f c A.\n. 107 3'. QO /M l v ' l 1 ^/ "^ Issued November 8, 1916. PORTO RICO AGRICULTURAL EXPERIMENT STATION, Do W. MAY, Agronomist in Charge, Mayaguez, P. R. Bulletin No. 20. EXPERIMENTS ON THE SUPPOSED DETERIORATION OF VARIETIES OF VEGETABLES IN PORTO RICO, WITH SUGGESTIONS FOR SEED PRESERVATION BY C. F. KINMAN, Horticulturist, AND T. B. McCLELLAND, j£Q££*3\ Assistant Horticulturist. rJc^*** X% UNDER THE SUPERVISION OF "^ v/?f*iiLrtf& STATES RELATIONS SERVICE, Office of Experiment Stations, DOCUMENTS D EPT. U. S. DEPARTMENT OP AGRICtfiLTTTKE. U.S. DEPOSITORY WASHINGTON: GOVERNMENT PRINTING OFFICE. 1916. Issued November 8, 1916. PORTO RICO AGRICULTURAL EXPERIMENT STATION, D. W. MAY, Agronomist in Charge, Mayaguez, P. R. Bulletin No. 20. EXPERIMENTS ON THE SUPPOSED DETERIORATION OF VARIETIES OF VEGETABLES IN PORTO RICO, WITH SUGGESTIONS FOR SEED PRESERVATION BY C. F. KINMAN, Horticulturist, AND t. b. McClelland, Assistant Horticulturist. UNDER THE SUPERVISION OF STATES RELATIONS SERVICE, Office of Experiment Stations, U. S. DEPARTMENT OF AGRICULTURE. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1916, PORTO RICO AGRICULTURAL EXPERIMENT STATION. [Under the supervision of A. C. True, Director of the States Relations Service, United States Depart ment of Agriculture.] E. W. Allen, Chief of Office of Experiment Statioyis. Walter H. Evans, Chief of Division of Insular Stations, Office of Experiment Stations. STATION STAFF. D. W. May. Agronomist in Charge. P. L. Gile, Chemist. C. F. Kinman, Horticulturist. R. H. Van Zwaluwenburg, Entomologist. E. W\ Brandes, Plant Pathologist. T. B. McClelland, Assistant Horticulturist. J. O. Carrero, Assistant Chemist. C. Alemar, Jr., Clerk. (2) 53548 LETTER OF TRANSMITTAL. Porto Rico Agricultural Experiment Station, Mayaguez, P. R., October 19, 191-5. Sir: I have the honor to submit herewith a manuscript on Some Experiments on the Supposed Deterioration of Varieties of Vegetables in Porto Bico ; with Suggestions for Seed Preservation, by C. F. Kinman, horticulturist, and T. B. McClelland, assistant horticulturist. This paper gives the results of several years' study on the question of the supposed deterioration of varieties of vegetables when grown through several generations in the Tropics. The information con- tained herein should be of great benefit, not only in Porto Rico, but throughout the Tropics generally, or wherever similar conditions exist. In connection with these investigations it was found that the season of planting affected the yield and character of vegetables to a remarkable degree. The data presented will, it is believed, go far toward developing vegetable growing, an industry that is greatly neglected in Porto Rico. Vegetable seeds of all kinds deteriorate rapidly in the hot, moist atmosphere of the Tropics, and a method has been developed whereby the viability of seed can be retained for- a much longer period than is possible under normal conditions. This information will doubtless prove valuable to seedsmen and planters generally. I respectfully recommend that the manuscript be published as Bulletin No. 20 of this station. Respectfully. D. W. May, Agronomist in Charge. Dr. A. C. True Director States Relations Service, U. S. Department of Agriculture . Washington, D. C. Recommended for publication. A. C. True, Director. Publication authorized. D. F. Houston, Secretary of Agriculture. (3) CONTENTS Page. Introduction 5 Practices followed in the experiments 6 Records of the crops 7 Beans 10 Okra 1$ Tomatoes 18 Lettuce 23 Preserving the viability of vegetable seed 26 Summary 29 ILLUSTRATIONS. Page. Fig. 1. Monthly rainfall at Mayaguez. P. R., for the years 1910 to 1914 8 2. Average yields of bean generations 14 3. Yields of different bean generations as affected by planting season. ... 15 4. Okra yield as affected by planting season 18 5. Average yields of okra generations 19 6. Average yields of Royal Red tomato generations 21 7. Average yields of Royal Red tomato generations, showing seasonal variations 22 8. Ail-- tight jar illustrating method of preserving seed over calcium chlorid. 27 9. Results of germination tests of seed 28 10. Loss of viability of imported vegetable seed exposed to the open air. . . 29 11. Loss of viability of home-grown vegetable seed exposed to the open air. . 30 (4) EXPERIMENTS ON THE SUPPOSED DETERIORA- TION OF VARIETIES OF VEGETABLES IN PORTO RICO, WITH SUGGESTIONS FOR SEED PRESER- VATION. INTRODUCTION. Vegetable gardening in Porto Rico is a small industry and one in which there has been little progress for many years. Aside from the culture of tomatoes, okra, beans, pigeon peas, and starchy root crops, which are grown as horticultural crops, the farm garden is seldom seen. The commercial gardens are few and small and devoted to the culture of few vegetables, these often of inferior type. The poor development of the vegetable industry results from the unfavorable soil and climatic conditions, a lack of knowledge regarding the value of the best varieties and concerning good seed and its storage, and the production of seed at home. There is a belief that n orthern -grown vegetables degenerate quickly in Porto Rico and that the seed from even the first generation produces plants which are inferior to their parents. This belief results partly from the common practices in vegetable growing. On account of the heavy summer rains, the crops which are grown from seed are for the most part planted in the fall or winter. If seed harvested from these plantings is immediately sown, the plants as a rule give poor results, as they must contend with heavy rains, intense heat, and other unfavorable conditions. The unsatisfactory development and yield of the plants are naturally but unfairly attributed to the rapid degeneration of the seed resulting from the effect of the tropical climate. Because of the high humidity, seed deteriorates rapidly in the Tropics if exposed to the open air, and, therefore, much of the seed offered for sale by local dealers is almost valueless. Imported seed is often several months old when it arrives here, and it soon loses its viability, leaving the home product to supply the demand. The defective methods employed in collecting, curing, and storing seed render much of the seed harvested in winter useless for planting the following winter, and seed produced during the rainy season, when seed production is very low, often must supply the demand. In 1910 experiments were undertaken to determine the degenerat- ing influence of the climate of Porto Rico on the growth and produc- es) tiveness of a few garden crops imported from the North. During the progress of the work, observations have also been recorded regarding seasonal effects on the production of different plants, methods for storing seed, and tests made to ascertain the period of viability of seed imported from the North and of seed grown in Porto Rico. The results of this work should prove valuable in changing the common belief and practices which have practically stopped progress in the cul- ture of some vegetables. PRACTICES FOLLOWED IN THE EXPERIMENTS. In undertaking this work seed of a number of common garden vege- tables, including tomato, pepper, okra, beans, lettuce, radish, beet, squash, and cantaloup, which had been grown in the North, was imported. When the seed first arrived, a portion of it was placed in closed jars which were kept dry by placing in the bottom a few ounces of calcium chlorid. (See fig. 8, p. 27.) The seed treated in this way re- mained in first-class condition until the experiment was concluded or until the seed was exhausted by planting or testing otherwise. Throughout the experiment, the crops were grown in the open field under normal field conditions, and on a site very similar to those employed for gardens in the vicinity. This field is in an almost level valley where the soil is a heavy dark loam which is fairly fertile, but has little vegetable matter and poor subdrainage. To avoid damage from surface moisture and to insure soil aeration, the land was well stirred, and raised beds with a space of 1 foot between them, were made just previous to each planting. As plantings of each successive crop could not be made in the same place, there was danger of influ- encing the yield by soil variation between the different plats, although all the land used for the experiment was apparently very uniform. To overcome as much as possible any variation in the soil fertility, a heavy and uniform application of fertilizer was given each crop. The fer- tilizer contained 3 per cent nitrogen, 9 per cent phosphoric acid, and 10 per cent potash. It was applied at a rate slightly in excess of 1 ton per acre. In addition, a uniform dressing of stable manure was given the beds. The nitrogen in the form of nitrate of soda was applied after the plants were well established. The quantity of fertilizer was intended to be in excess of the needs of the plants so as to prevent minor differences in the natural soil fertility of the different plats influ- encing the growth of the plants. This would make the climate responsible for the growth and productiveness of the plants of different generations. As the imported seed and seed from succeeding genera- tions were planted in beds which were side by side, the variation in soil conditions could have had little influence on the yield. In order to eliminate as far as possible the factor of varying weather conditions, simultaneous plantings were made of as many different generations as possible. Care was taken that the crops should not degenerate nor be improved by seed or plant selection or by crossing with other varieties. At the time of planting a much larger number of seed was sown than the number of plants desired for the experiment. The young seed- lings were later thinned, leaving the desired number of plants of average size and vigor. Where there was danger of blossoms being fertilized from other varieties, the blossoms which were to produce seed for future plantings were hand-pollinated and bagged. The term season in this discussion is used with reference to periods of rain and drought, as the temperature in Porto Rico varies by only a few degrees throughout the year and is always conducive to the growth of vegetation. The seasons of rain and drought and the rain- fall for the different months during which the experiments herein con- sidered were in progress are shown graphically in figure 1 . By com- paring the rainfall and the production of the crops included in this experiment it will be seen that the variation in the amount of rainfall in different months is probably responsible for much of the variation in yield between different times of the year and therefore must be considered in calculating the results of the harvests. As the rainfall varies considerably in different localities in Porto Rico, the results secured with reference to the effect of climate on the production of vegetables are applicable to sections where the distribution of rain- fall throughout the year is similar to that at Mayaguez. The results which were obtained during the few years' work with vegetables are given below in detail. In order that the system of naming may be understood, a word of explanation is necessary. Each planting of northern-grown seed is denoted by a Spanish ordinal number, as Primera, Segunda, Ter- cera, indicating the first, second, and third plantings, followed by 1st, indicating first generation or northern seed. In each instance the plants raised from seed of these plantings bear the same Spanish ordinal as the parent planting, followed by 2d, indicating the genera- tion, and so forth, lineal generations being always denoted by the same Spanish ordinal. In some instances several plantings other than first generations were made from the same lot of seed. These are distinguished by letters following the English ordinal. RECORDS OF THE CROPS. After the work had been in progress for only a short time it was seen that a few of the vegetables included in the experiment were not suited to the work, as climatic and soil conditions were not favorable 22 2/ 20 /9 /e /z /6 /s /4 /s I" ^ /o s 8 6 5 3 2 / O km \ \ is ^ $ - /Q K AZ7/Ly /s// • • n -/a/3 -/3/4 i\ 1 I i \ 1 A r \ 1 1 \ 1 •V \ i • - I r i /A v > i /I \ 1 1 i /A) v/ K x \ \ i \ \ fi\ '/ 1 '/ 1 / A 1 \\ 1 A : 1 | \ i V / M A // I '/ 1/ 7 i / / V 4 / / ' M / to \ ,/J i \\ \ \ r" / \\ \ . H 1 y / ^ r / 1 \ \\ i i 1 V \ \ \ \ 1 pi / / / ;/ 1 \ \ 1 -J rh // X 7 / 1 1 \ X 7 / / / V i Fig. 1.— Monthly rainfall at Mayaguez, P. R., for the years 1910 to 1914. to seed production. Cantaloups and squash were not grown after the first season on account of troublesome insects and weather con- ditions, which cause a very poor yield and would probably influence results by the necessity of planting, at least from later generations, seed which was not from average and representative plants and fruits. Beets and radishes were not grown beyond the first generation, as attempts to produce seed failed. The leaves and roots made a thrifty growth, but the plants all died without sending up seed stalks. While these plants occasionally produce seed in Porto Rico, the crop is so uncertain and the yield so small that for the present at least it is necessary to import the seed required for these crops. With peppers the Ruby King variety was planted, but it was impos- sible to get data of size and weight of fruits, as they were nearly all attacked by rot before they matured. The plants grew poorly and produced such a small crop, as compared with the kinds commonly grown throughout the island, that it was apparent that seed of this type of peppers should not be imported. Further tests were made, however, to compare the growth and yield of the types commonly grown in Porto Rico with those of a number of similar imported varieties, including Ruby King, Chinese Giant, and Tabasco, to determine the advisability of producing seed from them. Plantings of these varieties were made in the spring when the weather favored a good plant growth, but later the rainfall was so heavy that con- ditions were unfavorable for growth and blossoming. Tbe imported varieties made a rapid growth at first, but before fruit matured, the leaves began to fall, and the plants became barren stalks which did not regain their vigor. The average production per plant of Ruby King and Chinese Giant was between three and four marketable fruits, and that of Tabasco was small also. The growth of all the common Porto Rican types was vigorous from the first, and they continued thrifty and productive for several weeks after the northern varieties were dead, their average production being over 16 fruits per plant. The fruit from the common Porto Rican type resembled very closely in size and shape the best northern kinds and was not inferior to them in flavor. The types tested included both sweet and pungent varieties. As the ancestors of the seed imported from the North for this test were doubtless taken from tropical America many years ago, these results show that the marked change effected by acclimatization and selection to suit northern conditions resulted in rendering them unfit for tropical conditions. Results with other crops discussed here show that where seed or plant selection is not practiced the place effect on these crops is too slow to be of importance to gardeners. 59836°— Bull. 20—16 2 10 BEANS. The beans were of the Extra Early Valentine bush variety, grown in New York for J. M. Thorburn & Co., of New York City, and ob- tained in February, 1910. All plantings were made from this importa- tion except Segunda, which came from seed received September, 1911, and Octava, Novena, and Undecima, from seed received January, 1913. The seed was planted in two parallel rows about 6 inches apart and 25 feet long. Each bed was thinned to 130 to 135 plants. In some instances weather conditions retarded germination, and some plants did not appear until after the thinning, which usually took place within two weeks after planting, thus making a few more plants than the experiment called for. In other instances disease entered and reduced the number of plants after thinning to a number below the average of the experiment. The results of these plantings are shown in Table I. Table I. — Bean generations. Date of planting. Name. Number of pro- ducing plants. Number of days from planting until blossom- ing. Average number of pods per plant. Average number of beans per pod. Average number of beans per plant. Number of beans contain- ed in \ liter. Porto Rico plantings. Mar. 12, 1910 Primera 1st Primera2d Primera 3d Primera 4th Segunda 1st Tercera 1st Primera 5th Cuarta 1st Tercera 2d Primera 6th Quinta 1st Cuarta 2d . 137 61 137 126 112 127 124 148 131 138 127 135 131 127 116 103 119 86 117 57 73 88 84 88 95 53 73 84 95 71 77 79 71 90 79 137 145 146 32 31 35 37 30 30 31 36 36 36 32 34 34 34 34 34 35 35 34 35 35 34 34 35 35 32 32 32 32 32 32 32 32 32 32 9.31 3.02 6.03 5.97 4.22 4.48 4.99 3.57 2.88 3.91 11.17 10.78 11.99 5.10 6.12 5.12 5.17 2.73 2.26 2.79 1.86 2.19 2.11 2.01 2.22 11.89 10.21 8.64 8.58 9.51 9.10 8.14 9.80 9.30 9.22 14.67 12.10 12.79 3.34 2.64 2.78 2.24 2.61 2.61 2.29 2.13 1.97 2.19 3.21 3.18 3.21 2.60 2.78 2.56 2.64 2.02 2.01 2.08 1.76 1.91 1.S4 1.67 2.10 2.98 3.23 3.28 2.91 3.03 2.52 3.05 3.20 3.15 3.18 3.41 3.41 3.41 31.1 8.0 16.8 13.4 11.0 11.7 11.4 7.6 5.7 8.6 35.9 34.3 38.5 13.3 17.0 13.1 13.6 5.5 4.5 5.8 3.3 4.2 3.9 3.4 4.7 35.4 33.0 28.3 25.0 28.8 22.9 24.8 31.4 29.3 29.3 50.0 41.2 43.6 1 1,411 Sept. 17, 1910 Dec. 24, 1910 1 1,559 Mar. 11,1911.... 1 1,412 Sept. 16,1911... . 1 1,994 Do 1 1,966 Do 1 1,862 Dec. 22, 1911 1,420 Do 1 1,693 Do 1 1, 443 Mar. 12, 1912 1,268 Do 1,226 Do Primera 7th Octava 1st Minn. A 1st Minn. B 1st Minn. C 1st Novena 1st Decima 1st Quinta 2d B Octava 2d 1,226 Jan. 17, 1913 1,610 Do 1,560 Do 1 1, 455 Do 1,512 Nov. 8, 1913 Do Do Do Do Minn. A 2d Minn. B 2d Minn. G 2d Primera 8th B Undecima 1st Duodecima 1st Novena 2d Decima 2d Quinta 3d Do Do Do Mar. 26, 1914 1,230 Do 1,302 Do 1,213 Do 1,228 Do 1,273 Do... Octava 3d 1,234 Do Minn. A 3d Minn. B 3d Minn.C3d Primera 9th Sexta 1st 1,268 Do 1,264 Do 1,205 Do 1,238 Minnesota plantings. June 1, 1912 1,009 Do Quinta 2d A Primera 8th A 1,080 Do... 1,111 1 Estimated. 11 The first planting of beans, Primera 1st generation, was made March 12, 1910. The season was quite favorable for this crop, and the yield per plant averaged 9.31 pods of 3.34 beans each, or 31.1 beans per plant. From this crop two plantings were made during the following June. The first, made on June 11, was removed on account of injury from an insecticide. The second, planted on June 25, was almost completely destroyed by anthracnose and very wet weather, so that from 135 plants only 6 pods were picked. As these plantings were complete failures, they are not included in the table of bean genera- tions. A third planting of Primera 2d generation was made September 1 7. Even though the season was much drier than usual, with weather conditions seemingly exceptionally favorable, the plants were badly diseased and did not fully develop their seed, many being obtained in the yield which were not sufficiently developed for either eating or planting. The producing plants gave an average yield of only eight beans per plant, showing a marked falling off in number of both pods per plant and beans per pod. Primera 3d generation, planted December 24, gave a yield which more than doubled that of the parent generation, but it still amounted to little more than half of the yield of the original planting. The planting of Primera 4th generation was made March 11, 1911, a year later than the original planting. The yield amounted to less than half that of Primera 1st generation. In the light of further tests it is assumed that this planting fell below the usual production of plantings at this season because the amount of moisture was below the optimum, the rainfall for March, May, and June of 1911 being less than that for these months in any other year during the term of the experiments. As the original seed had been kept in the drier, some new seed, Segunda 1st, of the same variety, was obtained from the same seed house and planted simultaneously with seed of the original lot, Tercera 1st and Primera 5th, seed from the fourth generation. Fresh seed was procured for comparison with seed kept in the drier, since the seed for Primera 5th generation was quite fresh and could not be strictly compared with the old dried seed without a check on the latter. This planting was made September 16, 1911, and had to contend with very wet weather. There were 37 days with rain, 29.39 inches of which fell between planting and the first picking. The development of many pods and beans was stopped at an early stage by wet weather and disease fostered by such weather. The differences in average yield per plant of beans of sufficient development to be counted were so slight as to fail to show any degeneration, since the fifth generation yielded 0.4 bean more per 12 plant than one first generation and 0.3 bean less per plant than the other. Since the old seed kept in the drier gave a slightly larger crop than the fresh seed, the intense drying had not injured it. The number of beans per pod in the two first generations was the same. In the fifth generation it was less than in the first two, but this was balanced by the greater number of pods per plant. The next planting, which was made on December 22, 1911, con- sisted of a first, a second, and a sixth generation. In contrast with the preceding weather, the rainfall was entirely too small for a good crop, there being but six days with rain and the total precipitation amounting to only 1.68 inches from planting until the first picking of mature beans. As in the preceding case, the advanced generation gave the greater average number of pods per plant. The pod of Primera 6th generation averaged also a slightly greater number of beans than its contemporaries, which is in contrast with the behavior of Primera 5th. The second generation showed the smallest average number of pods per plant and beans per pod. The size of the beans of the second generation was also below that of its contemporaries. The next planting was made March 12, 1912. It consisted of a first, a second, and a seventh generation. In no other Porto Rican planting did the average yield of beans per plant for any generation whatever equal the average yield of this entire planting. The ad- vanced generation, Primera 7th, led with an average production of 38.5 beans per plant. This was followed by Quinta 1st generation with 35.9, and Cuarta 2d generation with 34.3 beans. Here it is seen that the order corresponds with that of the preceding planting, the advanced generation in the lead, followed by the 1st, and lastly by the 2d generation. In order to observe any evidence of changes which had been pro- duced in this crop by being grown in the Tropics for from one to several generations, it was decided to make several plantings in the North. For this purpose seed were sent to Minnesota. Through the courtesy of Dr. J. N. Tate, of the Minnesota School for the Deaf, plantings were made June 1, 1912, at Faribault, Minn., of Sexta 1st generation, being seed of the original lot kept over calcium chlorid, of Quinta 2d generation A, and of Primera 8th generation A. The beds were made side by side under very uniform conditions of soil and light. The soil was a friable, rich, black loam, now well drained but having formerly been swamp land. All received the same treat- ment, all blossomed about the same time, and all were picked on the same date. All made a strong, vigorous growth and were much taller and more luxuriant than any plantings in Porto Rico. Though all were thrifty and vigorous, those whose antecedents had been grown for seven generations in the Tropics showed a little less pronounced vigor than the other two, between which no distinction could be made. 13 The average number of beans per pod was identical for all — that is, 3.41. In number of pods per plant the 1st generation, plants whose forbears had never been grown in the Tropics to our knowledge led, followed by the 8th generation, and lastly the 2d generation, hi yield of beans, the average crop per plant ran 50, 43.6, and 41.2 beans, respectively, yields never equaled in any of the Porto Riean plantings. Though a counting of the number of beans of the original lot which would be contained in one-half liter was never made, new seed re- ceived from the North January, 1913, ran 1,103 beans. Those grown in Minnesota ranged from 1,009 to 1,111. It may be assumed from these four counts of northern-grown seed of this variety that the original lot of northern-grown seed ran somewhat the same. The seed harvested from Primera 7th generation ran 1,226, and from Quinta 1st, 1,268, beans to the half liter. These Porto Rican grown seed, then, were considerably smaller than the northern-grown seed, and this in itself might account for the larger crop from Sexta 1st. 1 The next planting in Porto Rico was made January 17, 1913. It consisted of four 1st generations in that the parent generation of each had been grown in the North. Octava 1st was from new seed just received from the North; the other three were from the Minnesota crops of the previous summer, Minn. A 1st from Sexta 1st, Minn. B 1st from Quinta 2d A, and Minn. C 1st from Primera 8th A. Minn. A 1st produced an average of 1 7 beans per plant, while the others ranged from 13.1 to 13.6 beans. It is thought that perhaps Minn. A 1st was in a slightly moister location than its contemporaries, as no other explanation for this difference can be seen. The next planting was made November 8, 1913. As a whole, it gave a lower average than any other lot of plantings. It consisted of two 1st generations, five 2d generations, and one 8th generation. The average yield of the two 1st generations was 5 beans, of the five 2d generations, 4.1 beans, and of the 8th generation, 4.7 beans per plant. The 2d generation average yields ranged from 3.3 to 5.8 beans per plant, including thus the smallest and the largest average yield. Such yields show the very marked seasonal effect, but little else. It is interesting here to compare the yields of two plantings made from the same lot of seed, Quinta 2d A, planted in Minnesota and yielding an average of 41.2 beans per plant, and Quinta 2d B, planted in Porto Rico and yielding 5.8 beans per plant. The final j)lanting was made March 26, 1914. It consisted of two 1st generations, two 2d generations, five 3d generations, and one 9th generation. The number of plants in all rows was somewhat reduced by a Sclerotium blight. Except for this, conditions were i See Vermont Sta. Bui. 177 (1914). 14 very favorable. The two 1st generations gave the largest crops, averaging 33 and 35.4 beans per plant. This may be accounted for, as mentioned before, by a difference in the size of the seed, since in no instance whatever did any Porto Rican grown seed equal in size those from the North, as indicated by the number of beans counted in one-half liter. The 2d generations averaged 26.6 beans per plant, the 3d generations 27.4 beans, and the 9th generation 29.3 beans. Though the 9th generation averaged more than the average of all the 3d generations, one of the latter equaled it and another exceeded it in yield. In nearly all plantings some plants produced either no pods or pods with beans of too small a size to be considered. These non- productive plants varied considerably in number, but in no very uniform or consistent manner. They are not included in the table or the preceding discussion, but deserve mention. In .the Septem- ber, 1911, planting of the two 1st generations, one produced 19 such plants and the other only 6, while the 5th generation included only Fig. 2.— Average yields of bean generations, the number of beans per plant being indicated by heavy lines, with lineal generations connected by dotted lines. The appended dates are those of planting. 4. In the next planting of a 1st, a 2d, and a 6th generation, 5 non- productive plants were included in each. In the November, 1913, planting these plants ranged from 2 to 5 in the 1st generations, and from 2 to 26 in the 2d generation, with 20 in the 8th generation. Though these tests did not afford sufficient data for a systematic classification of this tendency, more of such plants were observed in the fall plantings than in plantings made at other seasons. Two diagrams are appended in order that the results of these plant- ings may be clearly shown. In figure 2, the yields of the lineal generations are shown, and it is seen that the only uniformity indicated is a large yield at certain seasons and a poor yield at others. No indi- cations of degeneration are seen. In the other diagram, figure 3, the points determining the curve of the 1st generation were obtained by taking an average of all lst-generation plantings made in any one month. These points were then connected by lines, thus forming a curve showing the yields of lst-generation plantings at different seasons. The yields of the 2d and 3d generations are shown in the 15 same way. Of each of the more advanced general ions only a single planting was made. These are -indicated by crosses. This indicates very clearly the main factor governing the yield. In considering the many plantings that have been made, it is seen that the matter of a large yield or a small yield of this particular crop is not influenced by the progenitors having been previously grown in the Tropics for a long time or a short time or not at all, but is chiefly a matter of the planting being made at the proper season. At certain I ^ « !? S s.4 sr o ^ iO S 3 S 3 $ § § s $ 40 36 36 J* 7 £4 \ee \'* \ /6 \ /o e 6 2 O ^ ^ ^ 3 * 5 I 1 K *7& 1 A 1 \: rjd )sy / k* /? *3 ^ -'73 /M ay !/ J.' * *2T/Y X ~ 3 ■w>c f ^ b w N s _, Si \ so j- | so \ \ 40 so eo /o o >*^"? j AH f*» : /*MA>4 vj ~tf- \ ««» - --; '< ■•'-" ■£f-'-= !VW'M c 1 ^ Aaete >-M-2- We ~ - 2 --- J -■ J '-■ .i 9 Si S S » 8 » /e- /& -Average yields of Royal Red tomato generations, the number of fruits being indicated by heavy lines and the lineal generations connected by dotted lines. their development materially decreasing the yield of this row. The other rows gave an average yield of 27 to 32.3 tomatoes per plant, the 4th and 2d generations leading in number of fruits produced. In size, the tomatoes of the 2d generation were the largest and those of the 4th generation were the smallest. Excepting Primera 6th C, the yield of the poorest row planted at this date exceeded the best yield of any other planting. In figure 6, the number of fruits produced by each planting is dia- grammatically shown in connection with the number produced by its lineal forbears and its descendants. The planting dates are appended in each case. This diagram is so similar in form to that showing the total yield by weight of fruit produced by the average plant that the latter diagram is not included. These diagrams clearly show that there is no uniformity of behavior where the number of ancestral generations grown in the Tropics is considered as the main factor. This is particularly noticeable in the 22 Primera group, of which more plantings were made than of any other. The one striking feature is the grouping of the late winter and spring dates on the one hand and the autumn planting dates on the other, showing the planting date to have been the real factor in determining the yield. In figure 7 this is clearly demonstrated where all plant- ings made on the same date are connected by lines. The small yield of the 6th generation in the October, 1913, planting, which is the only thing to mar the symmetry of this chart, has already been explained. The tomatoes planted in the fall on the whole had a slightly larger average size than those planted in the late winter and spring, but this was not uniformly the case. However, the main factor seemed S? A? /o \ y \ i r ,^ / y' \ \ ; 1 1 l g I r'W-- I l I . I w 1 1 1 1 l Mill /£? 3"? 4™ /sr 5>» /^ r 2Z° S w 6^ /& 2*-° 3-° 6V 7^ /& 2%.° 3& 4?? 6^ 7& A ^ A. 1 Af/lff'/O/W*'// OCT'// MAR /2, /3/? ttPrtAfSCff /4, /3/2 f£-5/?0/lf?y 24, /S/S OC7DS£R 27 /S/S. Fig. 7.— Average yields of Royal Red tomato generations, the number of fruits being indicated by heavy lines and simultaneous plantings connected by dotted lines. to be the time of planting. The number of ancestral generations which had been grown in the Tropics failed to show any effect on size. An analysis of the whole subject shows that, though tomatoes from northern seed were grown for seven generations in the Tropics, this change of habitat apparently had no effect on either size or number of fruits produced. The one factor influencing production which rose far above all others was the season at which the planting was made. Honor Bright. — The seed of Honor Bright tomatoes was from the 1910 crop grown in California for J. M. Thorburn & Co., of New York City. In tests of this variety, eight plants were set 4 feet apart in the row. The general treatment followed was the same as that for the Royal Red variety, except that less manure was used in the preparation of the beds. The results of the plantings are given in Table IV. 23 Table IV. — Tomato Honor Bright generations. Date of planting. Name. Num- ber of days from plant- ing to blos- som- ing. Num- ber of days from blos- som- ing to fruit pick- ing. Num- ber of days cov- ered. Num- ber of plants. Aver- age num- ber of fruits per plant. Aver- age weight of fruit. Horizontal diameters of average fruit. Ver- tical diam- eter of aver- age fruit. Dec. 10,1910 May 24, 1911 Primera 1st Primera 2d A Primera 2d B Primera 3d 57 82 60 60 48 196 45 42 208 278 161 210 6 8 8 7 51.2 .3 3.1 90.7 Grams. 104.6 13.0 78.9 104.9 Inches. 2.57 by 2.37 Inches. 1.77 Mar. 12, 1912 Dec. 10, 1912 2.20 by 2.01 2.43 by 2.40 1.53 1.80 The first planting was made December 10, 1910. The fruits were large to medium in size, and the average production per plant was 51.2 fruits. The first planting of the 2d generation was made May 24, 1911. By the middle of August the plants began to blossom. In early October it was noted that blossoms were seen only very rarely and that these did not set fruit. By this time one plant had died, and two more were nearly dead. Near the end of February of the follow- ing year, two tomatoes, one weighing 18 grams and the other 8 grams, were produced on plants which were then entirely leafless, only one plant still having green leaves. This demonstrated very forcibly the uselessness of attempting to produce tomatoes in the season of heavy rains. A second planting of the 2d generation was made March 12, 1912. The fruits produced were small in size and few in number. They, however, furnished sufficient seed for a planting of the 3d generation. The 3d generation, which was planted December 10, 1912, far excelled any other planting. The average size of the tomatoes was about the same as that of those of the 1st generation, but the number produced was 90.7 fruits per vine. The total weight of fruit per vine was about 21 pounds. The yields showed the seasonal effects in even more pronounced contrast than did the plantings of Royal Red tomato, though fewer plantings were made. The data serve to substantiate the conclu- sions reached in the tests of the other variety. LETTUCE. The lettuce plantings were from seed of Prize Head lettuce, grown in California and received in the Congressional seed distribution of 1909. The young plants were transplanted to individual pots 10 to 13 days after sowing. When sufficiently large, two or three weeks later, 25 to 30 plants were set in the field, 1 foot apart in the row. 24 Where a plant was injured so as seriously to retard its development, or varied decidedly from the others, it was removed. Alternate plants were left for seed, and the others were cut to supply the desired data. In the plantings of January, 1914, no plants were reserved for seed. The beds were cut at the first indication of a blossoming stalk starting up. Where one or several dried leaves were found, these were discarded and the head then weighed. The remaining leaves were then counted, including all an inch or more in length. Except- ing the plantings on the last two dates, measurements were taken of the length and breadth of the 12 outer leaves on six average heads of each planting, the length being measured along the midrib, and the breadth at the widest portion of the leaf. The results are given in Table V Table V. — Lettuce generations. Date of planting. Name. Number of days from planting to cut- ting. Average weight of head. Average number of good leaves. Average length of leaf. Average breadth of leaf. Mar 12 1910 Primera 1st 63 75 76 72 66 66 72 75 75 74 74 Grams. 169.2 97.9 54.2 124.4 195.0 223.1 63.3 199.4 136.1 127.8 109.2 24.7 19.5 18.7 27.3 29.3 30.8 18.0 28.9 24.6 25.1 24.2 Inches. 6.84 5.59 4.66 5.84 6.04 6.05 Inches. 6.55 Jan 2 1911 Primera 3d 6.40 July 6' 1911 Primera 4th 4.86 Dec 13 1911 Primera 5th A 6.45 Mar 12 1912 Segunda 1st 6.82 6.95 Nov 7 1912 Tercera 1st Mar 27 1913 Tercera 2d A . . . Jan 17 1914 Tercera 2d B The first planting, which was made March 12, 1910, gave a satis- factory yield. Seed from this generation were planted repeatedly, but sufficient plants for a test were never obtained. This same condition of poor seed with a low percentage of viability was found to be quite common where plants blossomed and seed was produced in rainy weather. Four 2d generation plants were finally obtained, but only one of these lived to blossom. From this plant seed ripened in December and was used for Primera 3d generation, which was planted January 2, 1911. Though this generation was left uncut 12 days longer than the 1st generation, the plants produced fewer and smaller leaves and weighed considerably less. Ripe seed was collected from these plants in May and June. Primera 4th generation was planted July 6, 1911. In number and size of leaves and weight of head it fell far below the 3d generation. Primera 5th A, which was planted December 13, 1911, produced a much better growth than either of its two nearest ancestral genera- tions, but in weight it did not equal the 1st generation. 25 Two years after the original planting a simultaneous planting was made of a 1st generation and a 5th generation. Owing to very wet weather, very few and poor seed were obtained from the 4th genera- tion, and the young seedlings from this generation were much slower in germinating and, as young plants, were vastly inferior in size to the plants of the contemporaneous 1st generation. However, before maturity the average size of plants of the 5th generation was con- siderably larger than that of the 1st generation plants, not only show- ing no degeneration, but excelling in size the 1st generation plants, even though handicapped by a poor start. The leaf measurements of these two simultaneous plantings showed no change to have taken place in the general shape of the leaves. The 5th generation showed a tendency toward earlier blossoming than the 1st generation. On account of rainy weather no seed was obtained from these plantings. Seed from 5th generation A and seed of the original lot were sent to Washington and planted at Arlington Experimental Farm through the kindness of the horticulturist, D. N. Shoemaker. He reported in letter of December 31, 1912, as follows: * * * On July 3 it was noted that the 6th generation lettuce in comparison with the other two was decidedly less frilled and not quite so deep in color. On July 13 it was noted that the 6th generation showed smoother edges and lighter color, but was larger and more vigorous in plant than the 1st generation or the Amer- ican seed. On July 30 it was noted again that the lettuce of the 6th generation was certainly not so deep in color and was less frilled to the edge of the leaf than the 1st generation. Also it was quite notable that it was tenderer than either of the other numbers planted for comparison. On September 4 we noted on some plants winch had been lifted and planted in a frame that the plants of the 6th generation grown in Porto Rico were larger and more vigorous than the others but were rather inferior in varietal character. Would say in summing up the observations that the difference between the lettuce of the 6th generation and first generation from Porto Rico seems very likely to be that which would result from a selection of the varieties in any region without a strict adherence to the varietal type; that is, that the difference observed here would very readily occur in the same number of generations on any American farm if the seed- bearing plants each year were not strictly compared with an ideal varietal type. Several other plantings were made in Porto Rico after this date. On November 7, 1912, another planting of 1st generation was made. The heads were next to the smallest of any produced. Measurements of the average length and breadth of the 12 outer leaves on six average heads showed that in the 1st generation the length and breadth of leaf were not in uniform relation, since on three heads the sum of the leaf lengths exceeded the sum of the breadths in each case, while the reverse was true with the three others. This lack of uniformity was also found in Primera 1st but did not appear in Primera 5th B or Segunda 1st. Its appearance in two 1st genera- 26 tion plantings showed it too unstable a factor to be regarded, so further leaf measurements were not taken. Two more plantings of 1st generations were made simultaneously with 2d generations, and in each instance the 1st generation gave the better yield. In only the earlier of these two plantings were plants left for seed. In this instance the 2d generation showed a tendency toward earlier blossoming than the 1st. In reviewing the various lettuce plantings, no degeneration caused by being grown for a number of generations in the Tropics is evident. On the contrary, indications point toward an improvement in the advanced generation. The production of well-formed and viable lettuce seed in some seasons here is practically impossible, and for this reason it will usually be found best to import seed for this crop. The difference in yield of 1st and 2d generations, when planted simultaneously, may or may not have been due in part to a difference in vigor caused by a difference in size of seed. As with the other crops previously discussed, the prime factor nfluencing production seemed to be the planting season. PRESERVING THE VIABILITY OF VEGETABLE SEED. The humidity of the air in Porto Rico is very high and causes vegetable seed exposed to the open air to lose their viability much sooner than would be the case in a drier atmosphere. The inability to keep seed in good condition is a serious hindrance to vegetable growing. When the experiments herein reported were undertaken, the following method for preserving seed was employed and was very satisfactory. The seed in cotton sacks was placed in air-tight glass jars in the bottom of which a few ounces of calcium chlorid had been placed. A small piece of wire screening separated the seed from the calcium chlorid below. (See fig. 8.) This method is simple and costs little, and is recommended for general use. While glass jars were used in experiments, metal or nonporous earthen vessels will serve as well if made air-tight. It must be remembered that the calcium chlorid placed in the bottom of the jar is used as a drying agent and not as a preservative in any other sense. Seed such as coffee and citrus, which lose their viability on drying, can not be kept viable in this way. If seed with a fairly high water content is to be stored, it may be necessary to renew the calcium chlorid, since, unless a sufficient quantity of calcium chlorid is used to take up the surplus water, the seed may not be kept sufficiently dry. Before the calcium chlorid becomes entirely moist, it should be replaced by a fresh supply. In handling it should be exposed as short a time as possible to the open air, since it takes up moisture from the air readily and so loses 27 its drying power. In removing seed from the container this should be remembered. Germination tests were made of seed of a number of different vege- tables, a Geneva seed tester being used. To bring out more clearly the results of these tests, they have been plotted as curves. The native tomato seed of which tests are recorded in figure 9 was col- lected in the winter of 1910. The other seed was imported, coining Fig. 8.— Air-tight jar illustrating method of preserving seed over calcium chlorid. presumably from the 1909 crop. A portion of each was kept in the drier and another portion was exposed to the open air. In the last germination test of bean seed which had been kept in the open the seed was put in the tester October 29, 1910, germinating 4 per cent. On this date, 100 per cent of the seed kept in the drier germinated. This seed from the drier still showed a 100 per cent germination nearly two years later, August 13, 1912, when the tests had to be discontinued on account of a scarcity of seed. 28 /&/o /3// /9/2 /&/3 /3/4 /S/S Fig. 9.— Results of germination tests of seed. The unbroken line shows the germination of seed kept in a closed jar containing calcium chlorid as a drier. The dotted line shows the germination of seed from the same lot kept in cotton sacks exposed to the air. 29 By the end of 1911 all of the imported seed kept in the open air had lost its viability. In 1915, 94 per cent of the radish seed and 84 per cent of the beet and tomato seed kept in the drier still germinated. Unfortunately, the supply of okra seed in the drier was too small to allow a long continuance of tests. On November 4, 1911, 100 per cent of the seed from the drier germinated, while of that in the open only 2 per cent germinated. It was found necessary before planting to file okra seed which had been kept in the drier. The native tomato seed, after more than five years in the drier, showed a germination of 93 per cent, whereas seed kept in the open lost all viability in less than half that time. In figure 10 the loss of viability of imported seed of a number of different vegetables is graphically shown. The very rapid loss of ZOO\ —T— £.. Fig. 10.— Loss of viabilit y of imported vegetable seed exposed to the open air. The tests begun Teb. 12 1910, were of seed presumably collected the preceding season; those begun Jan. 28, 1911, presumably were from the 1910 crop. The dates of the germination tests are given above. To eliminate as far a; possible errors of individual tests, the curves were plotted by connecting points midway between the germination percentages so as to occupy a mean position in respect to them. Tests not made were of kale Aug. 27, 1910; turnips, Dec. 31, 1910; okra, Sept. 9 and Nov. 4, 1911. viability of lettuce seed kept in the open air stands in marked con- trast with the results of tests of seed of the same lot which had been kept in the drier and which showed a germination of 90 per cent in 1915. In figure 1 1 the loss of viability of home-grown seed is shown. SUMMARY. That northern vegetables degenerate quickly when taken to the Tropics is a common belief in Porto Rico, resulting from the fact that seed loses its viability quickly when exposed to moist air and from a lack of knowledge regarding seasonal effect on vegetable production. To retain the viability of seed of the crops used in the experiments here reported, the seed was stored in air-tight jars in the bottom of which was placed a small quantity of calcium chlorid. This method was so satisfactory that it is recommended for general use. 30 Planting a few types of peppers, such as are commonly grown in Porto Rico, side by side with varieties imported from the North showed that the Porto Rican types are much more productive and therefore more desirable than imported varieties. Forty plantings of beans were made, including nine generations. Of the Porto Rico plantings, those made in March gave large crops, except one in 1911, which was hindered by an exceptional drought, while those made in other months, including June, September, November, December, and January, gave small harvests. No indi- cations that advanced generations were inferior to earlier ones were observed. Records of the five years' work with okra, during which time 32 plantings were made, reaching finally the eighth generation, show ^£S3 f "— -— . ^ -X- — , N? ^ 1 .. mm N N ^ \ — . -^ \ \ % ^ ^ \ % \ >9 \ V v *-„„ \ . "\ \ N -v **^v»^* !TT "*" =5fta4 A/L/MBER OF WEEKS *ETE& E>/C/