoided readily. First, poorly drained soil should be avoided for pea planting. The adoption of a long rotation on suitable soil should defer the appearance of disease for many years, perhaps indefinitely. The length of rotation re- quired to prevent serious development of the disease appears to be dependent to some degree on the soil type, being longer on clay soils than on loams. A rotation of five or six years duration is suggested as probably adequate on most Wisconsin soils. If it appears advisable for commercial reasons to plant peas as long as possible on the same ground, the field records collected here show that under average conditions it is possible to do this for three years before serious loss from rootrot need be anticipated. Occasionally they may be planted for a longer term of years. Generally, the disease appears in such fields for one or two years before it becomes destructive ; and thus a car-eful examination of fields for disease can readily determine when such fields have become unsafe for further planting. No serious loss from disease need be incurred from such practice if in- telligent supervision is employed. In districts where the disease is already well established avoidance of disease is not so easily accomplished. Fields in which peas have failed from rootrot are not safe for replanting for ten years after the failure on most soil types. Soil from such fields can serve to carry disease to other fields during this period of time, and thus much new land that has never grown peas in infested districts is unsafe for peas. Since it will be impossible in most cases to determine in advance just where such injured areas are, it will be impossible to avoid loss in all cases, even where a suitable rotation is adopted. As soon as infested tracts are located, they must be abandoned for pea culture. Transfer of soil from diseased fields should be avoided. Uncured silage from pea vines should never be fed or returned to fields as manure. The use of resistant varieties of peas may become profitable on infested soils under some conditions. Such varieties should be planted as early as possible. Under conditions favorable for the development of the disease even the most resistant varieties known at present may be damaged greatly, and in any case their growth increases soil infestation quite as much as those varieties which are readily destroyed. LESS IMPORTANT PEA DISEASES Fusarium Stem and Rootrot A stem and rootrot of peas caused by Fusarium martii App. & Wr. var. pisi F. R. Jones has been described (8) as occurring in Wisconsin and several other states. This Fusarium which was the only important parasite among several species and varieties tested produced typically its initial and most significant invasion at the base of the stem at or immediately above the point of attachment of the cotyledons. The resultant lesion becomes elongate. 20 Wisconsin Research Bulletin 64 extending up the stem as a wedge-shaped, dark brown or chocolate colored lesion, not appreciably shrunken until well advanced. This cortical rot may deepen and pentrate or even sever the vascular cylinder, after which, at higher soil temperatures the fungus invades the xylem for a short distance, producing a bright orange red or brown discoloration which may extend as far as the first node. Extensive vascular invasion is not a characteristic development. Rootlets may be attacked, in which case the symptoms are not visually distinct from the effects of several minor parasites. When this disease occurs alone as a stemrot it has been considered to be of easily recognized character. It has, however, almost always been found in association with rootrot where its presence rarely can be discovered except by the isolation of . the fungus. In the course of the survey only a few instances of the type of stemrot caused by Fusarium were discovered, and in all of these cases laboratory study showed the cause to be a phoma- like fungus which is mentioned below. Thus it appears that the Fusarium stem and rootrot of peas did not occur in Wisconsin this year as an independently recognizable disease. Laboratory study was not made to determine whether it occurred in association with rootrot. The absence of this disease this year may not have been due to the absence of the parasite. A study of conditions which make possible the development of this disease has shown that a mean soil temperature of 18° C. is necessary before conspicuous lesions on stems appear, and that a soil temperature of approximately 24° must be reached before plants are killed or conspicuously injured. Reference to the record of soil temperature prevailing this year discussed previously will show at once that this disease must have been delayed in development even more than rootrot, and that there was little opportunity for it to become destructive. Thus it is possible that in a warmer season this disease may appear again, though its seems unlikely that it will be important under Wisconsin condi- tions apart from its association with rootrot. Footrot, a Disease Resembling Fusarium Stem and Rootrot Early in June, 1924, before soil temperatures were favorable to the independent parasitism of Fusarium martii pisi, plants were found showing lesions typical of Fusarium stem and rootrot. Such lesions, when plated out, yielded cultures, not of Fusarium, but of a Phoma or phoma-like fungus which the senior author has isolated from pea root and stem lesions many times before. Inoculation experiments in the greenhouse have demonstrated that this fungus is capable of producing lesions, which resemble very closely those produced by Fusarium. Haenseler (5) has reported frequent isolations of Phoma species from peas in New Jersey. Footrot symptoms were encountered widely but sparingly in Wisconsin in 1924. Probably the disease will not prove of great importance. Seedling Injury Caused by Rhizoctonia The sterile or. Rhizoctonia stage of Corticium vagum B. & C. is another fungus capable of damaging the underground portions of the pea plant. Of wide occurrence in cultivated soils, this fungus is frequently encountered in pea fields where under some conditions it may assume considerable importance. Generally, however, it is of minor importance as a parasite of peas. Rhizoctonia may attack any underground portion of the pea plant, but it causes greatest injury when invading very young tissues. It may enter germinating seeds killing the embro or destroying the cotyledons, Pea Disease Survey in Wisconsin 21 removing the food reserve of the developing seedling. It may attack seedlings before emergence from the soil, injuring or completely destroying the growing points of roots and stem. When the stem tip is thus destroyed the pea frequently produces secondary shoots, one or more of which may escape similar destruction. Root tip injury may continue even after the plant is well established. This fungus may also produce lateral lesions on stems and roots of a type characteristic of this fungus on other hosts, being brownish, sunken and eroded, oval or oblong cankers. Coarse brown hyphae of the fungus frequently found on and around such lesions are helpful in diagnosis, but in general the injury caused by this fungus, par- ticularly upon roots, is not always readily distinguished under field condi- tions from that of some other parasites. Richards (13) has shown that the soil temperature most favoring the parasitism of Rhizoctonia on the pea is 18 °C., but that it is able to operate in a less important way through a wide range of temperatures, beginning as low as 9° and continuing up to 29°. The minimum temperature is thus below that of the major pea root parasites, and consequently Rhizoctonia injury occurs earlier than the more important root diseases. Late planted peas suffer greater injury than those planted early in cold soil. In the 1924 survey, it was not possible in all cases to distinguish the injury produced by Rhizoctonia under field conditions. Injury attributed to this fungus was noted in 35 fields ranging from the killing of 30 per cent of the plants in rare cases to reductions of stand that were negligible, and from reduction c 1 vigor that would amount to 25 per cent of the crop to that which would escape detection. Rhizoctonia injury was noted on soils ranging from light sandy loams to muckv clay loams, but the greater reductions of stand and vigor were limited to a few fields of sandy loams, Carrington silt loam and Miami silt loam. Seedling and Root Injury Caused by Species of Pythium When pea plants suffering from rootrot are examined in the laboratory, the species of Pythium .'ong known as a destructive seedling parasite will often be found present in the diseased tissue of many or all of the plants. Inoculation with some of the cultures of Pythium obtained in this way has shown that the fungus is capable of preventing germination of pea seed or of destroying many of the seedlings before they emerge from the ground, and occasionally some degiee of stem and rootrot is produced. Some preliminary work with this fungus earlier led the senior author (6) to express the opinion that it was the most important cause of pea rootrot — ail opinion which was not substantiated by further work, and which has since been corrected (10). However, more recently Stone (16) in Ontario has called attention to the association of Pythium with disease, ascribing to it a rotting of pea plants near the surface of the soil. Some attention has been given to Pythium species in relation to pea dis- ease during several years and though the study of the relation of species of this genus to root injury is far from complete, a few notes on the progress of the work may be presented. Although it will be shown in the following tables that species of Pythium capable of causing severe seedling injury under favoring conditions are present in adundance in some agricultural soils, yet the survey records no instance of important injury from these species. The most obvious explanation fur the failure of this group of fungi to produce injury is found in the comparatively high soil temperature required for their activity. An incomplete study of the more actively parasitic species indicates that a soil temperature of 16 °C. is necessary before much seedling injury occurs. Most peas have passed the stage at which seedling injury is possible before the mean soil temperature has reached this point. ?? Wisconsin Research Bulletin 64 The study of the relation of Pythium to root and stemrot has been greatly retarded by the fact that the cultures obtained from peas in the field have been found to belong to several species differing somewhat in pathogenicity and frequency of occurrence ; and furthermore it is not at all certain that present methods used in making isolations from plants secure cultures of all species present. Thus a vast deal of work will be required before the relation of Pythium species to stem and root injury of mature plants will be fully known. It may be stated, however, that inoculations made under controlled conditions have shown very slight ability in any species studied thus far to produce either stem or rootrot under -usual field conditions. During the summer, isolations were frequently made to secure cultures of any species of Pythium that might be present. Since previous experience had shewn that seme species arc almost as frequently found in association with , *oots apparently healthy as with those diseased, cultures wcie made from both. In making cultures, no sterilizing agents were applied to the surface of decaying tissue because they penetrate rapidly and destroy all Phycomycetous mycelium quickly. Thus mycelium adhering to the outside of roots may give rise to a culture, a difficulty which can hardly be avoided. The results of 76 isolations are summarized in Table VI. This table cor- roborates previous experience that cultures of Pythium are obtained with approximately the same frequency from healthy as from diseased plants. It also shows that in about one-third of the isolations two species were obtained, though in such cases no two species seem to be found more frequently associated than others. The cultures obtained in this way have been partially classified into species groups which are designated by letter in Table VII. Species A, B, D, and perhaps E, seem to have been included by pathologists under the name Pythium debaryanum ; and this group contains the more aggressive parasites. From this table it appears that the several species were obtained with approximately the same frequency from healthy plants as from those showing disease. In fact a small number of isolations from clover roots made at the same time with those from peas have given equal success in obtaining cultures of Pythium, the frequency of occurrence of the several species being somewhat different. From the observations made thus far it appears that some of these species of Pythium occurring abundantly in the soil may at times be responsible for the death of root ends, and for some rootlet injury; but only rarely for root and stemrot as it is usually known. It is possible that almost universal invasion of the root cortex of peas and clover by a mycorrhizal fungus (9) renders this tissue especially accessible to these species, and that it is from this superficial invasion that the fungus is most frequently obtained in culture. Table VI. — Frequency of Occurrence of One or More Species of Pythium in Isolations from Healthy and from Decaying Pea Roots and Stems. No culture of Pythium obtained 1 species of Pythium obtained 2 species of Pythium obtained Healthv pea roots. . . 10 22 14 Decaving pea roots and stems 6 16 8 An Undescribed Wilt Disease During the survey a disease was observed which in its effects upon the vines superficially resembles rootrot injury, but .which seems etiologically distinct from any of the known pea diseases in Wisconsin. It was character- Pea Disease Survey in Wisconsin 23 ized generally by rapid and complete withering of the vine without conspicuous rotting or discoloration of the cortex of roots and basal stem such as are typical of the better known diseases. Root tip injury was frequently found associated with it but not to a sufficient extent to account for the death of the plants. Fifty fields were encountered in which what appeared to be this disease was present. Infestation varied from small patches to 100 per cent of the field ; in the latter case crop destruction, especially of the sweet varieties of peas, was almost complete. In several factory districts this Table VII. — Frequency of Occurrence of Several Species of Pythium in Healthy and Diseased Pea Roots as Represented by Isolation. Spe- Spe- Spe- Species Unclas- cies A cies B cies C D and E sified Healthy pea roots 6 20 4 14 6 Decaying pea roots and stems . . . 3 9 2 11 7 disease caused greater losses than did rootrot, and in the total area surveyed it ranked second in destructiveness only to the disease caused by Aphanomyces. Slightly over half of the infested fields lay in Fond du Lac County, and nearly all of the remainder were in adjoining counties. Three- fourth of all such fields were on black soils of which Carrington silt loam was dominant. This disease appears to be correlated with the previous growth of peas, much the same as is rootrot. Leaf and Podspot or “Blight” Caused by Ascochyta The most widely known of the foliage diseases of peas is the leaf and podspot caused by Ascochyta pisi Lib., the conidial stage of MycospJiaerella pinodes (Berk. & Blox.) Stone. So abundant and important was this disease in Wisconsin in 1911 and following years that it was regarded as the chief cause of “blight” of peas. Recommendations made for its control assisted perhaps by climatic conditions during the past few years have brought about its almost complete disappearance. This fungus attacks all varieties of canning peas, and occurs on vetches. On pods, the lesions are rounded, somewhat sunken, light brown at first becoming darker with a light brown border, and with brown pycnidia in the center. On leaves the lesions are irregularly rounded with yellowish brown or ash}' centers and dark borders. On stems the lesions are elongate. If lesions are abundant they may become confluent killing a large part of the foliage, or even the entire plant. Lesions near the surface of the ground may spread over a considerable portion of the under- ground stem, which in the past has given rise to the impression that this fungus was the cause of much of the injury to roots that has later been found due to Aphanomyces. It has been demonstrated by Stone (15) and Vaugham (17) that this fungus lives over winter on diseased vines producing the ascigerous stage in the spring from the spores of which plants may be infected. The fungus has long been known to be carried in the seed. Several years ago when this fungus was very abundant the senior writer found nearly 10 per cent of peas in one sample of commercial seed carrying this fungus, though among many samples examined, only a few were found to contain infected peas. The fungus usually enters the seed coat and 24 Wisconsin Research Bulletin 64 cotyledons close to the plumule. When the seed germinates, the ends of the embryonic leaves in the plumule become invaded by the fungus. When the leaves are carried above ground and expand, the fungus carried in them produces lesions promptly, usually at the margins of these leaves. Pycnidia are soon formed, the spores are scattered over the foliage of other young plants by splashing rains, and the disease spreads from such centers of infection. Since the fungus penetrates deeply in the seed, no method of seed treatment thus far tried has been able to destroy it without damaging the seed. Leaf and pod spot caused by Ascochyta was of no importance in Wis- consin in 1924. Only 18 fields situated in 12 different districts were found to contain traces of the disease, and none of these were appreciably in- jured. The disease was first observed as early as June 2 where it was developing apparently from infected seed on plants only five inches tall. The disease was not found again until July 15. Its frequency of occurrence seemed to increase up to the end of the season, but no factory district was observed to contain more than three fields which showed traces of infection. Leafblotch Caused by Septoria Leafblotch of peas caused by Septoria pisi West is frequently associated with and confused with the leaf and podspot caused by Ascochyta. This disease was formerly considered an important factor in pea “blight.’’ Melhus has shown (12) that it is rarely difficult to distinguish these two diseases except perhaps in their later stages. The disease caused by Septoria is typically a blotch rather than a spot. Its margins are irregular, sometimes angular when restricted by the larger veins of the leaf, and without any distinct marginal band. They are yellowish green at first, turning brown upon the death of the tissue. Such blotches increase in size indefinitely, sometimes extending down the petiole and infecting the stem. Infected tissues produce numerous pycnidia, yellowish brown at first, becoming darker with maturity. The overwintering of the fungus has never been traced in a satisfactory manner, Melhus found the pycnospore short-lived and found no seed infection. The senior writer, however, found one collection of this fungus, kept in a sheltered location out of doors, that maintained viable pycnospores for at least a year. Seed infection is not infrequent. Infection usually occurs at the hilum which exhibits a very characteristics pink discoloration, involving more or less surrounding area. Although the fungus can be isolated readily from such seeds, no infested seedings have yet been observed from their germination. In 1924, Septoria leafblotch was more abundant than leaf and podspot. but still was of very little importance in the state as a whole. It was observed first June 6, and was collected at intervals during the summer without showing any marked increase with late summer rains. It was observed in only 29 fields in twelve districts. Most of these fields contained no more than a trace, while a few were damaged to the extent of about 10 per cent. Septoria was most prevalent in the northern districts, all of the heavily infested fields being situated north of the latitude of the north end of Lake Winnebago. A Septoria Leafspot New to Wisconsin In one field in Dodge County was found an unfamiliar leafspot caused by a fungus identified by Dr. J. J. Davis as Septoria flagellifera E. & E. Both the spots and the pycnospores of the fungus are distinct from those of Septoria pisi. Ellis and Everhart (4, p. 57) have described the spots and the fungus from material gathered in South Dakota as follows : Pea Disease Survey in Wisconsin 25 “Amphigenous, spots suborbicular, 0.25-1 cm., diameter, subzonate, with a slightly raised border, rusty-brown at first, whitening out in the center: perithecia hemispheric — prominent or subconical, dark amber color, 75-124 microns in diameter, sporules filiform, hyaline, nucleolate, only slightly curved, 80-120 x 2-2.5 microns. “Differs from S. pisi West, in the different character of the spots and the much longer sporules.” This fungus occurred sparsely in the one field where noted and showed no evidence of possible importance as a parasite. Anthracnose The anthracnose of peas caused by Colletotrichum pisi Pat. was recently described in Wisconsin by Jones and Vaughan (7). This disease closely resembles that caused by Ascochyta and may easily be confused with it in casual examination. Occurring on all aerial parts of the plant, it produces lesions that on pods are circular and sunken ; on leaves, irregular in out- line ; and on stems, elongate. These spots are generally brownish with some- what darker brown borders. Stem lesions when covered with spores from the numerous acervuli are ashen when dry, and copper colored when wet. In later stages, small black sclerotial bodies, which are helpful in identify- ing the disease, develop in the lesions. The life history of the fungus has not been followed through the winter ; although the fungus attacks pods freely, it has not been observed on the seed. Pea anthracnose has been reported in the United States only from Wis- consin thus far. Although recognized as a destructive parasite, its distribu- tion has been considered so limited that its importance in the state as a whole was minor. In 1924, both in severity of the disease in individual fields and in the total injury observed during the survey, anthracnose was far more import- ant than any other foliage disease. It was found in 42 fields in 13 districts lying in 6 counties. The majority of these fields were infected very lightly or were infected so late in the development of the crop that injury was slight, but there were a number of fields in which losses were severe, amounting to as much as 50 per cent of the crop in one 28 acre field. A few fields showed severe defoliation relatively early, but in the main heavy infection did not occur until near the end of the season when heavy rains caused rapid increase of the disease where it occurred. In a number of fields anthracnose was associated with bacterial blight in causing important foliage destruction late in the summer. Downy Mildew Downy mildew of peas caused by Peronospora viciae (Berk.) DeBy. is found widely but sparsely distributed almost wherever peas are grown. Rarely does it become conspicuous. Frequently a few plants may be found which have been completely overrun by the mildew in an apparently systemmic infection. Such plants are dwarfed beyond recovery, but con- stitute at most a small fraction of one per cent. Typically the mildew occurs as irregular downy patches of violet gray color on the under side of leaves. Such leaves are yellowish above and are usually recurved. The occurrence of downy mildew in 1924 was limited to a very few fields, in none of which did it cause important injury. Bacterial Blight Sackett (14) has described a bacterial disease of peas caused by Pseudomonas pisi Sackett which caused much loss in Colorado in 1915 and 26 Wisconsin Research Bulletin 64 following years. A similar if not identical disease occurs sporadically in Wisconsin, rarely causing important crop reduction. The disease is characterized by the production of water-soaked lesions of olive green to olive brown color which may remain small spots, or, under favoring conditions, may spread rapidly to include large portions of leaves and stem. Such lesions become darker as they dry. In wet weather bacterial ooze may appear on lesions. Infection is through stomata and wounds. Lesions may develop on pods, and mature peas beneath lesions are sometimes found to bear flakes of what appears to be the dried bacterial slime. Although it has been assumed that the bacteria may be carried alive in this way and thus infect seedlings, such seed transmission of the disease has not actually been demonstrated. Early spring infection occurring while the seedlings are still young may result in important reductions of stand. Severe infection at any time later is able to weaken the plants seriously, at times destroying practically the entire leaf surface. Bacterial blight was seen in 1924 both early and late in 23 fields in various districts. Of the infested fields 74 per cent had never grown peas before, or at least not in recent years, a fact which seems to indicate no important correlation between the occurrence of this disease and the previous growth of peas. Likewise no correlation was found between bacterial blight and source of seed. Early attacks were observed to weaken the plants and to cause uneven development. Usually such early infection was outgrown and did not lead to subsequent increase of the disease. A few fields were observed to show signs of recent infection during late June and early July, but the most severe cases were observed after the middle of July. At this time a few fields were seen badly damaged, in one of which the disease had developed freely over leaves, stems, and pods. Mosaic Pea mosaic, observed in experimental plantings at Madison in 1923. was first recorded as occurring in commercial plantings in Wisconsin in 1924, when it was found widely but sparingly distributed. In spite of its early appearance in 1924 at Madison it was not conspicuous in commercial fields until July 12, but after this date at least traces of mosaic were found in practically every factory district visited. In all, it was recorded in 63 fields in 10 counties. Most infested fields contained a mere trace,* while the heaviest estimated infestation was 20 per cent. A considerable number of fields showed from 5 to 15 per cent of the plants infected. Varieties found diseased with the number of fields of each are : Green Admiral, 20 ; Alaska, 14 ; Horsford, 13 ; Perfection, 10 ; Winner, 3 ; Advancer, 2. Not only did Admirals show the greatest number of infested fields, but also the heaviest infestation and the most evidence of injury. Injury from mosaic appeared to be negligible except in a few fields where infestation was heavy. In some of these fields, it appeared that mosaic plants were somewhat dwarfed and failed to fill as many pods as healthy plants. Judged by its behavior under conditions prevailing in Wis- consin in 1924 mosaic of peas can hardly be regarded as such a menance as mosaic diseases of some other crops have been. The origin of mosaic in these pea fields can be only conjectured at present. Dickson (1) has reported the appearance of the disease in several varieties of field peas from seed transmission. On the other hand, Doolittle (2), using seed from mosaic plants in experimental plantings at Madison, Wisconsin, and McMillian, Michigan, has planted nearly 1,000 seed from Alaska peas and smaller numbers from other varieties under controlled conditions without obtaining a single mosaic plant. The field occurrence of the disease in 1924 did not suggest seed trans- Pea Disease Survey in Wisconsin 27 mission, for different varieties and peas from seed from different sources appeared to develop the disease almost simultaneously in certain districts. In one, for example, mosaic was found in only four fields representing three varieties from different sources. On the other hand Doolittle (2) has produced mosaic in pea plants from mosaic red clover by transfer of aphids and by artificial inoculation. Inas- much as many pea aphids migrate to peas from red clover, on which they winter, it seems likely that mosaic clover plants which are abundant locally are the source of the disease in commercial plantings. FIG. 8.— LOCATION OF PEA CANNERIES IN WISCONSIN Circles indicate the location of the pea canning factories in Wisconsin. Black dots designate the factory districts surveyed for disease in 1924. 28 Wisconsin Research Bulletin 64 SUMMARY 1. — The Wisconsin pea crop of 1924 represented a total farm value of over $7,000,000. Pea diseases play a major part in de- termining the systems of pea culture employed and in reducing profits to both growers and canners. 2. — In 1924 a detailed survey was made of 688 fields compris- ing 5,416 acres representatively distributed in the pea growing sections of Wisconsin to determine the importance of the various pea diseases and especially to study the development of rootrot in relation to cropping practices, soil types, and other factors which appeared to influence its occurrence and destructiveness. 3. — This bulletin is a summary of the findings of this survey, supplemented with notes from pea disease investigation conducted in this state by the U. S. Department of Agriculture in cooper- ation with the Wisconsin Experiment Station. 4. — The rootrot chiefly considered in this survey is that caused by the fungus Aphanomyces mteiches Drechsler. This fungus is assumed to be indigenous is Wisconsin soils, occurring especially in wet locations. It increases rapidly in the soil with culture of peas. 5. — The season of 1924 was so cool and favorable for the de- velopment of peas that fields infested with rootrot did not appear to suffer as great damage as in other years. 6. — The rootrot caused by Aphanomyces was more destructive in 1924 than all other fungous and bacterial diseases of peas com- bined, considering the state as a whole. In some localities a new- ly observed “wilt” disease was more destructive, and in the state it ranked second to the Aphanomyces rootrot. Anthracnose caused by Colletotrichum pisi Pat. was the most destructive of the foliage diseases, causing important losses in several districts. 7. — Rootrot was found in 32 per cent of all fields examined. Eleven per cent of all fields were severely infested. The total loss in inspected fields is estimated at 8 per cent of the total yield. Since diseased fields were especially sought in the survey it is believed that the pea crop in the state as a whole did not suffer as great a loss as this. Even if the loss in the entire state amounted to only half this amount or 4 per cent of the total yield it would represent a loss to the growers of about $300,000 in addition to losses incurred by the canning companies. 8. — Of the fields inspected 48 per cent were growing their first crop of peas. Fields which had been planted more than once to peas were found to have on the average a rotation period of about two and one-half years. 9. — Rootrot increases both in frequency of occurrence and in severity with the number of crops grown. Rootrot occurred but rarely in fields growing the first crop of peas, while all fields growing the fifth crop were more or less infested. The occur- rence of severe infestation does not increase rapidly during the Pea Disease Survey in Wisconsin 29 first four crops ; but it rose to 56 per cent of the fields growing the fifth crop. 10. — Peas were found growing on 27 soil types and seven groups of incompletely classified soils, thus making the number of fields on most types too small for comparison. No soil type showed prospects of providing environment in which rootrot can- not develop. The two soil types which include nearly half of all the fields examined — Miami silt loam and Carrington silt loam — show little difference in behavior. In general, with similar crop- ping, clays and clay loams have a larger percentage of severely in- fested fields than loams and silt loams or lighter soils. In fields including more than one soil type, disease usually appears first in the soil with greater moisture holding capacity, or in poorly drain- ed spots. Greater precautions to avoid rootrot are needed on heavy or wet soils than on well drained, medium, or light soils. 11. — Rootrot was found to persist in some Wisconsin soils for 10 years after it had caused crop failure. After such failure.no fields were found entirely free from rootrot in less than 10 years. 12. — The only commercial variety of pea that showed an im- portant degree of resistance was the Green Admiral ; and even this variety was greatly damaged when not planted early. 13. — A five or six year rotation is suggested as a method of control of this disease which should prevent its appearance on most Wisconsin soils not already infested. When a shorter rota- tion seems advisable, careful inspection of fields can detect its development before it becomes destructive. Resistant varieties are being tested which may be of value in some situations. 14. — Other diseases discussed in this bulletin are as follows: Stem and rootrot caused primarily by species of Fusarium was not found in 1924. A new but apparently relatively unimportant footrot caused by a species of Phoma was noted. Seedling injury caused by Rhizoctonia solani Kuhn was noted in 35 fields, but for the most part was not important. The relation of species of Pythium to seedling and root in- jury is discussed briefly. A new wilt disease was found in 50 fields, in some localities causing greater losses than Aphanomyces rootrot. The cause has not yet been determined. Leaf and podspot caused by Ascochyta pisi Lib. was rare and unimportant. Leafblotch caused by Septoria pisi West, was not abundant but was important in a few fields in northern districts. A leafspot new to Wisconsin caused by Septoria flagellifera E. and E. was noted. Anthracnose caused by CoUetotrichum pisi Pat. was the most important foliage disease encountered, causing considerable dam- age late in the season. 30 Wisconsin Research Bulletin 64 Downy mildew caused by Peronospora viciae Berk, was rare and unimportant. Bacterial blight caused by Pseudomonas pisi Sackett was en- countered occasionally both on early planted peas on wet soil, and on foliage of mature plants late in the season. Mosaic was encountered frequently late in the season, but rarely appeared to reduce yields. LITERATURE CITED (1) Dickson, B. T. 1922 Studies concerning mosaic diseases. MacDonald College Technical Bui. 2: 1-125. illus. (2) Doolitle, S. P. and Jones', Fred Reuel. The mosaic disease in the garden pea and other legumes. Phy- topath. (in press). (3) Drechsler, Charles 1925 Root-rot of peas in the middle Atlantic states in 1924. Phytopath. 15: 110-114. (4) Ellis, J. B., and Everhart, B. M. 1900 New species of fungi from various localities with notes on some published species. Bui. Torrey Bot. Club 27 : 49-64. (5) Haenseler, C. M. 1924 Pea root rot investigation. New Jersey State Agr. Exp. Sta. Rpt. 44: 366-375. (6) Jones, Fred Reuel 1920 Pythium as a causal factor in “pea blight.” (Abstract) Phy- topath. 10 : 67. (7) and Vaughan, R. E. 1921 Anthracnose of the garden pea. Phytopath. 11 : 500-503. illus. ( 8 ) 1923 Stem and rootrot of peas in the United States caused by species of Fusarium. Agr. Res. 26 : 459-475. illus. (9) 1924 A mycorrhizal fungus in the roots of legumes and some other plants. Jour. Agr. Res. 29: 459-470. illus. (10) and Drechsler, Charles 1925 Rootrot of peas in the United States caused by Aphanomyces euteiches n.sp. Jour. Agr. Res. 30: 293-325. illus. (11) Linford, M. B. and Vaughan, R. E. 1925 Rootrot of peas. Some ways to avoid it. Wis. Agr. Extension Cir. 188: 1-10. illus. (12) Melhus, I. E. 1913 Septoria pisi in relation to pea blight. Phytopath. 3: 51-58. illus. (13) Richards, B. L. 1923 Soil temperature as a factor affecting the pathogenicity of Corticium vagum on the pea and the bean. Jour. Agr. Res. 25 . 431-449. illus. (14) Sackett, Walter G. 1916 A bacterial stem blight of field and garden peas. Colo. Agr. Exp. Sta. Bui. 218: 1-43. illus. (15) Stone, R. E. 1912 The life history of Aschyta on some leguminous plants. Ann. Mycol. 10: 564-592. (16) 1924 Root rot and blight of canning peas. (Abstract) Phytopath 13: 348-349. (17) Vaughan, R. E. 1913 Mycospherella pinodes the ascigerous stage of Ascochyta pisi. (Abstract) Phytopath. 3 : 71-72. 3 C y/is Research Bulletin 65 October, 1925 Fertilizer Experiments: Methods of Application and Effect on Germination, Early Growth, Hardiness, Root Growth, Lodging, Maturity, Quality and Yield 1 Emil Truog, H. J. Harper, O. C. Magistad, F. W. Parker and James Sykora Agricultural Experiment Station of the University of Wisconsin Madison Contents Introduction - 1 The function of fertilizers 1 Methods of using fertilizers 1 Problems in the use of fertilizers 1 Effect of fertilizers on germination of seeds in general 2 Effect when applied in different ways 3 Relation of osmotic pressure of seeds and sprouts 3 Fertilizer experiments with corn 5 Effect of amount and method of application on germination and growth .. 5 Relation of moisture content of soil on influence of fertilizers on germination 11 Summary of experiments on germination .. 15 Greenhouse studies on effect of hill fertilization on root de- velopment 16 Summary on effect of fertilizers on root development 20 Field experiments in 1919 and 1920 20 Summary of field experiment in 1920 .. 26 Field experiment in 1921 27 Field experiments in 1922 27 Field experiments in 1923 .. 31 Field studies on effect of hill fertilization on root development 31 Influence of fertilizers in protecting young corn against freezing 32 The theory and practice of corn fertilization 32 Machinery for applying fertilizers to corn — 34 Fertilizer experiments with oats 36 Experiment with oats in 1920 37 Experiment with oats in 1921 .. .. 42 Experiments with oats in 1922 42 Experiments with oats in 1923 42 Fertilizer experiments with cabbage 42 Experiment with cabbage in 1919 42 Experiment with cabbage in 1922 45 Fertilizer experiments with potatoes on methods of application 45 Machinery for applying fertilizers to potatoes 47 Utilization of ammoniacal and nitrate nitrogen by plants — 48 Summary on utilization of ammoniacal and nitrate nitrogen 50 Secondary effects of fertilizers «... 50 General Summary 53 Effect of fertilizers on germination 53 Corn experiments 54 Oats experiments 54 Cabbage experiments 55 Potato experiments 55 Ammoniacal and nitrate nitrogen 55 Secondary effects of fertilizers 55 Fertilizer Experiments: Methods of Application and Effect on Germination, Early Growth, Hardiness, Root Growth, Lodging, Maturity, Quality and Yield 1 Emil Truog, H. J. Harper, O. C. M agist ad, F. W. Parker and James Sykora F ERTILIZERS are added to soils for the purpose of supplying plant food elements which are present in too small an amount or in an unbalanced proportion. Aside from increasing yields, fertilizers when properly used may' produce a number of desirable re- sults as follows: 1. The quality of the crop may be improved. 2. Early growth, root development, and hardiness may be promoted. 3. Lodging may be lessened. 4. Maturity may be hastened. 5. Danger of injury from frosts and other unfavorable weather con- ditions as well as from insects and diseases may be lessened. The present investigation was undertaken for the purpose of studying .the various effects which fertilizers have on crops. It was soon found that the method of applying the fertilizer has a very important bearing on the results that may be obtained, and hence a considerable portion of the present investigation was devoted to a study 7 of methods of appli- cation. Fertilizer usage may be divided into two kinds. There is first the use of fertilizers, especially phosphates, in comparatively heavy' broadcast applications for the purpose of building up the basal supply' of essential elements which are known to be too low in the soil under consideration for practically* all crops. Second there is the use of fertilizers in a con- centrated way in the hill or drill row to supply the special needs of certain crops and conditions so as to produce the various favorable effects mentioned. By' using a fertilizer according to the second method it is possible to surround the young plant with a much higher concen- tration of essential elements than is practicable with the ordinary' broadcast applications, and hence it should be possible to promote early growth and some of the other favorable effects *on the young plant with much less fertilizer and at less cost than with the broadcast method. The use of fertilizer in the hill or drill row, however, involves certain problems which must be recognized if the use in this way is to be suc- cessful. The problems are: (1) The danger of delaying or even pre- venting germination if too much fertilizer is improperly' placed near or in contact with the seed. (2) The danger of producing too much top growth in comparison with the root growth causing what is called "fireing” (a drying up of the leaves) in time of drought. (3) The danger of making the field ununiform in fertility', causing the succeeding crop to be streaky and patchy. (4) The problem of knowing the best ^hese experiments were conducted under a fellowship grant from the Soil Improvement Committee of the National Fertilizer Association. The authors are indebted to A. R. Albert for assistance in starting the experiments. 2 Wisconsin Research Bulletin 65 fertilizer to use for each particular case. These four problems as well as others were studied and are discussed in the present report. The investigation was conducted in the laboatory, greenhouse and field for a period of five years. A large portion of the investigation was carried on with corn because of its great importance as a crop and adaptability to special fertilization. EFFECT OF FERTILIZERS ON GERMINATION OF SEEDS A considerable number of investigations have been made on the effect of fertilizers and salts on germination, but comparatively few in which the method of application, moisture content of the soil, and the soil class have been considered . 2 The results of these investigators indicate that fertilizers retard germination more readily on sands than on the heavier soils ; that the injury is greatest at the low moisture contents ; and that the injury is greatest when the fertilizer is applied in direct contact with the seed. The results obtained by Allison regarding the effect of the method of application on germination are not in agreement with those obtained by Hicks, Hutcheson and Sherwin. This may be due to the small amount of soil used by Allison. Experiments were conducted on the effect of fertilizers - on the ger- mination of seeds of a number of field and garden plants . 3 The results of these greenhouse and garden experiments are sum- marized in Table I. The data in Table I show that the field pea, cowpea, navy bean, and soybean are very sensitive to the application of fertilizer in contact ^Sigmund (.17) and Rusche (14) have conducted extensive experiments on the effect of a large variety of fertilizer salts and other chemicals on the germina- tion of different seeds. Rusche has also studied their effect on the early growth of the plants. Buffum (4), Slosson (19) and Stewart (20) have reported experiments on the effect of the salts common in alkali soils on the germination of different seeds. Slosson (18) and Rudolfs (13) have shown that one cause of the delayed germination is the reduced rate of imbibition of w-ater by seeds in the salt solutions. There was a general relation between the osmotic pressure of the solution and its effect on germination, but there were some exceptions to this general statement. Harris (7) has reported some experiments on germination, using different salts and salt mixtures. The germination tests were carried out in tumblers and in most cases the salts were added in solution. He found greater in- jury to germination at low than at high moisture contents of the soil, and the injury was much greater in a sandy soil than in a loam. Hicks (8) found that the injury to germination was greater when the fertilizer was placed in the row in direct contact with the seed than when it was thoroughly mixed with the top layer of soil. He also found that nitrate of soda and muriate of potash retard germination more than acidified bone black. Hutcheson (10) reports similar experiments. Fertilizer salts were applied broadcast, and in the row with the seed on a sandy loam, and on a silt loam. The injury to germination was greatest on the sandy loam, especially when the fertilizer was applied in the drill row with the seed. Allison (1) studied the effect of ammonium phosphate and some other salts on germination. He found that sandy soils require only one-tenth as heavy an application of salts to cause injury to germination as do loams and clays. The germination tests were made in tumblers containing about two hundred grams of soil. Little or no difference was found when the fertilizer w r as applied in direct contact with the seed, one inch above the seed, one inch below' the seed, or mixed with all the soil in the tumbler. A field experiment indicated that 150 pounds per acre of ammonium phosphate applied in the drill row with the seed was injurious to the germination and early growth of corn. The soil was a loam, and the season was rather wet so tne soil was at a high moisture content most of the time. Sherwin (16) reports results similar to Hicks. Two trials were made in the greenhouse, using one gallon jars filled w r ith Plainfield sand, which was kept at a moisture content of 13 per cent. Seeds as follows were planted at depths indicated: lettuce and carrots at a depth of / inch; cucumber, muskmelon, watermelon, and scruash at a depth of 1 inch; corn, bean, pea, soybean, and cowpea at a depth of 1 y 2 inches; and potato at a depth of 2 inches. A 3-10-4 commercial fertilizer w r as applied at the rate of 500 pounds per acre, except with lettuce and carrots, in w r hich Fertilizers 3 with the seed, but that application above or below the seed largely eliminates the detrimental effect. Because of convenience and practi- bility, application above the seed is advisable with these four plants when fertilizer is applied in the hill or drill row. The germination of lettuce and carrot is affected and delayed most by the application of fertilizer near the surface. For these plants it is best to work the fertilizer into the soil several inches prior to planting. Table I. — Effect of Fertilizer on Germination of Seeds When Applied In Ways Indicated Kind of Seed Retarding effect of fertilizer on germination With the seed Above the seed Below the seed Field pea Very marked Very slight Slight Cowpea Marked Very slight Very slight Navy bean Very marked Slight Very slight Soybean Very marked Slight Slight Lettuce Slight Marked Very slight Carrot Slight Marked Very slight Squash Very marked Slight Marked Cucumber Marked Slight Marked Muskmelon Very marked Slight Very marked Watermelon Very marked Slight Marked Potato Marked Marked Very slight Corn Marked Very slight Very slight Squash, cucumber, watermelon, and muskmelon are quite sensitive to fertilizers', and for these it is best to apply the fertilizer about one- half inch above the seed if the fertilizer is applied in the hill. Potato sprouts are very sensitive to fertilizer, and hence, in potato fertilization the fertilizer should be placed either to the side of the seed or else below the seed so that the sprouts' will not come in direct con- tact with high concentrations of fertilizer in coming up through the soil. Corn seed requires considerable water to germinate, and if fertilizer is applied in direct contact with the seed, the taking up of the water is either prevented or delayed so much by the attraction of the fertilizer salts for the water, that the corn germinates slowly or not at all. When the corn once sprouts, the sprout unlike the potato sprout, is able to pass through a considerable concentration of fertilizer, and hence application of fertilizer above the seed for corn is a very successful method in hill and drill row application. The marked difference in sensitiveness of corn and potato sprouts to fertilizer raises the important question as to why this is the case. It was thought that, possibly, the sap of corn sprouts has' a higher osmotic pressure than the sap of potato sprouts, making it more difficult for the fertilizer salts' to draw water out of the corn sprouts, and thus cause them to lose turgidity and shrivel up. Freezing point ’determin- ations gave an osmotic pressure of 7.18 atmospheres for corn sprouts and 6.15 atmospheres for potato sprouts, indicating that the possibility ca S es 1,000 pounds was applied. The fertilizer was applied at three different depths to each crop, viz : y 2 inch helow the seed, at the same level as the seed, and (with the exception of lettuce and carrots) y 2 inch above the seed Since lettuce and carrots must be planted shallow, the fertilizer could not ^keTinl l^ttlT" thC SCed ’ S ° tt sim *> iy »PPUed “SadSlt and The experiment just described was repeated twice with some of the croDS in the garden on a Miami silt loam. Fertilizer applications of the same kind and rate were made as those of the greenhouse experiments. The seed I Se planted m rows and the fertilizer was applied in a band 2J4 inches wfd P o OOOOOOOOOOOOOOOOOOOOOOOOOOOO oo I>tO!D05 00NNM(»Tji®T#CK0ONM®NOM'0 00 00»' cm oo tj« cs © in in m ■*!* t-i oo oo oo © co r- co ® rf no «o no ® in in oo © © oo •> CMod©©©©r-< T 4noCO®'’tf®TA©o6cioO©cOrHl>I>nO©C<0©CMo6 ■^ ■^•*7s oE <*« °.g . 2 .2 o ‘3 a a o o uu cm cm CM CM j •PQ j co 3 : -C« = i co : j-ft : o w : 00 -ft 00 73 73 73*73 73 73* 73 73* ‘ © 73 CS V3*°°< -«◄< «i«i< m^cq JJ • o ■<73 -d ^'.W WWW WWW WEE W< O 5 «j HHW WHH « JJ u « 3 oj a £ S a S 2 o o o uoo OINM NMN »£« ail O _ „ go© -ft © © U cm oo oo CO C3 CO ^ CM CM CM ^ •. oi as OCO^S cm o n m f^T<«00 u u ai «J « C3 saa CM CM <•;> CM CM ^ „ „ „ O O oooocooooocoooooeo' ! t'^ , Tt'T}>'>o < n<'9 , Tf'>T_ )< w « <1 4J 3 T3 o g *; a) ,q ft C3 OX5 a ° ft h B * ©£ 22 V Vh ’>>>> £ 3 O £? J5 24 Wisconsin Research Bulletin 65 giving the best results. The manure caused an increased growth on the unfertilized plots. However, with the best fertilizer plots' the manured and unmanured series were about the same. The manure had caused very little increased growth over that produced by the fertilizer alone. During the latter part of July a count was made of the number of hills in silk and in tassel. This indicated that maturity had been hast- ened to a considerable extent by fertilization. The checks' on the un- manured series were about 10 to 25 per cent in tassel, while the fertiliz- ed plots were 75 to 100 per cent in tassel. On the manured series the checks were 75 to 100 per cent in tassel and the fertilized plots 100 per cent. The most silks and tassels were found on the plots receiving the complete fertilizer, 2-12-2. Acid phosphate hastened silking but not as much as when it was used with nitrogen or potash. With Silver King corn, 2-12-2 and 3-10-4 seemed to be hastening maturity to a very considerble extent. The season was very dry thus' giving an excellent opporunity to study the effect of drought upon corn which had been fertilized in the hill. If fertilization in the hill restricts root growth, it should also cause the corn to suffer from drought. There was a five-week period in July and August in which there was only one-tenth of an inch of rain. This was right at the time when corn uses the most water, and is the most critical period in relation to moisture. The lack of rain un- doubtedly reduced the yield of all plots, but there was not at any time any indication that the plots which received fertilizer in the hill suffer- ed any more from drought than the unfertilized plots or those which received fertilizer applied broadcast. The corn on all the plots suffered to some extent, but the ears' on the plots which received hill applica- tions were longer and better filled out than those on the unfertilized plots. Therefore, it seems evident that root growth was not restricted by fertilization in the hill. The results in the field are in entire agree- ment with those obtained in the greenhouse with the large cylinders. As a further test of the effect of fertilization on maturity, twenty of the plots were sampled for moisture determinations on September 28. The samples were obtained by taking the ears of every tenth hill in the plot. This gave a sample of from five to eight pounds of ear corn. The corn was weighed at once, then placed in sacks and hung up in a warm, well-ventilated room to dry. At the end of five weeks it was taken down and weighed and the percentage of moisture determined. The results are tabulated in table XI. The moisture content of the corn on the fertilized plots' was in al- most all cases lower than that of the adjoining check plots. The 3-10-4 fertilizer applied in the hill on the late corn caused the greatest decrease in moisture content compared to the check, decreasing it from 39.3 per cent to 23.7 per cent on the unmanured soil and from 37.0 per cent to 31.2 per cent on the manured. The form of nitrogen in the 2-12-2 fertilizer seemed to effect maturity some, as is indicated by the percent- age of moisture in the ear corn. The corn from plot 17 receiving 2-12-2 fertilizer with nitrogen in the form of (NH^aSQ* contained less moisture than that from adjoining plots receiving the same application of 2-12-2 fertilizer but with nitrogen in the form of NaN0 3 or blood meal. The nitrogen in the commercial 2-12-2 was not as effective in this respect as was the (NH^SCX The same relation holds on both the manured and unmanured series'. Acid phosphate alone hastened matur- ity some, as is indicated by the percentage of moisture in the ear corn. On the unmanured series 2-12-0 hastened maturity more than 0-12-2 or 0-12-4, while on the manured series the reverse order wa§ found. The results obtained by this method of indicating maturity gave the same general indications as a study of the time of silking. Fertilizers 25 The corn was cut and shocked by October 6. It was husked and the grain and stover weighed separately about a month later. The yield of ear corn and stover are given in table X. Discussion of Yields: When plots 4 and 20 are excluded, the yields of the check plots indicate that the field was fairly uniform. The yield on plot 4 is low due to the fact that when planting, the planter was not working well, and one row had to be replanted on June 1. Plot 20 gave a poor yield due to water injury. Acid phosphate gave relatively small increases of grain and stover. Applications in the hill gave better results than broadcast application. Commercial 2-12-2 gave excellent results when applied in the hill at the rate of 120 or 200 pounds per acre. Plot 9 on the unmanured series which received 120 pounds in the hill gave an increase of 17 bushels of corn and 820 pounds of stover. This is an increase of 45 and 43 per cent respectively over the unfertilized plots. On the manured series the in- crease was 14 bushels of grain and 900 pounds of stover. Theses results would be very profitable with normal prices for grain and fertilizer. Table XI. — Percentage of Moisture in Ear Corn With Treatments Indicated. All Fertilizer Treatments Included in This Table Were Applied in the Hill Above the Seed Plot No. Pounds per acre and kind of fertilizer Moisture in corn Unmanured series Per cent Manured series Per cent Golden Glow Corn 3 90, 0-16-0 : 34.4 38.9 4 Check 35.9 37.8 6 180, 0-16-0 31.8 32.2 8 Check 35.3 34.3 9 120, 2-12-2 commercial mix 33.5 32.9 16 Check 34.6 37.6 17 200, 2-12-2, N as Am. sulfate 27.7 30.4 18 200, 2-12-2, N as So. nitrate 32.3 33.2 19 200, 2-12-2, N as Blood meal 32.0 32.8 20 Check 32.4 38.7 43 Check 38.6 36.7 44 200, 0-12-2, commercial mix 33.4 32.7 45 200, 2-12-0, commercial mix 32.1 36.0 46 200, 0-12-4, commercial mix 35.5 33.8 47 Check 41.5 36.1 Silver King Corn 52 | Check | 39.3 40.2 53 j 200, 2-12-2, commercial mix 1 41.3 38.3 58 Check 39.4 38.4 59 200, 3-10-4, commercial mix 23.7 31.2 60 Check 39.2 35.6 Two hundred pounds of 2-12-2 gave results of the same order as 120 pounds. The results indicate that the 120 pound application on this soil is more desirable and profitable. It probabily furnishes about all the additional phosphorous the corn plant can utilize. Possibly if a fertilizer containing a higher percentage of nitrogen and potassium were used, heavier applications would prove more desirable. If the applica- tions are made broadcast or in the drill row, the rate of application must be increased. Of the forms of nitrogen in the complete fertilizer, ammonium sulfate or nitrogen in the commercial 2-12-2 gave the best results, but the differences were not large. 26 Wisconsin Research Bulletin 65 The 0-12-2, 2-12-0, and 0-12-4 gave very similar results. On the un- manured series the yields' of stover with these three fertilizers were practically the same, although the 2-12-0 did give a somewhat smaller yield of corn than the other two. On the manured series the grain yields were the same, but 2-12-0 gave 400 pounds less stover. With Silver King corn the best results were obtained with the 3-10-4. This gave an increase of 20.2 bushels of corn and 1,210 pounds of stover on the unmanured series. On the manured series the increase was' 17.3 bushels of corn and 760 pounds of stover. This fertilizer was also very effective in hastening the maturity, especially on the unman- ured series. Summary of Field Experiment with Corn in 1920 1. Germination on a silt loam soil was not appreciably retarded by the proper application of 120 or 200 pounds of mixed fertilizer in the hill. 2. Fertilizers, especially 2-12-2 and 3-10-4, decidedly increased the early growth of the corn. 3 Maturity was hastened a week or more by proper fertilization. The most marked effect was with 3-10-4 on the late corn. The 2-12-2 and acid phosphate also hastened maturity. 4. Applications in the hill gave better results than those made in the drill row or by the broadcast method. 5. Nitrogen as found in commercial 2-12-2 and especially as ammonium sulfate produced slightly better growth and higher yields than the other forms of nitrogen when used in a complete fertilizer. Field Experiments with Corn in 1921 In 1921 a field was selected on the University Hill Farm for carrying on fertilizer experiments with corn similar to the previous year. The soil is a silt loam in a good state of fertility.® Field Experiments With Corn in 1922 on Use of Fertilizer In 1922 fertilizer experiments with corn were conducted on several fields. One of these fields was on a fertile Miami silt loam on the Hill Farm, where the experiment of 1921 had been conducted. The field had been in clover the previous year and received an application of 12 tons of manure before being plowed for corn. The corn (Wisconsin No. 7) was planted May 20, and the fertilizer in the hill was applied with an attachment on the corn planter. Table XII gives the treatments applied and the yields. Hill applications' proved to be superior to broadcast applications both as regards growth during the growing season and final yields. August was very dry, and the corn on the fertilized plots ripened more rapidly than on the unfertilized plots. Early September rains gave the check plots quite an advantage over the fertilized ones, since the leaves of the plants on the unfertilized plots' were still green, and the corn was able to utilize the water, and thus make additional growth. However, “The yields in this test did not indicate any marked benefit from the use of fertilizers regardless of the method of application. There may have been several reasons for this. First, the soil was a silt loam in a high state of fertility, producing on some check plots over 70 bushels of corn to the acre, which may have been near the maximum possible, under the existing weather conditions of that season. While not conclusive, the results indicate that it does not pay to fertilize corn on a soil as fertile as this. It is possible, how- ever, that with a more favorable season and more intensive cultivation, it might be profitable to use fertilizer on this soil. Fertilizers 27 this additional growth did not overcome the lead secured by the fer- tilized corn earlier in the season. The corn was cut and shocked, and at the time of husking, the mois- ture content of the check plots was from 2 to 7 per cent higher than that of the fertilized plots, indicating an appreciable effect of the fer- tilizer on maturity. The increases due to fertilizer treatment were not large, but the best producing treatment which was 120 pounds of 3-10-4 in the hill, netted an increase in value of corn and stover of four to five dollars per acre. Broadcast application of fertilizer produced very little increase.. Lack of moisture was a limiting factor at certain stages, otherwise greater increases might have resulted from the use of fer- tilizers. A second experiment was conducted on a Plainfield sandy loam near Mazomanie, Wisconsin. 10 A third experiment was conducted on a Miami silt loam at the Men- dota State Hospital Farm. This experiment compared fertilizer applica- tion in the hill below the seed with application in the hill above the seed. The plots consisted of 100 hills each and the planting and fertilizer ap- plication were made by hand. The fertilizer was applied in a thin band 4 inches by 8 inches either one-half inch above or below the seed as the case required. On June 6 the fertilized corn was about twice the size of the unfertilized, and the application above the seed seemed to have a slight advantage. The results of this experiment are given in table XIII. The fertilized plots had been distinctly superior to the surrounding corn during the whole growing season, and the yields showed an increase of 22 bushels per acre due to fertilization. This experiment shows what a marked result may be produced by the use of a small amount of fertilizer if it is properly applied. In this case, it did not make much difference whether the fertilizer was applied above or below the seed. Since application above the seed is more easily made, especially with a planter, it is probably the more desirable method. A fourth experiment was conducted on a neutral peat soil on the Home Acres Farm near Madison. The potassium chloride was applied by hand around the corn hills as soon as the corn came up. The 0-10-10 fertilizer was applied with an attachment on the corn planter. The corn on the fertilized plots grew rapidly from the start, and produced an en- ormous growth of stalks. The unfertilized corn grew poorly from the start. The yields are given in table XIV. The results indicated a great lack of potash and phosphate in this soil, and show strikingly what a small amount of fertilizer may do in a case of this kind when applied properly in the hill* The cost of the 0-10-10 treatment which increased the yield 60.9 bushels per acre was about $2.60 per acre. A fifth experiment was conducted on a medium acid Carrington silt loam on the Kayser Farm near Madison. The field had been in oats the previous year. The Golden Glow variety of corn was planted. The fertilizers' were applied with an attachment on the corn planter. None j In j^ ls experiment, different fertilizer mixtures, different forms of nitrogen and different methods of application were compared. The fertilizer treat- ments were beginning to show by the first of July, but unfortunately, a long drought of five weeks occurred in July and August which caused water to be the limiting factor and only slight increases were produced bv fertilization, several important points were, however, suggested by this experiment. The soil was very acid, and it seems that the use of ammonium sulfate and potash saits in considerable amounts in the hill liberated so much soluble acidity that nitrification was markedly decreased which affected the corn adversely. This same effect had been noticed on other occasions when large amounts of potash were applied. A detailed consideration of this matter is given on page 50. e* eo co cm HH c/) X c w -2 1-J pq < H P u ^ w • 5? w g ot w c >r‘: C/3 C/3 ffi ffi C/3 C/2 <1 ffi X X3 8* 43 S O Fertilizers 29 of the treatments retarded germination appreciably. The different treat- ments and yields are given in table XV. The soil showed a decided response to phosphate fertilization. A 300 pound broadcast application of fertilizer did not produce as much increase as a much lesser amount applied in the hill. A number of the fertilizer treatments in the hill produced marked and profitable returns. Due to the fact that the soil was acid and in rather poor physical condition, and that the cultivation was insufficient to properly keep down the weeds, the yields were probably kept down and the results of fertilization decreased. A number of very interesting and important points were raised by this experiment, similar to those noted on page 27, relative to the effect of high amounts' of potash and ammonium saults in decreasing the yields under certain conditions. To illustrate : an 0-12-2 produced a vig- orous growth, while an 0-12-4 produced no response; when nitrogen Table XIII. — Yields of Corn Secured with Treatments Indicated in 1922 on Miami Silt Loam on Mendota State Farm Amount per acre and kind of fertilizer Method of application Yield per acre Bu grain Lbs. stover 180 lbs. 2-12-2 In hill above seed 54.6 2600 180 lbs. 2-12-2 In hill below seed ' 54.3 2525 diprlf 32.1 1550 was supplied with the high potash, the fertilizer produced a response ; in a 2-12-2 fertilizer, nitrogen as ammonium sulfate was superior to nitrate, while in a 4-10-2 the reverse was true. These results may be explained as follows : High amounts of potash salts or ammonium sulfate on acid soils liberate so much soluble acidity that nitrification and other bacterial activity is hindered. If available nitrogen is applied along with high amounts of potash, the detrimental effect of the high potash is overcome due to the crop not having to depend on nitrification for available nitrogen. Table XIV. — Yields of Corn Secured with Treatments Indicated in 1922 on Peat Soil on Home Acres Farm Amount per acre Method of application Yield per acre and kind of fertilizer Bu grain Lbs. stover Check 19.6 2395 100 lbs. KC1 In hill above seed 64.4 5005 135 lbs. 0-10-10 In hill above seed 80.5 5298 Field Experiments with Corn in 1923 on Use of Fertilizers In 1923, corn experiments to test the use of different fertilizers in the hill were conducted on a Carrington silt loam on the Kayser Farm near Madison, on a Miami silt loam on the Johnson Farm, also near Madison, and on a Plainfield Sand on the Bower and Walton Farm near Arena, Wisconsin. Unfortunately, the season was rather unfavorable for corn, and a frost September 13, killed all the corn before it was ripe. All three fields had been manured, and increases from the use of fertilizer with the best treatments ranged from 5 to 12 bushels per acre. Phos- Table XV. — Yields of Corn Secured With Fertilizer Treatments Indicated, in 1922 on Carrington Silt Loam, on Kayser Farm. (Abbreviations used in table are, N. for nitrogen; Am. Sul. for ammonium sulfate; So. Nit. for sodium nitrate; Bl. Ml. for blood meal; I. H. A. S. for in hill above seed; I. D. A. S. for in drill above seed; Be. for broadcast.) ooiflioiaoiflooifioiowoifliomiaoo o.o c ic w c in m Mn C!c>fflo«cn(Oin!OiMfflooMnot>oooo'*^ , fi«ooinoi>N^o OOiHOMriOCnTffflOO^t'CnMMONMMMOCSNOCIOOOlCiMn CSJfO®Hf®00-'tf®©CJCa©®©'^©in5rH®in©Clt^®®fNlrH®CO© ®36rHrHinMai>oe>6rH®LnC^CO?Co6ciiCCOC'io®iOo6'®»nCOt> ■^intoinininMimnio^mT)>inin^^Trininin^'ii''?'jininTf(Din «3W ‘o3 02 03 ‘02 02 02 02 03 02 02 02 02 02 02 02 02 << << oW Hai-si v s ii • o o 1 V V ^ Ca ® •'OM so 75 75 -£?43 43. M i ~ A 2 44 = 43 * 43 S 44 H 43 44 43 * 2 £ 44 £ o2 o - S y o 1—1 o ~ o o s ©4>o «®® 45®© £®® 2© S®°0 2© ■-«ooo43o lf3,i H < *' J '' : f43©ca.— cajM.C-^.fi^caca'fici UoirHOMt'U^NUM^C’ H nUnU TH n r iUn " -i s o © « - tj< pC tF ^ thUW^ T- ®o > .» a OTJO 43 CO TH 00 CO CM CO cm 03 .« o GO fc .^44 'S « 03 CS to „ o u S bfi 5- 03 be 03 Si c G «-5 fa g iq-d 73 G C .5 344 o eu o O © CO © © 44 o 44 O 44 CM 44 O 44 CO 44 CM r* i CM 44 CM 44 t CM 44 CM 0 01 43 © in u 03 43 o LQ o »00»0 M«'^t , »r;©t>i00©©T-icMM'^'00©«i-i NC«MC^CSC^C^C^fOCCMeQPOCOMCCtCfO'V'>r MecTfintoi^oo© 40 Wisconsin Research Bulletin 65 unfertilized were just beginning to turn yellow. The plots receiving 3-10-4 and 2-12-2, either with or above the seed, were a week earlier in maturing than the oats on the unfertilized plots. Acid phosphate advanced the maturity about four days, but nitrogen or potash alone had no effect. Potash with phosphate had a greater effect than phos- phate alone or with nitrogen. The fertilizer treatments that hastened the ripening of the grain were the same as the ones that caused it to head out earlier. Discussion of Yields: The plots were sampled July 24 and 25. The bundles were weighed and threshed August 24. Table XVI gives the yield of grain and straw secured with the fertilizer treatments indicated. The increase in yield over the average of the two adjoining checks is also given. An examination of table XVI shows that in many instances' fertili- zation gave a considerable increase in yield and that the field was in a good state of fertility. The twenty-three unfertilized plots gave an average yield of 73.1 bushels of grain and 2,707 pounds of straw per acre. The average yield for the twenty-three fertilized plots was 81.5 bushels' of grain and 2,978 pounds of straw. This is an average increase of 8.4 bushels and 271 pounds. The average increase in the ten best fertilized plots was 12.4 bushels of grain and 302.3 pounds of straw. The yields of grain indicate that the application of a complete fertiliz- er above the seed is better than with the seed. These two methods of application were used with four different 2-12-2 fertilizers and in each case the application above the seed gave a greater increase than the ap- plication with the seed as is shown in table XVli. Table XVII. — The Effect of Form of Nitrogen in a 2-12-2 Fertilizer and Method of Application on the Yield of Oats when 200 Pounds" per Acre was Used Form of nitrogen Applied above the seed Applied with the seed Yield per acre Increase over check Yield per acre Increase over check Bu. Bu. Bu. Bu. NaN0 3 93.8 23.4 83.9 11.3 (NH 4 ) 2 so 4 86.3 14.6 79.1 9.6 Blood meal 83.8 10.2 77.0 1.9 Commercial 83.8 10.2 76.1 1.0 Table XVII also affords a comparison of the effects of different forms of nitrogen in a 2-12-2 fertilizer. The nitrogen in 2-12-2 fertilizer was supplied in four different forms ; namely sodium nitrate, ammonium sulfate, blood meal, and nitrogen as found in a commercial fertilizer. The nitrogen in the commercial 2-12-2 was mostly in an or- ganic form, probably as tankage. The sodium nitrate was the best form of nitrogen when applied either above or with the seed. Plot 24 which received 200 pounds of 2-12-2 above the seed with nitrogen as sodium nitrate, gave the largest increase of any plot. The increase in grain was 32 per cent and the increase in straw was 25 per cent. The ammonium sulfate, blood meal and nitrogen in commercial 2-12-2 rank after sodium nitrate in the order named when the increase in the yield of grain is the basis of comparison. The last two forms gave practical- ly the same increase. Fertilizers 41 The yields of straw do not show the same differences as the yields of grain. The form of nitrogen in the fertilizer seems to affect the yield of straw to a considerable extent. With the nitrogen as blood meal, the largest increase of straw was secured when the application was made with the seed. With the nitrogen as nitrate of soda, the largest increase was obtained when the application was made above the seed. The method of application did not affect the yield of straw greatly when the other forms of nitrogen were used. Plot 37 receiving 2-12-0 and plots 39 and 41 receiving 0-12-2 gave practically the same yield and about the same increase as was secured from the use of commercial 2-12-2. When the amount of potash was increased to 0-12-4 the results were somewhat poorer. Plots 43 and 45 which received 0-12-4 gave only an average increase of 4.1 bushels of grain and 73 pounds of straw. When the fertilizer constituents were used singly, acid phosphate gave the best results. Sodium nitrate gave a slight increase and potas- sium chloride alone reduced the yield of both grain and straw. This experiment indicates that fertilization of oats on a soil in a good state of fertility may be profitable, and that fertilization may have a number of desirable physiological effects on the plant. The results emphasize the importance of the proper method of application. The season was very favorable for the best growth of oats. There was an abundance of rain in June. The rainfall in July was light and the weather conditions were ideal for the ripening of the crop. No doubt the seasonal factor is important but just how important and the direction of its influence on an experiment of this kind is unknown at present. Therefore, one should be very careful in drawing conclu- sions from one year’s' results of an experiment of this nature. Experiment With Oats in 1921 This' experiment was conducted on a field adjoining the one used for the oat experiment the preceding year. 14 Experiments With Oats in 1922 In 1922 fertilizer test plots similar to previous years were carried on with oats at the Hill Farm on a Miami silt loam, at the Kayser Farm on a Carrington silt loam, and at the Mendota State Hospital on a Miami silt loam. On the Hill Farm the untreated plots yielded 60 bushels per acre which was probably near the maximum production possible under the weather conditions of the season. As a result, fertilization produced only a slight increase. On the Kayser Farm the oats rusted very badly just as they were filling, and the yields were greatly reduced. However, the fertilized 14 The rotation on this field is clover, corn, and oats. The field was in a higher state of fertility than the one used in 1920, having been quite heavily manured the previous year. Thirty-one plots were laid out in duplicate, using much the same plan of treatments as the previous year. A late variety of oats was planted April 5. At no time during the growing season was there an appreciable difference in the growth of the oats on the fertilized and unfertilized plots. The oats on all plots made a very rank growth during the early part of the growing sea- son. Later, the season became very hot and dry, and the oats did very poorly. They started lodging over the entire field about the middle of June. Then they were attacked by rust which became very bad. By the latter part of June they were lying flat on the ground. The heads filled scarcely at all. As there were no differences apparent and since it would have been practically Impossible to sample for yields the plots w'ere abandoned. 42 Wisconsin Research Bulletin 65 plots were farther advanced when the rust struck and produced from 5 to 7 bushels more per acre. Sodium nitrate proved to be a better source of nitrogen than ammonium sulfate. On the Mendota State Hospital Farm the check plots produced around 30 bushels’ per acre, and 240 pounds of 3-10-4 applied in the drill row increased the yield to 43 bushels per acre. The results of the three oat fields may be summarized as follows : Nitrate nitrogen proved superior to ammoniacal nitrogen ; broadcast application did not give as good results’ as application in the drill row ; and there was little difference between application above or with the seed in the drill row. Experiments With Oats in 1923 In 1923 experiments similar to those of the previous years were con- ducted with oats on two fields. The season was unfavorable and only on one field were the returns with the best treatments sufficient to pay for the cost of the fertilizer. Phosphates hastened maturity on one of the fields. This season, nitrate nitrogen did not prove superior to am- moniacal nitrogen. FERTILIZER EXPERIMENTS WITH CABBAGE ON THE EF- FECT OF DIFFERENT METHODS OF APPLICATION ON GROWTH AND YIELD Experiment With Cabbage in 1919 In 1919 an experiment was conducted with cabbage on one-thirtieth acre plots on a peat soil at Home Acres Farm in cooperation with L. P. Hanson. The field had never been cultivated or fertilized. Pot ex-, periments with corn indicated a decided response to phosphorous and potassium. The field selected for the experiment seemed to be very uniform and was well drained. The experiment was planned to compare broadcast applications of fertilizer with drill row applications as is made by a cabbage trans- planting machine. The only fertilizer used was a commercial 0-10-10. Broadcast applications were made at the rate of 600 and 1200 pounds per acre. These applications were made just previous to setting out the cabbage and the fertilizer was thoroughly mixed with the surface soil by harrowing. Applications in the row were made with a fertilizer attachment on a transplanting machine at the rate of 150, 300 and 450 pounds per acre. This attachment drops the fertilizer in a continuous stream just in front of the shovel that opens the furrow in which the cabbage plants are set. The packing wheels then press the soil con- taining the fertilizer around the plant. This assures good mixing of the fertilizer with the soil thereby preventing an excessive concentration of the fertilizer about the roots of the young plant. The cabbage was set out June 16. The rows were three feet apart, and the cabbage was set about two to two and one-half feet in the row. There were about 160 plants to the plot. It was very hot at the time of planting, and the plants wilted rather badly. Notes on Growth: There was so much dry hot weather the latter part of June and in July that the plants got a rather poor start. How- ever, by the middle of July, a beneficial effect due to the fertilizer was 1 evident. The plots receiving 600 and 1,200 pounds of fertilizer broadcast seemed to be the best. By the latter part of the month there was a very decided difference between the fertilized and check plots. At this Fertilizers 43 time plots 1, 2, 5 and 6 were about equal. It seemed that the heavier applications in the row did not promote early growth as much as broad- cast applications, but shortly thereafter differences' were not apparent. As the cabbage headed out, it was evident that there would be marked differences in the number and size of the heads on the different plots. Table XVIII. — Yields of Marketable Cabbage, Percentage of Plants Forming Heads, and Average Weight of Heads, With 0-10-10 Fertilizer Treatment Indicated Plot Fertilizer treatment Yield per acre Percentage of plants forming head Average wt. of heads Lbs. Per Cent Lbs. 1 600 lbs. broadcast 7380 64.0 2.56 2 1,200 lbs. broadcast 9120 75.6 2.72 3 Check 4020 48.0 1.86 4 150 lbs. in row 6840 75.0 1.90 5 300 lbs. in row 9780 1 87.5 | 1 2.33 6 450 lbs. in row 6300 64.8 2.06 7 Check 2580 40.0 1.53 1 The cabbage was cut and weighed September 24th. Only the heads were harvested, but all heads were cut regardless of size. The heads were trimmed as 1 for shipment so the yields given represent marketable cabbage except that many of the heads were too small for the market. This was especially true of the heads from the check plots The number of heads harvested from each plot was determined, making it possible to calculate the percentage of plants forming heads and the average weight of the heads. The results are given in table XVIII. All treatments gave good increases over the check plots. An increase of the broadcast application from 600 pounds to 1,200 pounds per acre produced a decidedly better yield, higher percentage of heads formed, and a greater average weight per head. While broadcast applications gave good results, they were not so economical as those obtained by fertilization in the row. Three hundred pounds in the row gave the highest percentage of heads and also the highest yield per acre, but the average weight per head was lower than with broadcast fertilization. When the row application was in- creased to 450 pounds, the yield, percentage of heads, and weight per head decreased. This may have been due to too high a concentration of fertilizer about the plant roots. Before conclusions can be drawn regarding this question further experiments are necessary. Experiment With Cabbage in 1922 In 1922 another fertilizer test with cabbage was made on the same farm as in 1919. The soil was a neutral peat that had just been brought under cultivation. Broadcast applications of fertilizer were made be- s ri 0) os .« U < fe sl s C/3 « ^ Sh •5J 4J « {* ft P£ U < w -■d Sr* cs S xfl ££ o 2 ffi O > Jh Z £ ^ ft LJ cn rt ft U W X3 h a < o hh O M-t H . Z U WPQ H *3 ^ £ tf g W ^ N , h - d Pi w c !_ W O H ^ iSl* o M < . < < S Kl P ;*» a; cs 3TJ i> ^ C « O o> cfl as * £>ft i'' "* u c „ « o 2 E q « ^ C'O C5 •< fO t* c w a> o qT S CS •*-> O fS cfl 3 •h a a 55 ’O « o © r4 in in ~ c4 ni Tf t'- CO CO IM TH * z bC tc Fertilizers 53 plication of fertilizers in the hill or drill row is sometimes desirable in order to supply the special needs of certain crops and thus promote early growth, hasten maturity, and increase yields. This use of fertiliz- ers in the hill or drill row, together with closely related problems, has been the subject of this investigation ; and the results may be sum- marized as follows: Methods of Application and Effect on Germination 1. — When considerable concentrations of fertilizer are applied near the seed, such as occur when 100 to 200 pounds are applied in the hill for corn, the fertilizer should never be allowed to come in direct contact with the seed, otherwise germination may be greatly delayed or entirely prevented. Fertilizer should always be mixed, at least some- what, with the soil, and in no case should it be dropped and left in masses. 2. — Field peas, cowpeas, navy beans, and soybeans are very sensitive to fertilizer in contact with the seed. Application of fertilizer about inch above the seed, the same as for corn, is probably the best method in the case of hill and drill row application. 3. — Squash, cucumbers, watermelons, and muskmelons are quite sensi- tive to fertilizers and, in case of hill application, it is best to apply the fertilizer about 54 inch above the seed. 4. — Germination of lettuce and carrots is affected and delayed most by application of fertilizer near the surface. For these plants it is best to work the fertilizer into the soil several inches. 5. — Potato sprouts are very much more sensitive to fertilizer than corn sprouts, and it is believed that this is, at least partly, due to the lower osmotic pressure of the potato sprout sap than the corn sprout sap. Best methods of applying fertilizer to corn and potatoes are described under experiments with these crops. 6. —' There are at least three factors, inherent in the seed or sprout itself, which determine the amount of effect of fertilizers on the ger- mination of seed ; viz., osmotic pressure of the seed, osmotic pressure of the sprout, and amount of protective covering on the sprout. 7. - — The experiments with corn agree with the results of other in- vestigators that the germination of seeds is affected by fertilizers more easily on sandy soils than on peats and heavier soils, and more easily at low than at high moisture contents. Corn Experiments 1. - — The experiments indicate that under Southern Wisconsin condi- tions, if a soil will not produce more than 40 or 50 bushels of corn per acre due to lack of essential elements and not moisture, hill or drill row application of fertilizer will probably be profitable; but if the yield is 60 to 75 bushels, a profitable increase from fertilization becomes questionable and usually can be obtained only under conditions of in- tensive cultivation and favorable weather conditions. 2. — Of the fertilizer mixtures used in hill fertilization, the 2-12-2 and 3-10-4 mixtures have given the best results. Undoubtedly, 2-14-2, 3-12-4, and 2-12-6 mixtures would give equally good results. As a source of nitrogen for this purpose ammonium sulfate was somewhat superior to the other forms. 3. — An application of 120 pounds per acre in the hill was a satisfactory rate. In the case of all except sandy soils, the rate may be increased to 200 pounds per acre without danger of injuring germination, providing the fertilizer is properly applied. If the corn and fertilizer are drilled, 300 to 400 pounds per acre may safely be applied. 54 Wisconsin Research Bulletin 65 4.— The best method of hill application is to apply the fertilizer in a band about 4 inches wide and 8 inches long from y 2 to y A inch directly above the seed. The fertilizer should never be dropped directly in contact with the seed. Many of the attachments on corn planters do not apply the fertilizer properly and this situation has probably pre- vented better results in practice, more than anything else. 5— When fertilizer is applied as directed it promotes early growth which facilitates early cultivation. Maturity may in some cases be hastened from one to two weeks. 6.— Numerous greenhouse and field tests have given no evidence that hill fertilization restricts root growth. 7— Experiments indicate that medium heavy applications of fertilizer, especially when applied in the hill so that the fertilizer is in position to produce an immediate effect, may lower the freezing point of young corn plants one to two degrees and thus prevent freezing. This effect would be greatest with poor soils and less with the more fertile soils. Oat Experiments T — The experiments were usually conducted on fields in a good state of fertility and the returns from the use of fertilizer have not been uniformly profitable. ' 2. - — In the season of 1920, which was very favorable for oats, the un- fertilized plots produced an average of 73 bushels of oats per acre and the best fertilizer treatments increased the yield 10 to 20 bushels per acre, hastened maturity, and lessened lodging. In this case, even though the yields on the unfertilized plots were high, fertilization when properly made was profitable. 3. — There are indications that nitrogen in the form of sodium nitrate is better than the other forms for oats. 4. — Application of mixed fertilizers in the drill row has given better results than when broadcasted. 5. — Application of fertilizer in the drill row just a little above the seed is probably better than mixing right with the seed, although ap- parently 200 pounds of mixed fertilizer per acre can be applied so as to mix with the seed without appreciable injury. This probably applies to the other small grains as well as oats. 6. — The maturity of oats was hastened more by the addition of phos- phorus than potassium or nitrogen. The combination of phosphorus and potassium hastened the maturity of oats most of all. 7. — Under dairy farming conditions it does not seem probable that the addition of mixed fertilizers for oats on most soils will prove pro- fitable. Weather conditions so often become the limiting factor in oats production that the addition of mixed fertilizers often does not pay as regards' the oat crop, but might if subsequent crops are con- sidered. Cabbage Experiments 1. — In the case of a peat soil, 300 pounds per acre of 0-10-10 fertilizer applied in the row increased the yield of cabbage more than when 1,200 pounds was applied broadcast. Application in the row thus seems to be more economical than broadcast application for cabbage. Potato Experiments 1. — Application of fertilizer below or to the side of the potato seed usually results in a much better stand and yield than application with or directly above the seed. 2. — Potato sprouts are very tender and are easily injured if they come in contact with high concentrations of fertilizers as occurs with applications directly above the seed. Fertilizers 55 Ammoniacal and Nitrate Nitrogen 1. — In quartz cultures, sodium nitrate produced in all cases as good or better growth than ammonium sulfate. 2. - — Any salt which contains the ammonium radical seems to be toxic to oats in quartz cultures. 3. — With field applications the conditions are quite different, since in this case the ammonium radical combines largely with soil acids and its concentration in the soil solution is greatly reduced and toxic action prevented. In this way excessive leaching is prevented and through nitrification, gradually the nitrate form is produced to supply the plants as needed. Under certain conditions the ammoniacal form may thus be more economical than the nitrate form. 4. — In case of a concentrated application near the seed, it is probably best to apply the nitrogen in several forms : some nitrate for immediate use, some ammoniacal nitrogen but not enough to be toxic for a little later use, and some organic nitrogen for still later use. Nitrate nitro- gen is not fixed in the soil and herjce its use above a certain amount in humid regions is certain to result in a loss. Secondary Effects of Fertilizers 1. — Acid phosphate when applied in high concentrations greatly stim- ulates nitrification and may in this way produce a decided secondary effect on plant growth. 2. — Potassium chloride increases the hydrogen-ion concentration of acid soils, and may in this way when applied in heavy amounts decrease plant growth due to lessened nitrification or other effects. 3. — The secondary effects 1 of fertilizers are probably much more im- portant than usually appreciated, and a further study of these would probably help greatly to answer many perplexing questions regarding fertilizer usage. LITERATURE CITED (1) Allison, F. E. 1918 Some .availability studies' with ammonium phosphate and its chemical and biological effects upon the soil. In Soil Sci., V. 5, p. 1 — 80. (2) Brigham, R. O. 1917. Assimilation of organic nitrogen by zea mays and the in- fluence of bacillus subtilus on such assimilation. In Soil Sci., V. 3, p. 155-95. (3) Brown, P. E. and Gowda, R. N. 1924. The effect of certain fertilizers on nitrification. In Jour. Am. Soc. Agron., V. 16, p. 137-146. (4) Buffum, B. C. • 1899. (a) Alkali : Some observations and experiments. In Wyo. Agr. Expt. Sta. Bui. 29, 1896. (b) Alkali Studies, III Pub. as part of Wyo. Agr. Expt. Sta. 9th Ann. Rpt. (5) Coe, D. G. 1922. Fertilizing the potato crop — What is the best way to do it? In Hints to Potato Growers, N. J. State Potato Assoc., V. 2, No. 10. (6) Gowda, R. N. 1924. Nitrates and nitrification in field soils. In Soil Sci., V. 17, p. 333-342. (7) Harris, F. S. 1915 Effect of alkali salts in soils on the germination and growth of crops. In Jour. Agr. Res., V. 5, p. 1 -53. (8) Hicks, G. H. 1900 The germination of seeds as affected by certain chemical fertilizers. In U. S. Dept. Agr. Div. Bot. Bui. 24. (9) Hutchinson, H. B. and Miller, N. H. J. 1911 The direct assimilation of inorganic and organic forms of nitrogen by higher plants. In Centr. Bakt. Abt. II. Bd. 30, p. 513-47. (10) Hutcheson, T. B. and Wolfe, T. K. The effect of fertilizers on the germination of seeds. In Va. Agr. Expt. Sta. Ann. Rpt. for 1918-19, p. 33-37. (11) McCall, A. G. 1917 A new method for harvesting small grain and grass plots. In Jour. Amer. Soc. Agron., V. 9 , p. 138-40. (12) Pantanelli, E. and Severini, G. 1910 Experiments on the use of various ammonium salts as plant food. In Staz. Sper. Agrar. Ital., V. 43, p. 449-544. (13) Rudolfs, W . 1921 Effects of salt solutions having definite osmotic concen- tration values upon absorption by seeds. In Soil Sci., V. 11, p. 277-93. (14) Rusche, A. 1912 Beeinflussung der Keimfahigkeit verschiedener Kulturp- flanzen durch Salzdiingnung. In Jour, f, Landw., Bd. 60, p. 305-65. (15) Salter, R. M. and Mcllvaine, T. C. 1920 Effect of reaction of solution on germination of seeds and on growth of seedlings. In Jour.' Agr. Res., V. 19, p. 73-95. (16) Sherwin, M. E. 1923 The effect of fertilizers on germination and seedling growth. In Jour. Amer. Soc. Agron., V. 15, p. 66-73. (17) Sigmund, W. 1896 Uber die Einwirkung chemischer Agentien auf die Kei- mung. In Landw. Vers. Sta., Bd. 47, p. 1-58. (18) Slosson, E. E. and Buffum, B. C. 1898 Alkali Studies : II. In Wyo. Agr. Expt. Sta. Bui. 39. (19) Slosson, E. E. 1899 Alkali Studies IV. In Wyo. Agr. Expt. Sta. 9th Ann. Rpt. (20) Stewart, J. 1898 Effect of alkali on seed germination. In Utah Agr. Expt. Sta., 9th Ann. Rpt., p. 26. / UNIVERSITY OF ILLINOIS-URBANA 630.7W75RE C002 RESEARCH BULLETIN MADISON 51-65 1921-25 3 0112 019935995