." ■ v,\v.. h. &* - X p U. S. DEPARTMENT OF AGRICULTURE DIVISION OF CHEMISTRY BULLETIN No. 19. A 7. 3; /? METHODS OF ANALYSIS OF COMMERCIAL FERTILIZERS, CATTLE FOODS. DAIRY PRODUCTS, SUGAR, AND FERMENTED LIQUORS, ADOPTED Al THE FIFTH ANNUAL CONVENTION OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, HELD AT THE U. S. DEPART- MENT OF AGRICULTURE AUGU^|r^9-T«iXlLjXL 1888. EDITKD BY CLIFFORD RICHARDSON, Sl'CHETAKV OF Till: ASSOCIATION. 4>- WASH INCTOX: GtOVERNMBNT PRINTING OFFICE. 18SS. 1 V ! Firman J. Golman, wommeddumel 0/ tSaatrat/tieie. U.S. DEPARTMENT OF AGRICULTURE. DIVISION OF CHEMISTRY. BULLETIN No. 19. METHODS OF ANALYSIS COMMERCIAL FERTILIZERS, CATTLE FOODS. DAIRY PRODUCTS, SUGAR, AM FERMENTED LIQUORS, ADOPTED AT THE FIFTH ANNUAL CONVENTION OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS, HELD AT THE U. s. DEPART- MENT OF AGRICULTURE AUGUST 9 AND 10, 1888. EDITED BY CLIFFORD RICHARDSON Si CRETAR1 Ol 1 HI A J80< !\ I ION. WA SB I NGTOK: GOVERN M l. .\ l PRINTING OFFICE 1888, it— linii r.» LETTERS OF TRANSMITTAL. Department of Agriculture, Division of Chemistry, Washington, D. C, September •!, 1888. Dear Sir: Complying with your invitation the Association of Offi- cial Agricultural Chemists met in the rooms of the Chemical Division August 9 and 10 this year. I have the honor to submit for your ap- proval, with a view to publication as Bulletin 19 of this division, the official proceedings of that meeting, as attested by the inclosed letter from the secretary. Respectfully, H. W. Wiley, Chemist. Hon. X. J. Colman, Commissioner of Agriculture. Washington, August 31, 1S88. Dear Sir: I have the honor to hand you herewith, for publication, an abstract of the proceedings of the fifth annual convention of the As- sociation of Official Agricultural Chemists, together with the methods adopted for official use daring the ensuing year for the analysis of commercial fertilizers, cattle foods, dairy products, sugar and sugar pro- ducts, and fermented liquors. To tliis has been appended a list of the reporters upon the different subjects for 1888-'89, of the officers of the association, and the consti- tution as amended. The demand for information of the nature given in the proceedings is rapidly increasing, and the methods of the association are almost uni- versally adopted in this country as standards. Very respectfully, Clifford Richardson, Secretary, Association of Official Agricultural Chemists. Dr. H. W. Wiley, Chem%9t } etc. Digitized by the Internet Archive in 2013 http://archive.org/details/meanalysiOOedit UNIFORM METHODS FOR THE ANALYSIS OF FERTILIZERS, CATTLE FOODS, DAIRY PRODDI TS, SI GAR, AM) FERMENTED LIQUORS. PROCEEDINGS OF THE FIFTH ANNUAL CONVENTION OF THE ASSOCIA- TION OF OFFIi TALAGBIi ULTUBAL CHEMISTS, HELD AT WASHINGTON, AUGUST 9 AND 1". 1888. MORKING SESSION, THURSDAY. In accordance with the call of the secretary, the Association met in the library of the Department of Agriculture at 10 o'clock, the president, Mi. P. B. Ghazal, in the chair. The following members and ''others interested in the objects of the Association" were present: The president, Mr. I*. E. ChazaJ, State chemist of South Carolina. The vice-president, Dr. W. J. Gascoyne, of Baltimore. The secretary, Mr. Clifford Richardson, of Washington, D. C. Of'tjie executive committee, Prof. John A. Meyers, of the West Vir- ginia Agricultural Experiment Station. Dr. 11. W. Wiley, Washington, 1). 0. ; J>r. C. A. Crampton, Washing- ton, D. C.J Prof. William C. StUDDS, of Louisiana: Prof. B. A. von Schweinitz, of Salem. X. C; Prof. Richard II. Gaines, of Richmond, Va.j Prof. William Frear, of State College, Pa.j Prof. H. J. Patterson, of Agricultural College, Md.j Prof. W. L. Bntchinson, of Agricultural College, Miss.; Prof. G. 8. Fellows, of Washington, D.C.; Prof B. B. Vooi bees, of New Brunswick, N. J. j Mr. r>. B.Ross, of Louisiana; Mr. Edgar Richards, of Washington, I>. C. ; Prof M. A. Scovell, of Lexing- ton, Ky. : Prof. E. II. Fan Lngton, of Hanover, N. II. : Prof. C. W, Dab- oey, jr., of Knoxville, Tenn.; Prof. F. A. Bolton, of Iowa City,Iowaj Mr. \V. M. Saundera, of Providence, R. I.: Prof. B.C. White, of Athens, Ga. : Messrs. Knoir. Fake, Edson, and Dngan, of the U. S Department o( culture. Letters of regret were read from l>r. E. II. Jenkins, of Connecticut ; Dr. (i. 0. ('aldw.-ii, of N<-\\ fork ; Prof. ( F.Cook, of New Jer- Sey, and otic The president, in calling the meeting to order, said thai be would omit the usual address, as there was a desire to proceed at oner to busi- ness, * ith the \ iew of accomplishing the objects of the conventioo in as ahorl a time as possible. B On a call for reports of committees, Mr. Richardson, in the absence of the chairman, presented the report of the Committee on Cattle Foods, as follows : REPORT OF THE COMMITTEE OX CATTLE FOODS. Your committee offer the following report on the work done during the year on the analysis of fodders and feeding stuffs: Only six out of a dozen or more who undertook to make these analyses for the com- mittee were able to do the work. Iu many cases so much time was consumed, in the latter part of the year, in the reorganization of the experiment stations, under the provisions of the Hatch hill, that too little was left for these deferred analyses. More- over, some of the reports were received so late that no copy of the results sent by all the analysts could be sent to each of the parties doing the work, so that if any one so desired he might repeat the analyses with a view to some possible explanation of the discrepancies that occur. There is no reason to doubt that these analyses, the results of which are given be- low, were made by one method and with careful attention to the directions given by your committee, and by analysts at least quite up to the average, in respect to skill and special training, of those who commonly do this sort of work at the experiment stations, and yet there are some very serious differences iu the results. It appears to your committee, therefore, that there is occasion for more work on this line, and it may reasonably be hoped that with the large number of experi- ment stations in operation now a much larger number of participants in this work can be found who can undertake the analyses and complete them in season for a more careful preparation of the report by the committee having it in charge. Your committee would, therefore, suggest that, if the work is continued, the sam- ples, three in number, as hitherto, be sent out before the middle of October; that the analyses be made strictly according to one scheme, without, however, excluding alter- nate methods that any one may wish to try in addition to those prescribed by the com- mittee ; that the time for receiving the reports be limited to February 1 ; that a tabu- lated copy of all results received be sent out within two weeks thereafter to eaeli contributor, and that then a limited time be allowed, say one month, for any re- \ i -ion of his work that he may wish to make. As one illustration of such possible revision, the uniformly higher results reported in the first line of each list below, on nitrogen and protein, may bo due to some fault in the standardization of the solutions used fortho titration of the distilled ammonia solutions. Additional illustrations may occur to others on inspection of the figures. Only one report was received on any alternate methods. At the Connecticut station all the samples were analyzed also b\ the method in common use thero ; but no de- scription of the methods followed accompanied the results. These results given for the air dried BUDStance were iu Very many of the cases identically the same as those obtained by the committee method, and in nooase was i hero any serious difference; but as the moisture was in all caso nearly or a little over 1 peroent. higher, they would, if calculated to dry substance, lie BOmewhal higher lhan those obtained by t be committee method. The following interesting results were given also DJ the same station on different methods of determining the moisture and the tat iu the dried residue: fodder. Bran. Cotton st'fl nic.il. Moisture by the committee method I .it in i be aboi ■ ire at ion in -i ream <■! hydrogen ... i it in i be .iiw. e re at 1 1" 115° in -i ream <»i b] rirogen I .it m tin- ;ili'.\ i- 7. 88 I. 78 7. 73 1.68 7.84 4.00 !». 1 1 4. "J.-. 12. OS 12.42 7.03 12. SO Reported by— Moisture. In the dry substance. Name of substance. Ash. Ether extract. Fiber. Crude protein. X-free extract. Caldwell 5.80 6.67 6.56 11. S3 9.04 7.04 8.45 8. 35 7.64 11.10 9 90 8.46 6.76 6.98 6.43 8.37 7.60 6.14 10.59 10.34 6.60 8.53 8.7^ 9.48 7.52 7.34 7.01 6.39 6.16 6.: 9 9.23 9.C6 8.10 8 12 8.01 7.96 2.?5 2.15 1. 55 3.31 2.18 1.43 4.54 4.51 4. IT. 5.79 4.99 4.01 10.68 13.92 12. 88 14.71 13.76 12.28 29. 25 33. 79 29.90 31.40 27. ^8 24. il 10.86 9.97 840 9.64 9.10 7.71 6.70 4. 25 2. 86 3.56 3.27 2.51 10.90 10.59 9.90 9.79 9.88 9.56 19.57 17.78 18.13 16.90 17.96 10.75 50.9°. 47. 85 49.0 5 46.13 47.55 46.06 47.00 43. 13 Jenkins 52.05 46.97 Weber 51.29 Peter . . 58. l J 8 Caldwell , 5". 51 Fi ear Jenkins 60.40 62.21 61. 18 Weber 61.69 Peter 66.68 22. 46 24. 92 27. 13 27.48 Weber : 24.41 Peter 25. 05 In the dry substance. Corn fodder. Bran. Cotton-seed meal. Total. Albuminoids. Total. Albuminoids. Total. Albuminoids. 1.69 1.57 1.53 1.02 1.34 1.21 2.85 2.91 2.68 2.40 2.64 2.53 7. 65 7.84 7.37 6.20 7.64 Peter 7.00 Respectfully submitted. G. C. Caldwell, Chairman. Clifford Richardson, Wm. H. Jordan, CommHtee. Mr. Richardson said that, in view of the large amount of discrepancy in the results, he believed it advisable to confine fuhire work to one method without alternates. He added that, with the establishment of so many experiment stations during the present year, all of which would be interested in this subject, a very large increase in the number of analysts was to be expected, and the results at the next convention would admit of drawing more definite conclusions. He had found that in the alternate method for determining liber, recommended last year, the use of bottles could only be made a success when live steam was available. In a water-bath, owing to sudden and unequal heating, cracking seemed unavoidable. Dr. Prear found difficulty in seeming concordant results on cotton seed meal, and regretted the absence of all the analysts but himself, which prevented an interchange of experiences On motion of the latter gentleman, the report and recommendations of the committee were adopted. 8 Dr. Wiley then presented the report of the committee on dairy prod- ucts, as follows: RETORT OF THE COMMITTEE OS DAIRY ERODUCTS. By H. W. Wii.i v. Mr. President: I found it impracticable to secure a meeting of the committee ap- pointed at the last meeting to consider the subject of dairy products, and have therefore no report considered by all of the members of the committee to present. In the absence of such a document I thought ir lust to submit a brief resume of tin 1 work which has been doDe in dairy products during the year since our convention last met. The methods proposed at onr last meeting for the analysis of dairy products have given fairly good satisfaction, and I therefore recommend that they be con- tinued unchanged for another } ear. Many points which are now in dispute will have been settled by that time, and at our next annual meetingany necessary changes in methods of analysis can be made. With this introduction I beg to submit the following abstracts of the progress made in the analysis of dairy products during the year just passed. MILK ANALYSIS. UOE OF ASBESTOS CLOTH INSTEAD OP BLOTTING PAPER IX THE ADAMS METHOD. Johnstone has reported to the Society of Public Analysts favorable results attend- ing his experiments in substituting asbestos paper for blotting-paper in the Adams method offal estimation.— (Abstract from the Analyst. December, 1887, p. 834.) [NOTE.— I called attention to the use of asbestos paper for this purpose two yean ago, as will be seen on page 81 of Bulletin No. 13, Part 1.] AN INSTRUMENT Foil CALCULATING MILK RESULTS. Richmond has described an instrument for calculating milk analyses based on the relations existing between fat, solids not fat, and specific gravity in milk as estab- lished by the paper of Hehner and Richmond, published in the Analyst, vol. 13, p. 26. The instrument consists of a slide rule on one side of which a scale, of which one division equals 1 inch, represents total solids. The other scale, one division of which equals 1.164 inches, represents the fat; while the scale representing specific gravity has a division equal to 1.254 inches. The instrument is Itased upon the formula T — .854 (i = L.164 P. To use the instru- ment the lines indicating total solids and specific gravity found by analysis are placed together. The fat is then read off by an arrow <>u the other side. Where many analyses are t<» be made the instrument is especially valuable since ii elimi- nates all chances of error in calculation.— (Abstract from the Analyst, 1888^ vol. i:'». No. 1 U, p. 65.) ESTIMATION 01 i u in MILK and CREAM. Werner Bohmid proposes the following method of estimating milk fat : A test tube of about 50 cubic centimeters oapaoitj is graduated in tenths of a cubic oenti meter ; LO cubic centimeters of milk or 5 cubic centimeters <>f cream are placed in the tube and 10 cubic centimeters concentrated hydrochloric acid. The contents of the tube BTC bailed with constant shaking until the liquid is dark brown. After cooling in w ate r add 30 cubic centimeters ether, shake, allow to stand until the ether solution eparated, measure its volume and remove LOoubio centimeters with a pipette; place m ;i porcelain crucible, evaporate, and dry at 100° in air-bath. Calculate to total volume of the ether solution. Ph< results are said to be exact.— (Zeit. Anal, Cllem.. 87, p. H.I.) THE RELATION OF SPECIFIC GRAVITY, FAT, AND SOLIDS NOT FAT, IN MILK. Heliner and Richmond have made an exhaustive study of the relations existing be- tween the specific gravity, fat, and solids not fat, in milk. Three sets of determina- tions of the fat were made, viz, extraction from a paper coil, from plaster of Paris, and the direct extraction of the dried total solids. All previous formula; re- lating to the above relations have been based upon imperfect methods of fat extrac- tion. The work of Hehner and Richmond is therefore especially valuable on account of being based upon the Adams method of extraction. Forty-two analyses of various kinds of milk were made. For discussion of the formula' the original paper is cited. The working formula obtained is T — .254 G = 1.164 F. A table accompanying the paper g'ves the percentage of fat, calculated for specific gravities ranging from 1024 to 1034.0, and for total solids from 10 per cent, to 15 per cent., inclusive. The author strongly recommends that no milk analyses be accepted as correct which do not correspond closely with the calculated results.— (Abstract from the Analyst, vol. 13, No. 142, p. 2(5.) SOURCES OF ERROR IN SOXIILET'S METHOD OF ESTIMATING FAT IN MILK. Weinwurm has determined the error which may arise by allowing the ether fat so- lution obtained by Soxhlet's method to stand for varying lengths of time before its specific gravity is determined. As a result of his experiments it appears that the ap- parent percentage of fat is raised by allowing the solution to stand for a long time. For the first twenty-four hours, however, this increase is only small, the amount, at most, being a few hundredths of a per cent. By standing a long time the apparent percentage of fat gradually increases, and appears to reach a maximum at the end of ten days. The amount of iucrease in that time may be .25 of 1 per cent. — (Abstract from Repertorium No. 19 of the Chemiker-Zeitung, 1663, p. 151, and M. Zg. 1888, 17, p. 401.) ESTIMATION OF DKY SUBSTANCE AND FAT IN MILK BY MEANS OF WOOD FIBER. Ganttex proposes the use of wood fibe? for the estimation of fat in milk, which he a follows : Two grams of wood fiber are dried at 105 c to constant weight in a dish containing ;i small glass rod. The total weight of the dish, glass rod. and fiber is noted. After weighing the dish, which is covered with a neat-fitting cover to prevent the ab- sorption of water, 5 to (5 grams of the milk are poured upon the wood fiber and the exact amount taken determined by reweighing the bottle. During evaporation the iii-* — of fiber is pressed from time to time against the sides of the dish, so that no par- ticles thereof remain attached thereto. The final drying is made in an air-bath, and the whole time required Cot the evaporation and drying should be abonl two hours and a half. From the increase in weight the amount of dry substance is determined. The mas-, of fiber i^ now transferred to a Soxblet extract ion apparatus, and, if 116068- sary, the dish rinsed with petroleum ether. The extraction of fat ami the weighing thereof are carried on in the usual way. The author has also used this method for ih.) determination of water in butter and other fats and oils.— (Abstract from the Zeitschrifl fiir analytische Chemie, vol. 26, No. 6, p. •*»?-). THE 'Hi laCAL ACTION <-i SOME MM RO ORGANISMS in mi: k. Warington has studied the effect of certain micro-organisms on the curdling of milk< This curdling is effected either by the formation of a rennet-like ferment or by the production of lactic acid. The amount of lactic acid required t<> curdle milk depends on the temperature, the am it growing less as the temperature rises. Five different organisms : imineil which bail the power of cnrdling milk, but in very diffi 10 degrees. They were Staphylococcus candidus : B. termo ; M. gplatinosus, B. fluoresced liquescens, and M. urea. Some of them readily curdled milk at a temperature as low as 10°. They do uot, however, at that temperature produce an appreciable acidity. Thoy therefore act plainly as a ferment. A large quantity of gas is evolved during the action of the same on milk. Five organisms were also found to act as peptoniscrs: they are, B. snbtili*. B. unthracis, B. floccus, B. toruliformis, and Tinkler's comma. The milk treated with these organ- isms at 22° becomes clear after a few days. The clear fluid is rich in pepton. A few- organisms render milk after a time decidedly alkaline; two of these are, B. fluorescent von liquescens and the Bacillus of septicemia. Cultivation in milk is an excellent method of distinguishing micro-organisms. — (Ab- stract from the Chemical News, June 22, 1883.) REDUCTION AND ROTATION TOWER OF MILK SUGAR. Deniges aud Bonnans have reviewed the literature concerning the redaction of cop- per solution by milk sugar. As compared with pure glucose the reducing power of lactose is as 9G to 136. The copper solution employed contained 34.65 grams of pure crystallized copper sulphate and 10 cubic centimeters pure sulphuric acid per liter. The alkaline solution was composed of 250 cubic centimeters of soda lye of 3ti J strength, in which was dissolved 150 grams of crystallized Rochelle salts. Ten cubic centimeters of this solution was added to each 10 cubic centimeters of the copper solution. For the determination of the rotatory power of lactose the author used a pure sugar, obtained by three crystallizations. The specific rotatory power for the anhydrous sugar at 20° was 55.30 for concentrations varying from 4 to 36 per cent. ; for the crys- tallized sugar the specific rotatory power in the same condition is 52.53. The general formula for the crystallized salt would be [a] =52.53 + (20 — T) X -055. The number .055 is the correction to be applied for each variation of 1 D in temperature. The num- bers obtained by the authors are exactly those given by Schmoger. — (Abstract taken from the Journal de Pharmacio et de Chimie, April and May, 1888j pp. 303 and 411.) ESTIMATION OF SUGAR IX MILK BY THE POLARI8COPB. Vieth discusses the method adopted by the Association of Official Agricultural Chemists for the estimation of sugar in milk, which was originally described by me in the American Chemical Journal, vol. 6, No. 5. The author is of the opinion that acetic acid and subacetate of lead can be used with equal advantage with mercuric nitrate and nitric acid for preparing the milk for polarization. He, however, prefers the use of the mercury salt, lie calls attention to the correction which should be made in the reading of the polariscope for the volume of the precipitate. According t.» Vieth this collection in the method adopted by t he association is too small. The volume of the precipitate is equal to the combined vol nines of the precipitated casein and fat carried down with it. The method employed by With for making the eor- r ction for the volume of the precipitate is illustrated by the following example : Given a milk with a specific gravity of L.03'25 which contains 3.72 of fat. Add to 50 cubic centimeters of this milk u cubic centimeters of the mercuric nitrate solution. The polarization of the nitrate shows 5.1 milk Bugar. The accessary calculations fat the correction of this number are as follows: 93 : 100 :: 372 ! X whence \ = 4 the volume occupied by the fat. 103.8C : 96 :: 51 ; X whence X = 4.74. The number 96 in the above formula represents LOO oubic centimeters of the Liquid ; , ni.ie centimeters, the rolurae of the tat. The number 1.74 represents the per- 11 centage of crystallized milk sugar. The percentage of the anhydrous milk sugar would therefore be 4.5. — (Abstract from the Analyst, vol. 13, Xo. 144, p. G3. ) short's method of determining fat in milk. [Bui. VTis. Ag. Ex. Station, Xo. 1G.] The process depends on the following facts: That when a mixture of milk and a strong alkali is heated to the temperature of boiling water for a sufficient time the fat of the milk unites with the alkali and forms a soap which is dissolved in the hot liquid; at the same time the casein and albumen are disintegrated and become much more easily soluble. After the heating has continued for about two hours the mixt- ure of milk and alkali becomes homogeneous and of a dark brown color. On the ad- dition of an acid the soap is decomposed, the fatty acids are set free, and rise to the surface, while the albumen, casein, etc., are first precipitated and then dissolved. The insoluble fatty acids thus obtained constitute very nearly 87 per cent, of the total fat of the milk. APPARATUS, The process requires the following apparatus: (1) Tubes made of soft lead glass about one-sixteenth inch thick. The lower part of the tube is about 5 iuches long and fifteen-sixteenths of an inch in diameter. The upper part of the tube 5 inches long and one-fourth inch inside diameter. (2) Three pipettes, one holding when filled up to the mark on the neck 20 cubic centimeters (about two-thirds of an ounce), this being the exact amount of milk to be taken for analysis ; the other two pipettes, holding 10 cubic centimeters each, for measuring the alkali and acid used. (3) A scale, divided in millimeters, for measuring the column of fat when the analysis is finished. The one used by the writer is a folding boxwood rule, but any rule divided in millimeters will answer the purpose. (4) A water bath made of sheet copper. It is provided with a rack to hold the tubes while being heated; also a feed aud overflow to keep the water in the bath at a constant level. (5) A wash bottle to hold hot water. SOLUTIONS REQUIRED. The solutions required for the process are as follows: No. I. — 8.75 ounces (250 grams) caustic soda and 10.7ounces(300grams) canst ie potash dissolved in 4 pounds (1,809 grams; water. Use 1<) cubic centimeters for each analysis. No. 2. — Equal parts of commercial sulphuric and acetic acids. The acetic acid should be of 1.04? specific gravity. Ose 10 cubic centimeters of the mixed acids for each analysis. Dim t i [ONS rOB ANALYSIS. Taking tamples. — Mix the milk thoroughly by pouring from one vessel to another, avoiding as much as possible the -formation of air bubbles; warming the milk to 80 ..i '.mi degrees Fahrenheit will prevent frothing to a large extent. Alter mixing, allow the milk to stand one or two minutes, to permit the air bubbles to escape before tak- ing samples. Fill the 20 cubic centimeter pipette by placing the lower end in the milk and Bucking until the milk rises in the tube above the mark on the side. Hat e the finger quickly on the top of the tube and allow the milk to run out slowly until it falls to the mark on the side of the tube; then let the contents of the pipette run into one of the analytical tubes, blowing out the last few drops. Adding the alkali, -Fill one of the 10 cubic centimeter pipettes to the mark on the side with alkali, and allow tlie solution to Bow Into the milk jut measured I 12 the linger ou the top of the tube and shake the tube until the milk and alkali are well mixed. A rubber cot on the finger will protect it from the action of the alkali. Treat all samples in the same way. Place the tubes in the rack, set the rack and tubes iu the water bath, and heat the bath until the water boils; continue boiling for two hours, or until the contents of the tube become homogeneous and of a dark brown color similar to that of sorghum mo Yfter the tubes have boiled for one hour remove the rack and tubes from the bath and examine the tubes to see if the contents are well mixed. If a whitish layer of casein and fat is found floating on the surface of the liquid gently shake and roll the tubes until the contents are well mixed. Re- turn tubes to water bath and boil one hour. The tubes are then ready for the addi- tion of the acid. .hiding the acid. — Remove tie- rack with the tubes from the water and allow them to cool to about 150 degrees Fahrenheit. Then, by means of the pipette, add 10 cubic centimeters of the acid mixture to each tube, slowly, so as not to cause the contents of the tube to froth over. Mix the acid with the contents by running a small glass tube to the bottom of the mixture and blowing gently. Place the rack and tubes again iu the bath and heat to boiling for one hour. Remove the tubes from the water, and then, by means of the water-bottle, rill the tubes with hot water to within 1 inch of the top. The fat will then rise to the top of the water. Replace the tubes in the bath and allow them to staud in the hot water, without boiling, for one hour. At the end of this time remove the tubes from the bath, one at a time, and measure while hot. Measuring the fat. — The analyst will observe that the line ;iSing the upper and lower limits of the column of fat do not extend straight across the tubes, but are slightly curved. In measuring the column of fat place the rule on the tube so that the lower line will come opposite the lowest part of the curved line of the fat; then read up the scale to the division coming opposite the lowest part of the upper curved line The number of divisions on the rule is the length of the column of fat iu milli- meters. The per cent, of fat in the milk is then calculated from the following for- mula and data : Amount of milk taken, 20 cubic centimeters. Specific gravity of milk, 1.032. Specific gravity of insoluble fatty acids, .014 Per cent, of insoluble fatty acids in butter fat, 87. From the above data we have the following formula : 100 aXfrXQ = g d X e Where a = the length of the column of fat in millimeters. 6 = the value of one linear millimeter of measured fat expressed in cubic centimeters. The value of //will vary according to the si/c of the tube used. c = specific gra> ity of the insoluble fatty acids. d 20 <>i grams or the volume of milk taken for analysis multiplied by its specific gravity. 6= per Cent. Offal present in sample of milk taken for analysis. Bubsl ii uting t be figures obtained by an act nal analysis t be formula would bo — 100 30 X.027X.914 = g=418 20.64 x .-■ per iriii. of fat in sample of milk analyzed. \mu:.- The neck of the tubes can be graduated in cubic centimeters, removing the necessity of the troublesome measurements mentioned shove. — H. W. W. ) If the process has been conducted according to the above directions the column of fat will be free from impurities and the line of separation between the water and fat 13 will be perfectly clear. On first rising to the surface the fat is slightly turbid owiug to the presence of a small quantity of water. Although this will make no apprecia- ble difference in the measurement of the fat it may, if desired, be obtained perfectly clear by removing the tubes from the bath and allowing them to cool slowly. The crystallization of the fat causes the finely divided water, which- is distributed through the fat, to collect in drops, which sink to the bottom when the tubes are again heated, leaving the fat perfectly clear. The analyst may fail to obtain correct results from the following causes: Either the column of fat may contain flecks of nndecomposed casein, which would iuciease the volume of fat, thereby giving too high a per cent., or a small quantity of butter fat may remain unsaponified, which will also give too high results. These errors are both caused by insufficient heating of the milk with the alkali, and may bo easily obviated by taking care to heat the mixture of milk and alkali for two hours at least. If not pressed for time it is better to heat two and one-half hours, and thereby re- move all risks of the above errors. If milk containing more than G per cent, is to he tested the mixture of milk and alkali should be heated at least three hours. In such case it would be better, perhaps, to take 10 cubic centimeters in place of the usual amount. Before adding acid the tubes and contents must be allowed to cool to 150° Fahren- heit at least. If added at a higher temperature the contact of the strong acid with the hot alkali solution will generate sufficient heat to cause the contents of the tube to boil with explosive violence, throwing out the contents of the tube and spoiling the analysis. If after the addition of hot water the tubes are allowed to stand in boiling water, small bubbles of gas are given off by the continued action of the acid on the casein. These bubbles rise through the column of fat, rendering it turbid, and causing difficulty in measuring. The bottles containing the solutions of acid and alkali should be kept corked when not in use. If the acid bottle bo left open the acetic acid will evaporate and the acid will not dissolve the casein. The alkali bottle should bo kept closed to prevent absorption of carbonic acid and consequent weakening of the solu- tion. (Note. — The method of Short has been compared in my laboratory with the anhy- drous copper method of Piggott &, Morse. A mean of 12 determinations gave by the anhydrous copper method 3.20 per cent, of fat, and by the Short method o.lS fat. The same samples treated by the Adams method gave results considerably higher. The process of Short has also been improved by graduating the neck of the tube into tenths of a cubic centimeter. For use at stations where expensive chemical apparatus can not be had, the method appears to have great value. — H. W. W.) BICARBONATE OF SODIUM IN MILK. Proust makes an energetic protest against the addition of bicarbonate of sodium to milk for the purpose of preserving it. His principal objection to t be rise of the above salt is based upon its action on lactic acid, the lactate of soda being a salt injurious to children. — (Abstract from Chemisehes Cent ral-Blat t , No. 24, I — . p. B37.) BUTTEB ANALYSIS. ACTION OF ALCOHOL OS BUTTEB PAT. Cochran has made experiments on the solubilities of the different glycerides of but- ter in alcohol with reference to the quantity of volatile aoid which the dissolved and undissolved portions would yield when treated by Eteiohert's method. As :i result «>i the work it Ls seen that that portion of the glycerides dissolved bj ethyl or methyl alcohol is richer in volatile acids than the Undissolved portions. The iodine number <>( the dissolved fat is less than the undissolved fat. The melting point of the fats lived by alcohol is LeSS than that of the undissoh ed !' | 14 The above facts would be of importance in the examination of batters 10 which it is suspected artificial butyrates may have been added. — (Abstract from the Analyst, vol. 13, p. 55.) A QUICK METnOD OF DISTINGUISHING OLEOMARGARINE IN BUTTER. Dubernard has proposed the following method for a qualitative examination of butter for oleomargarine. The method rests upon the observation that pure butter vigor- ously shaken with ammonia at a temperature oi from 70° to 80° and afterwards heated to 100° produces only a .small amount of foam, which does not last a long time. Margarine, on the contrary, produces a large quantity of lasting foam. In order to distinguish butter from margarine by this method, the following manipulation is pur- sued. In a test tube place about 3 grams of the butter to be examined. Heat to 95° or 100° in a water bath, and tben allow to cool to about 80°. Add to the melted substance 5 cubic centimeters of ammonia, shake vigorously, and place again in the water bath and heat to 95° or 100°. The ammonia is volatilized with the formation of foam. The foam in the tube rises more or less according as the butter contains a larger or smaller quantity of oleomargarine. In mixtures of pure butter with known quanti- ties of margarine the tube can be graduated and the relative quantities of butter and margarine thus approximately determined. — (Abstract from Chemiker-Zeitung, 1888, No. 46, p. 760.) DETECTION OF ADULTERATIONS IN BUTTER. Bockairy proposes the following method for the detection of adulterations in but- ter which rest upon the different solubilities of fats in toluene : Place in a test tube 15 cubic centimeters of pure toluene, add 1' cubic centimeters of the filtered fat and 40 cubic centimeters of strong alcohol at 18° j the toluene hold- ing in solution the fatty matter remains at the bottom of the tube. By means of a water bath the tube is heated to 50° and thoroughly shaken. If the sample is of but- ter fat or a mixture of butter fat and some other kind, there is no turbidity ; but if no butter be present a turbidity is at once manifested. The tube is placed in water at a temperature of 40° for half an hour ; pure butter gives no turbidity at the end of that time, but if other fats are present the mixture will appear turbid. — (Abstract taken from the Chemical News, July (!, 1888, p. 11, and Bui. de la Societe Chimique de Paris, vol. 49, No. 5, p. 331.) DIFFERENCE BETWEEN NATURAL AND ARTIFICIAL BUTTER. C. J. van Lookeren proposes the following for a characteristic test between pure and falsified butter: A small amount of butter is melted in a teaspoon, and a drop thereof placed in boil- ing water contained in a watch glass, [f the butter be pure a thin libn of fat is formed upon the hoi water, which breaks up into numerous fat globules, which tend to collect quickly at the periphery. With oleomargarine, etc.. a thin film of fat is formed in the same way, which, however, breaks up into only a few large drops, which remain distributed over the whole surface of the water. The conditions for the success of the e \ peri men t are that the water be perfectly pure and clear, and the melted butter fat very hot. — (Abstract in the KYpert oriuin of the Chemiker-Zei t ung, No. 18, i---. p. I 13, from the Milchat , L888, No, 17, p. 3G2.) a MODIFICATION <>i KOETTSTOEFER'S AXD KEICHERT'S PROCESSES. Low e has proposed the following method of determining the volatility and saponi- fication equivalent of butter and othez fats: About 2 grams of the filtered tat are saponified with 10 cubic centimeters of normal alcoholic potash in a stoppered flask. The alcohol is then boiled off and the soap 15 dissolved in 50 cubic centimeters of hot water. The excess of potash is then deter- mined with semi-normal sulphuric acid solution. The saponification numbers having thus been determined enough additional sulphuric acid is run in to make 25 cubic centimeters in all. The flask is then connected to a condenser and 50 cubic centi- meters distilled off, filtered, and determined in the usual way. The same flask is used throughout the whole process, one of 200 cubic centimeters capacity answering the purpose very well. The whole process does not occupy more than two hours, and can be completed often in less time. Tyrer, criticising the process of Mr. Lowe, was of the opinion that the quantity of fat taken, 2 grams, was too small, since it gave only about one-tenth of a gram of volatile fatty acids. The alkali also dissolved some of the glass, thus introducing an error of which no account was taken. Carbonic acid, moreover, expelled by this method, increased the apparent estimate of filtered fatty acids. — (Abstract from the Journal of the Society of Chemical Industry, March, 1888, p. 185; May, liSS, p. 376.) SAPONIFICATION WITHOUT TUE USE OF ALCOHOL. Mausfeld, in order to avoid the losses due to etherification during saponification in the presence of alcohol, has proposed to carry on the process of saponification, pre- paratory to the estimation of volatile acids in butter, without the use of alcohol. The Diethod employed is as follows: Five grams of the, melted and filtered butter fat are taken ; into the melted fat are allowed to run 2 cubic centimeters of a potash lye containing 100 grams caustic pot- ash in 100 cubic centimeters of water. The flask is closed with a stopper carrying a glass rube drawn out to a capillary point; the flask is then placed in an air bath heated to about 100° and allowed to remain for two hours. At the end of that time the saponification is completed. One hundred cubic centimeters of water are now added, the flask placed in a water bath until the soap is dissolved, which is decom- posed and subjected to distillation in the usual way. — (Abstract in Chemisches Cen- tral-Blatt, June 2:5, 1888, p. 870, from M. Z., No. 17, pp. 281-b3). Wol.I.NVS CRITICISM OF THE KEICIIEKT-MEISSL METHOD. The method of determining volatile fatty acids devised by Beichert has, during the past year, been subjected to an extended criticism by Dr. R. Wollny, of Kiel. Dr. Wollny, as a result of 98 analytical tests, was convinced of the truth of the state- incut of Professor Fresenius, that the R< icherl method was totally unreliable for the determination of very small quantities of butter in oleomargarine. A- a result of his analytical work Dr. Wollny was convinced that the chief source of trior in the Reichert work was due to the absorption of carbonic acid. B< asserts that the error which may arise from this source must amount to as much as 10 per cent, hotter, and renders the results obtained by the method quite inacenrate. It beiti£ very difficult to secure an alkali free from carbonate, the author decided to use a 50 per cent, solution of caustic soda instead of potash, since in a solution of that strength the chloride, nitrate, sulphate, and carbonate of soda are quite insoluble- An arrangement was also devised by which the solution of caustic soda could be kept and drawn off for use without risk of absorption ofcarbonic acid. Such a solu- tion made ami kept in this way gave for :'. cubic centimeters treated by the ordinary process "t Reichert's distillation an amount of volatile acid sufficient to neutralise from .•_> t«» ■'■'' cubic centimeters of the deci-nonnal barium hydrate solution. The author's precautions to prevent absorption of CO arc wholly unnecessary when the saponification i- carried on in closed flasks. Dr. Wollny further discusses the errors due to tin- formation of butyric ethers during saponification and distillation, ami also the error due to the mechanical trans- lation of particles of insoluble fatty acids during the process of distillation, and the 16 error due to the shape and size of the vessel in whirl) (he distillation is carried ou and the time of its duration. The magnitude of these errors he states as follows: (L) Dae to absorption of carbonic aeid +10 per cent. (2 and 3) Formation of bu- tyric ether — 13 per cent. (4) Cohesion of fatty acids —30 per cent. (5) Shape and size of vessel -+- or —5 per cent. — (M. Z., 1887, HTos. 32, 33, 34, 35, and the Analyst, L887, Xos. 139, 140, 141, 142.) Dr. Wollny farther describes the exact methods to be employed in the examination and the results obtained, for the details of which 1 refer to the original papers. As president of tin 1 butter commission of the German Dairy Association, Dr. Wollny has proposed the following method to be employed in the comparative examinations of butter for the forthcoming report of the commission (see M. Z No. •-'•">. 1888). In the preparation of the deci-normal bariotn solution each one of those taking part in the analyses lias been furnished with samples of normal sulphuric acid, pure crystal- lized chloride of barium, and pure peroxalate of potash. In addition to this for test- ing the refractometer samples of olive oil, uitrobeuzol anil nionobromnaphtalin have been sent. Following are the details given by Dr. Wollny for the preparation of tho reagents and the conduct of the analysis. METHOD PROPOSED BY DR. WOLLNY FOR THE EXAMINATION OF BUTTER FOP. THE COM- MISSION OF THE GERMAN DAIRY UNION. [From the Milch Zeitun^. 1888, No. 25 et seq.J (1) PBOTINfl TIIE WEIGHTS AND BURETTES. The weights which are to be used must be carefully compared and the burettes calibrated. (2) ESTIMATION OF TIIE STRENGTH OF NORMAL SULPHURIC ACID. A (lean dry dropping bottle of about GO grains content, the tip of wbich is touched with vaseline on the outside, is carefully weighed : :-.."> or 40 grams of normal sulphuric acid are then placed in it and it is again weighed. Afterwards live [tort ions in dupli- cate of about 'A grams are weighed in a beaker glass of 300 cubic centimeters capacity. The drop bottle being weighed after each portion, the exact weight of the normal sulphuric acid is determined. One hundred cubic centiniett rs of recently boiled dis- tilled water an- added to each of the portions. One hundred cubic cent imeters of this distilled water must not require more than '2 drops of barium solution to give color after 1 cubic centimeter of the phenolphtalein solution has been added. The samples are beated in t lie water bath to the boiling point and to each portion as many cubic centimeters of dilute barium chloride solution added that for each gram of the normal sulphuric acid there are present 10 cubic centimeters of the 15a Cb. The barium chloride solution contains 15 grams of the salt iii l liter of distilled water. Afterwards the beakers are COYered with watch glasses and allowed to stand lor fifteen minutes over the water bath. The precipitates are to be collected upon ash free filters, of 9 centimeters diameter ami washed with hot distilled water until the chlorine action disappears. After drying and incineration the weight obtained is to be multiplied by t he factor .34331. Tin- live duplicate portions, after the addition of 1 oubic centimeter of phenolphtalein solution, are treated with barium solution nut il the red color ap- The numbers obtained are to be calculated to 100 grams of the normal sul- phuric acid. I I ITMATIOH 01 SATURATION OF TIIE NORMAL SULPHURIC kCID II THE BARIUM sr.LlTION. From the results obtained by the above analyses the exacl strength of the deci- normal barium solution is to be determined. (4) TITRATION OF BARIUM SOLUTION Mini POTASSIUM PIROIALATR. One gram of the peroxalate of potassium, after drying for twelve hours in adesiccator ovi i sulphuric acid, is placed in a weighing bottle similar to the one described above, 17 Sixty grams of hot distilled water are then added, and after the salt has been dissolved and the solution cooled the weighing bottle is again weighed. About 12 grams of the solution arejiow run into each of five beaker glasses, and the exact amount in each one determined by reweighing the weighing bottle. Ninety cubic centimeters of re- cently boiled distilled water are now added, together with 1 cubic centimeter of phenol phtalein solution. The barium solution is now added until the red color, which at first disappears, remains for at least five minutes. The results are calculated to the strength of the barium solution according to the formula — lOOOOXj; (6—0 T= 84.5 n (l—t) rhich t = the weight of the weighing bottle empty. & = the weight of the weighing bottle -f- the peroxalate. Z = the weight of the weighing bottle -f- the peroxalate solution. p = the weight of the portion of the solution employed. ?t = the number of cubic centimeters of bariumsolutionused. (5) ESTIMATION OF THE VOLATILE ACIDS AFTER SAPONIFICATION WITH THE AID OF ALCOHOL. Five grams of the butter fat are weighed into an Erlenmeyer flask ; 10 cubic centi- meters of alcohol at 96 per cent, and 2 cubic centimeters of concentrated soda lye at oO per cent., which has been preserved in an atmosphere free of carbonic acid, are added. The flask, furnished with a reflux condenser, is heated, w itli occasional shaking, in a boiling water bath for one-quarter of an hour. The alcohol is then distilled off by al- lowing the flask to remain for three-quarters of an hour in a boiling water bath. One hundred cnbio centimetersof recently boiled distilled water are then added and allowed to remain in the water bath until the soap is dissolved. The soap solution is then im- mediately decomposed with 40 cubic centimeters of dilute sulphuric acid (25 cubic centimeters sulphuric acid to 1 liter), and the flask immediate^ connected with the condenser. This connection is made by means of a 7 millimeters diameter glass tube which, 1 centimeter above the cork, is blown into a bulb 2 centimeters in diameter ; the glass tube is now carried obliquely upwards about (i centimeters and then bent obliquely downward; it is connected with the condenser by a not too short rubber tube. The flask is now warmed by a small flame until the insoluble acids are melted to a clear transparent liquid. The flame is now turned on with such strength that within half an hour exactly 110 cubic centimeters are distilled off. One bundled cubic centimeters of tin- distillate; are now filtered off, placed in a beakei ula>>, 1 oubio centimeter of phenolphtaleio solution added and titrated with barium solu- tion ; when the red color is shown the contents of tie beaker glass are poured back into the measuring ^.lass in which the 100 eubie centimeters was measured, again poured back into the beaker, and again titrated with the barium solution until the red OOloi becomes permanent. The distillation should take place in as nearly thirty minutes as possible. J ESTIMATION OF Til V0LATII.K ACIDS LFTEB SAPONIFICATION WITHOUT ILCOHOL Five grams of the butter t'.n are saponified with 2 cubic centimeters of concentrate d potash Lye preserved from contact with carbonic acid. The Lye solution is made by dissolving 100 grams of the potash in 58 grams of water. Bj gentle rotation of the flask the lye and tat are intimately mixed together; the flask is then placed in a \ ertioal position over a boiling water bath until t lie mixture becomes solid. It is then placed obliquely in the water bath. After it has remained lei'' for two hoars n i^ taken out and the remainder of the process continued as above. 7717— No. 1!) 2 18 (7) ESTIMATION OF THE MEAN MOLECULAR WEIGHT OF THE VOLATILE FAT ACIDS. The soap solution obtained by the titration of barium hydrate in number G is placed in a weighed platinum dish evaporated to dryness in the water bath, dried in a dry- ing oven for two hours and again weighed. (S) ESTIMATION OF VOLATILE FAT ACIDS BY DISTILLATION FROM MAGXESII M SALTS. Eight grams of butter fat are placed in an Erlenmyer flask, treated with 3 cubic centimeters of concentrated potash lye and 15 cubic centimeters of alcohol as in number 5, and the alcohol then distilled off. The soap, dissolved in 100 cubic centi- meters of water, is washed with "2.*>0 cubic centimeters of water in a measuring flask of 500 cubic centimeters capacity and tooled to the temperature of the room. Then by means of a pipette, with constant shaking, 50 cubic centimeters of magnesium sulphate solution, 15 per cent, strength, are added, the tlask tilled up to the mark with water and vigorously shaken. The mixture is now placed upon an 18 cen- timeter filter and 300 cubic centimeters filtered into a measuring tlask. The fil- trate is brought into a tlask holding about 500 cubic centimeters, which on the one side is joined with a steam generator and on the other with a condenser. Two cubic centimeters of strong sulphuric acid together with two pieces of pumice stone are added. Two hundred and fifty to 275 cubic centimeters are now distilled off in to a meas- uring flask of 500 cubic centimeters capacity, the distillate being filtered into the flask through a moistened filter. The gas fla-ne under the distillation flask is then turned down low and the distillation continued in a current of steam until the 500 cubic cen- timeters flask is full. After the filter which has been Lised has been washed -\\ ith a little water the distillate is placed in a large flask of 1 liter capacity and 2 cubic ceu- timeters of phenolphtalein added and titrated. (9) ESTIMATION OF TnE VOLATILE FATTY ACIDS FROM COPPER SALTS. The estimation is carried on exactly as in No. 8 substituting 50 cubic centime- ters of copper sulphate solution for the 50 cubic contimeters of magnesium salts. In the present case the distillate may be collected directly in the one-half-liter tlask without filtration, since by precipitation with copper sulphate solution no insoluble fatty acids pass over during distillation. (10) ESTIMATION OF VOLATILE FATTV ACIDS RY DISTILLATION IX A CURRENT OF STEAM. AC- ('(HIDING TO GOLDMANN. Five grama of butter fat arc weighed in along-necked tlask of about 300 cubic ecu timeters capacity. Ten cubic centimeters alcohol and *J cubic centimeters potash lye are added and the flask connected, on the one hand, with a steam generator, and on the other with a condenser. The contents of the flask are now w armed with a small flame for twenty minutes the condenser being directed obliquely down ward and, in a measuring cylinder, 6 cubic centimeters dial illedoff. The steam is then directed into the flask and 50 cubic centimeters more collected. The Soap is now decomposed with ."> Cubic centimeters of BUlphurio mid fJO ciilnc centimeters strong acid diluted to L00 cubic centimeters) the separated fat acids heated with a small flame until they form a char solution and in a current of Steam the volatile acids distilled off into a •"'()() cubic centimeter flask, The size oi the flame is so regulated that t he contents of the distilla- tion tlask remain constant at about 30 to 40 cubic centimeters, and the distillation is continued until each succeeding 500 cubic centimeters, after the addition of2 cubic centimeters phenolphtalein solution, require not more than 'J cubic cent i meters of I he deci-noi inal barium solution to produce the red color. (13j ESTIMATION OF Till: 1 \TTV \itds SOLUBLE iv 10 PEB n.\T. ALCOHOL (a) Eo$et mi tinxi. —12.5 grams of butterf.it are weighed into a graduated flask of 500 cubic cent i me i.- is capacity! 50 oubiooenl imetera of aloobolic potash lye (about 112 19 grams of potassium hydrate dissolved in absolute alcohol aud made up to 1 liter), are added and gently shaken. After five minutes as slight an excess of semi-normal sul- phuric acid as possible is added. After two minutes dilute with water to 490 cubic centimeters; to destroy the foam, add 5 cubic centimeters more absolute alcohol, fill up to the mark with water and then add as much more water as will correspond Lo the amount of fat taken. The flask is now shaken, contents filtered through a dry filter and 250 cubic centimeters of the filtrate titrated with deei-normal potash lye. (b) Modification of the above process by Dr. Wolhuj. — Weigh G.*25 grams of the fat in an Erlenmeyer flask of 300 cubic centimeters capacity. Add 25 cubic ceutimeters of potash lye, made with absolute alcohol as described above, and then diluted with absolute alcohol until not more than 74 cubic centimeters and not les^ than 73 cubic centimeters of semi-normal sulphuric acid are required to neutralize it. The flask is then closed with a stopper which carries a glass tube 10 millimeters wide and 50 centimeters loug; heat with a very small flame until the contents are near the boiliug point for about twenty minutes and until the smell of the butter ether has en- tirely disappeared. The flask is then cooled in water and 150 cubic centimeters of recently-boiled distilled water added. When the soap is dissolved and the whole mass cooled to about 15°, 75 cubic centimeters semi-normal sulphuric acid are added. The flask is then closed with a rubber stopper and shaken vigorously for one minute. Filter through a strong dry -folded filter into a graduated flask holding 200 cubic cen- timeters, pour the 200 cubic centimeters into a large beaker glass, and titrate with barium solution. (13) ESTIMATION OF INSOLIBLE FAT ACIDS. Treat 3 to 4 grams of the butter fat with 2 cubic centimeters concentrated soda lye and 10 cubic centimeters alcohol in a porcelain dish on the water bath and evapo- rate the soap to dryness. Dissolve the soap with 100 cubic centimeters of hot water, add 5 cubic centimeters concentrated sulphuric acid, and heat the separated fatty acids for one hour on the water bath. Pour the contents of the dish on a filter of 11 centimeters diameter, made of tin; best thick Swedish filter paper previously dried and weighed. Upon this filter the acids are washed with 1.5 liters of boiling water. The filter with the insoluble fatty acids, alter they have solidified, is placed in a weighed beaker glass, dried for two hours in an air bath, and weighed. (16) ESTIMATION OF THE FREE FAT ACIDS, Ten grams of the butter fat are placed in a beaker glass with 20 cubic centime! ether and 10 cubic centimeters of alcohol and 1 cubic centimeter phenolphtalein so- lution. The free acids are then titrated with dcci-normal alcoholic potash. (11, 14, 15, 17, IS, 1!» COMBINED ESTIMATION OF THE SAPONIFICATION NUMBER, THE VOLA- TILE. SOLUBLE, AND INSOLUBLE FAT ACIDS. (a) Saponification number. — Five grams of the fat are weighed into an Erlenmeyer flask and 25 cubic centimeters of alcoholic potash added. (Twenty-five cubic centi- meters alcoholic potash .should saturate from Id to ">(> cubic centimeters semi-normal sulphuric acid.) The mixture .should he wanned as in No. 5 for a full hour with the reflux condenser. Meanwhile the same quantity of alcoholic potash should be ex- actly titrated with the semi-normal sulphurio acid, and later being titrated by the barium solution used and the number obtained noted. The alcoholic snap .solution is taken from the waterhath, the solid parts adhering to the walls of the flask dis- solved by gentle shaking, and, alter the addition ot 1 oubic centimeter phenolphta- lein solution, titrated with semi-normal sulphuric acid. The number obtained is subtracted from the number obtained in the blank experiment ami calculated hack to the corresponding number of cubic centimeter's barium solution. (b) Volatile fai aoida.- The alcohol is driven oil' from the exactly neutralized alco- holic soap solution obtained in tli kg experiment, and then an additional 20 amount of sulphuric acid added, amounting to 2 cubic centimeters more than neces- sary to neutralize the original potash lye employed. The mixture is then dilated with 140 cubic centimeters of recently-boiled distilled water and two pieces ofpamice stone added, and, as in No. 5, 110 cubic centimeters distilled off, of which 100 cubic centimeters, after filtration, is titrated. (c) Soluble fat acids. — After the end of the foregoing experiment the distillation llask is removed from the distillation tube and replaced by one filled with distilled wa^er and placed as a receptacle under the condenser. The condensation water is now allowed to flow from the condenser and the condensed insoluble fatty acids driven over into the original distillation flask by a current of steam. Iu like man- ner the insoluble acids collected upon the filter are washed into the original distilla- tion flask. After the fatty acids are solidified the whole of them are brought into a cylinder 3.5 centimeters wide and 15 centimeters long, which ends in a 6-millimeter wide tube carrying a goose-necked glass tube from its side; and on the lower end is closed with a rubber tube and pinch cock. The lower contraction of the cylinder is Btopped with a plug of glass wool which forms a thick filter but allows hot water to mil through easily. Under the goose neck of this wash vessel, which is held iu a ver- tical position by a retort holder, is placed a graduated llask of 500 cubic centimeters capacity into which the excess of water flows. When the acids have been brought from, the distillation llask, together with the pieces of pumice stone into the wash llask, any residue is washed out with the washing bottle with boiling water. The above must be so conducted that no drops of the fatty acid pass through the glass wool and that no fatty acid remains on the upper walls of the washing llask. The washing llask is furnished with a gum stopper cai r\ ing two holes. Through the one passes a tube for the inlet of steam ; through the other a tube at least 8 milli- meters in diameter, cut off obliquely at the lower end and divided above the stopper into two brain lies. By means of one of these brain lies it is connected with a perpen- dicular condenser, at least 8 millimeters wide, and through the other it is connected by means of a rubber tube and pinch cock with a hot-water reservoir by means of a U-tubo reaching to the bottom thereof. Through the steam tube, which is some- what bent at the end, so that the steam is directed obliquely downwards and out- wards against the sides of the washing bottle, a current of steam is conducted which stirs up in a lively mannerthe surface of the fatty acids, which swim upon the topof the water, while the condensed water drops back out of the reflux condenser in a boiling-hot condition and the excess flows through the goose neck into the measur- ing llask. The wash water collected in the 500 cubic centimeter llask is titrated with barium solution, with the addition of 2 cubic centimeters phenolphtalein solu- tion. This operation is repeated three times; in all, 1.5 liters of wash water being collected. The total amount of barium solution required in the three operations is collected in one number. (d) Insoluble fatty acids. — After the end of the foregoing experiment the wash flaek is cooled until the fatty acids are solidified. The pinoh COCK closing the lower open- ing of the, wash llask is opened, the water allowed to How off, and a currenl of air d through the apparatus by means of an aspirator after t he o [ten ing of (he 0O86 neck is lightly stoppered and the steam generator removed. The w ash flask together with the fatty acids and the glass wool plug are completely dried in this manner after, almost, half an hour. The steam generator is now replaced with a flask which con- tains ether free of water, alcohol, and acids. The ether is brought to the boiling point by means of warm water and it s v a por condensing in the washing llask dissolves the fatty acids and carries them into the weighed Mask of aboul loo cubic centimeters capacity placed below to receive them. When about 40 or 50 cubio centimeters of ether bave distilled over, the operation is broken, the outer end of the wash flask washed with ether and the weighed llask below is placed in a warm place in order to drive oh' the ether. The weighed flask containing the acids is now warmed in the water batfa ami placed under the receiver of an air pump. After the a< ids are completely 21 freed from ether by this method the flask is again weighed and the weight of the acids noted. (e) Mean molecular weight of the insoluble fat acids. — The fatty acids obtained in the foregoing experiment are melted at a low temperature and from .8 to 1 gram weighed in a beaker glass, dissolved in 50 cubic centimeters of alcohol, 1 cubic centimeter phenolphtalein solution added and titrated with barium solution. (20) DETERMINATION OF THE IODINE NUMBER. Weigh out from .8 to 1 gram butter fat in a flask of about 200 cubic centimeters ca- pacity furnished with a glass stopper; dissolve in 10 cubic centimeters chloroform, and add 20 cubic centimeters iodine solution. In case the liquid is not clear, more chloro- form must he added. It' the iodine color should rapidly disappear add from 5 to 10 cubic centimeters more. The amount of iodine added should be sufficient to secure strong colorization after standing two hours. Add 10 to 15 cubic centimeters iodide of potassium mixture and 150 cubic centimeters of water; titrate with thiosulphate of sodium solution until the iodine has almost disappeared. Add then a little starch solution, and continue titration until the blue color has disappeared. (21) ESTIMATION OF BEFBACTIYE INDEX. Set the large Abbe refractometer with water at 18° so that the index reads 1.33:30. The instrument should then be placed in a room which is kept at a temperature of about 25°. The estimation of the refractive index of butter fats can not be made below 25° on account of the solidification of the fat. The instrument is also to be tested at 20° with olive oil, nitrobenzol and monobromnaphtalin. (25) DETERMINATION OF THE SPECIFIC GBAYITY AT ZERO OK 15°. A platinum crucible is supplied with a platinum wire handle so that it can be sus- pended from the book of the balance. It is then weighed empty and afterwards its specific gra\ ity taken in water at zero. The water is kept at zero by being placed in a small beaker glass, which in turn is put in a larger beaker closed and surrounded with finely pounded ice. The crucible is then dried and 15 grams of melted butter fat placed therein. The butter tat is allowed to solidify slowly at a temperature not below 15°. The crucible with its contents is then weighed in the air and placed in distilled ice-cold water for an hour. The whole is then weighed in water at zero as before. From these results the specific gravity of the fat at zero is determined. In the same way determine the specific gravity at 15°. (20) ESTIMATION OF THE SPECIFIC GRAVITY IN BOILING VTATEB BT MEANS OF THE l'\K- NOMETEB. A small flask of ."•' , grams, the neck of which is drawn out to a narrow tube and marked at tin; narrow place, is weighed empty. It is then Idled with distilled water to the mark and placed for an hour in finely pounded ice. The contraction of the water is restored by the addition of ice-cold water until the mark is reached ; it is then dried, allowed to conic to the tempt rat me of the room, and weighed. The same flask is then tilled to the mark with boiling water, ami, after coining to the tempera- ture of the room, weighed. The flask is then emptied, dried, tilled with melted butter fat, kept in a boiling- water bath for an hour, the butter fat tilling to the mark, after- wards cooled to t be temperature of the room, and weighed. (J7) ESTIMATION OF THE SPECIFIC 0BAYIT1 UlNI Monks BALANCE, This is det ei n lined in t be usual manner by the aseof a specific gra\ Ltj bob, the ther- mometer of which is arranged so as to show a temperature of 100 . 22 ESTIMATION OF THE SPECIFIC fiSATITY WITH THE AREOMETER IN ROIUXf. WATER. This is determined by means of spindles furnished by the firm of G. C. Gerhard, in Bonn. These spindles are graduated especially for specific gravity at a temperature of L00°. (St) ESTIMATION OP THE MELTING POINT. This is.determined in straight or U-formed capillary tubes. They are tilled with bntter fat at 100° and placed in ice- water for a quarter of an hour. The capillary tube is then placed, with a thermometer, in a beaker glass with water at 20 and slowly wanned with a small flame. (SO) ESTIMATION OP THE SOLIDIFYING POINT. Place about 100 cubic centimeters of the melted fat in a test tube at 40 . Stir with a thermometer until the tat begins to solidity, the test tube meanwhile being placed in a Lai I containing water" at 20 . The temperature will rise slightly when the solidification begins, and the highest point reached is entered as the solidification point. Note.— Dr. Wollny has proposed in the above scheme an exhaustive study of but- ter fat. Many points appear to be superfluous and others elaborated into unnecessary detail. The method of determining insoluble tatty acids appears to be more objection- able than any other part of the scheme. The method of making the melting point first described in the Journal of Analytical Chemistry (vol. 1, No. 1, p. 39) appears i<» have escaped the notice of tin- German commission. 1 also note with regret the ap- parent ignorance of th ■ existence of the Gooch crucible existing in Germany. Points to be especially commended in these investigations are, sending samples ami standard re-agents, blanks for entering the analytical results, and minute printed directions for conducting the manipulation. The numbers prefixed to tin; several paragraphs correspond to the number of the column in the blank in which the results are to be entered. — H. W. \Y. CRITICISM OF WOLLNY'S METHODS. Goldmann criticises Wollny's original method chiefly because it gives only a part of the volatile acid present and takes no account of the whole. It has never been claimed, however, for the Reicherl process t hat it gave anything more than a portion of the volatile acid present, but the percentage of the total volatile acids obtained is approximately constant, and therefore the results are comparable among themselves. Goldmann proposes to remo\ e the Lasi t races of alcohol after saponification by dis- tillation in a current of steam. Before the distillation of any pari of the alcohol, however, the solution is heated with a reflux condenser for half an hour BO that any ethers which have been formed may be I'eeoinbined l>\ the exoess of alkali present. Tin- apparatus for the r< moral of the last traces of alcohol is illustrated in No. 1 1 oi' the Chemiker-Zeitung for L888, page '2ir>. Goldmann's paper is so long that for all details I can only refer to the original. The method finally adopted by him is a Hollows: ."> grams of the melted battel- fat are weighed in a flask of 300 cubic centimeters capacity, t he neck of which is L2 centi- meters long and 'i centimeters wide. Add b» cubic centimeters alcohol, 96 per cent., ami 'i cut lie centimeters of a 50 per oent. aqueous soda lye made and preserved out of COnta< t, with carbonic arid. The flask, furnished with a reflui condenser, is heated lor twenty minutes with a small flame. The condenser is turned to an angle of .">."> and 7 Pennsylvania avenue. Thompson's dairy, 511 Four-and-a-half street, south wesl Floral Hill dairy K. K. Ward's wagon. . 1. 0288 1.0334 1.0341 1. 0207 1.0820 1.0292 1.0811 1.0-270 1.0343 13 54 11.88 11.68 12.45 11.53 11.78 10.71 11.40 12 -47 «.t. 87 ii.:; 1 .) 11.68 8.93 7. .',7 8. 14 7.43 X. OS S. Ul 8 15 4. 61 8.71 8 80 :; 87 3.39 ::. 06 :;. 28 8 88 2. r>4 :t. oh :;. 51 & 36 2. 38 2. 59 :i. 95 ::. 41 :: lit 2. 74 :;. 17 a u 2.31 The milks were Collected between March S3 and 30. Of the eleven samples, lnit three were a1 all near the standard which would allow their sale in Massachusetts, New fork, or New Jersey, and of these three hut two were normal In composition. Of the entire eleven only sis were normal, according i»> the tables, these showing b very close agreement with theory, while the other five were so extremely discordant 25 as not to permit of errors of analysis. The normal milks agreed remarkably well, a3 may be seen from the following figures: Fat found by analysis. Fat calcu- lated fornor- mal milk. Difference. 3 97 3. 95 +.02 3.39 3.44 -.05 :;. 06 3.10 -.04 3. 38 3.47 -.09 3.88 3.95 -.07 2.54 2.49 +.05 These figures certainly show the tables of Hehner & Richmond to be equally appli- cable to American and continental milks. The differences in the remainder of the milks were as follows : Fat found 1 I ; at 1 " alcn -, byanaljs.s. malmilk Difference. 4.01 3.71 3. 80 3. 25 3.08 5. 30 2.38 2. 59 2.74 2.31 - .75 +1.33 + 1.21 + .54 + -77 In all but one case an excess of fat was found, which leads to the belief that oleo oil or similar substance had been churned into the milk to enrich a poor or skim milk, the difference being, it seems to nie, altogether too large and striking, after the coincidences found in the other samples, to make it readily supposed that the sam- ples were normal in character. The extremely poor character of the milks of this city point to the necessity of milk inspection, even if this is due to low-grade cattle and poor feed alone; and it would seem that the tables given by Dr. Wiley may prove of great value in pointing out suspicious milks for further examination. Dr. Wiley spoke of Bebner & Richmond's recommending the rejec- tion of all analyses not agreeing with their tables as erroneous, and said that this was certainly very dogmatic, as on adulterated milks such agreement is impossible. Mr. Richards spoke of Swedish machines for taking out cream and putting back cottonseed oil, illustrated in London Engineering, Professor Meyers spoke of the methods of emulsifying oil and skim- milk and its frequency. Cotton-seed oil can be switched in for call feed, but he did not think artificial cream can be made so. Oleo oil is probably used. The secretary then read the following commnnieat inn from Professor (i. B. Patrick, of [owa : VOTJ fi OS THE METHOD FOE SOLUBLE A \ D INSOLUBLE F I /' ACIDS, In the determination of soluble and insoluble fit acids I have recently found it con- venient, as well aa economical of time and material, t«> change Blightlj the modi- of procedure as laid down in the method (originally fcfuter's) adopted by this associa- tion at its last meeting (1887) and published in Bulletin No. it', from the Department of agriculture, Chemical Division, and 71. 26 The purpose of the modification is to avoid the use of a separate saponifying bot- tle, and the consequeut transfer of the soap solution to an Erlenmeyer flask, and evaporation of the alcohol added in this transfer. This end is accomplished by saponifying directly in the Erlenmeyer, the latter (with stopper tied down) being inclosed in a " conversion "jar with tight cover, and containing absolute alcohol (50 cubic centimeters), the vapor of which gives an out- side pressure on the Erlenmeyer equal to that within. The jar is made with vertical sides, and is only a little larger than the flash it is to inclose : its cover is held down by a screw-clamp in the usual way, a disc or washer of rubber making the union tight. Breakage by contact of the flask with the walls of the jar while agitating is pre- vented by a ring of corks strung upon a wire. Thus arranged, rrith a tight jar, there is no danger of breaking the Erlenmeyer if the work is executed with ordinary care. Fifty cubic centimeters of absolute alcohol iu the jar will suffice for many analyses. Two styles of jar have been tried, one of heavy glass, iron bound (made by Eimer Amend), the other of heavy sheet copper, made (excepting the clamp) by an ordi- nary tinsmith; the cover of each is a thick ground-glass plate, with a metal disk to receive the pressure of the screw. The glass one, allowing a view of the interior dur- ing saponification, is the more convenient. The saponification is conducted in a steam or air oven at a few degrees below 100 centimeters. Neither style of jar can be heated directly on the water or steam bath with safety to the contents. This needs further trial. Another slight deviation from the method of last year, which I find convenient, is to retain the cake of insoluble acids in the Erlenmeyer for weighing, using alcohol merely to dissolve into the flask whatever fat acids arc on the dried filter. This small amount of alcohol is quickly removed, first on the water bath and then in the oven, with a current of dry air or hydrogen. By these two changes in the method the entire 1 determination of insoluble acids is effected without any transfer of material from first to last. At the start the filtered, still liquid, and well-shaken fat is weighed into the tared Erlennieyer: at the end, after the insoluble acids are weighed, they are removed and the Mask is dried and tared again to insure against error from " nicking" or corrosion during the op- eration. However, in the few trials thus far made, the first and last weights of the flask have been practically identical. The report of the committee was then adopted. Dr. W. J. Gascovne Mien read the report of the Committee on Plios- phoric Acid, the president stating that, at the request of the Executive Committee, Dr. Qascoyne had retained the position of chairman on his resignation as state chemist of Virginia REPORT OF THE COMMITTEE ON PHOSPHORIC ACID. lis required by the constitution of the association, the Committee on Phosphoric Acid herewith presents its report. The report presents: l, brief notices of new ana. lytieal methods for determining phosphoric acid proposed during the year; \\ results of work done by t he committee and members of the association; 3, recommendations for t he oexl j ear. ( '. Molir fC in-iii. Zeit., 1 1, 117 11- : A i>s. J. chein. Soc., L887, p. 884) proposes the fol- lowing volumetric method for phosphoric acid: T WO grams of the phosphatic miIi- stance La treated with aacceseive small quantities of S per cent, sulphuric acid in a mortar. The residue is added to the extracts, and the whole made up to HM cubic centimeters and allowed to digesl for one hour: lo oubic oenti meters of the filtered solution, corresponding to 0.2 grams, are thee treated wiih potassium ferrooyanide so long as iron is precipitated. After the addition of sodium acetate, the phosphoric 27 acid is titrated iu the usual way with a standard solution of uranium acetate. It is stated that the end reaction is not influenced either by an excess of potassium ferro- cyauide or by the presence of Prussian blue. G. Kennepohl (Chem. Zeit., 11, 1089-1091; Abs. J. Chem. Soc, 1883, 321; J. Soc. Chem. Ind., C, 680) confirms the opinion expressed by Klein (Chem. Zeit., 10,721; J. Chem. Soc, 1886, 835) as to the practical non-occurrence of iron phosphide in nor- mal Thomas slag. The phosphoric acid may be accurately determined in the follow- ing manner: Ten grams of the finely powdered slag are introduced into a 500 cubic centimeter flask, moistened with alcohol to prevent adhesion, and heated in a water bath for at least half an hour with 40 cubic centimeters of HC1. (1.12 sp. gr.) and 40 cubic centimeters of water. After cooling the flask is filled to the graduation mark and the solution filtered. An aliquot part is mixed with ammonium nitrate and molybdic solution without previous removal of the silica. The solution, after belt- ing at about 80° for fifteen minutes, is filtered, and the precipitate washed with water containing 3 per cent, of nitric acid, redissolved in 24- per cent, ammonia, and then precipitated with magnesia mixture. The presence of silica does not interfere, owing to the ready solubility of ammonium silico-molybdate in the washing water. If the presence of ferrous salts should tend to cause the separation of molybdic- oxide, which dissolves but slowly in ammonia, an addition of nitric acid or bromine, before adding the molybdic solution, will insure the production of a precipitate in stantaueously soluble in the ammonia. Isbert and Stutzer (Zeit. Anal. Chem., 26, 583-587; Abs. J. Chem. Soc, 1888, 191: Chem. News, 57, 211; J. Soc. Chem. Ind., 7, 44; J. Anal. Chem., 2, 193). The method based on the determination of the ammonia in the phospho-molybdate precipitate (Chem, Zeit., 11, 223) is confirmed. A further simplification consists in washing the yellow precipitate with cold water, instead of with ammonium nitrate solution. The ammonium silico-molybdate is soluble in pure cold water, although insoluble in ammonium nitrate solution. On the other hand, the phospho-molybdate requires 10,000 parts of cold water for its solution. The analysis is conducted as follows : Five grams of the phosphate are dissolved iu hydrochloric acid or aqua-regia, diluted t<» 500 cubic centimeters and filtered. Fifty centimeters of the filtrate are mixed with an excess of ammonia, acidulated with nitric acid, and the phosphoric acid precipitated with ammonium niolybdate. The precipitate is allowed to settle at tit' t > 70 ( '. for fifteen minutes, and then filtered, the supernatent liquid being fust pass* f standard acid, which is then titrated back with baryta water, using rosa lie- acid as the indicator. One part of nitrogen in the precipitate corresponds with 1.654 parts oi' phosphoric acid. Test analyses with known quantities of phosphate, with and without silicic acid, ami comparative tots of phosphatic manures bj the above and gravimetric process, slow that for commercial purposes the method is sufficiently accurate. eh. Malot (Monit.Scicnt.. lf?87, 487; J. Pharoi., 16, 157-159; .1. chem. Boo., 1887, lor,:!; J. Soc. Chem. Ind., li, 563) describes a new volumetric method for 1\(). by ura- nium nit rate. Oxide of uranium gives a greec lake with cochineal, which property is ntilized for the determination of phosphoric acid. The phosphate i- treated according to Jonlie's method, {.«., dissolved ic EC1, the phosphoric acid precipitated with oitro-magnesinm mixture and the precipitate dis- solved in dilute nitric acid. A few drops of tincture of cochineal are added, then ammonia until the violel coloration just appears, and this in it^ turn is made to din- appeal with one to two drops of nitric acid, The eolation is now heated to inn c . 28 5 cubic centimeters of sodium acetate solution added and mixture titrated with uranium nitrate. Each drop of the latter causes a greenish-bine zone, which, on agi- tation, disappears again. As soon as precipitation is complete, the solution assumes a lasting greenish-blue color, remaining unchanged by excess of the uranium solution. The end reaction is most distinct. I Jy employing a very dilute solution of uranium nitrate, the determination is rendered very exact. The author uses solutions. 1 cubic centimeter of which represents 0.002 gram of phosphoric acid. A. Ernmerling (Zeits. Anal. Chem., 26, 244, Abs. ('hem. News. 57, 15; J. Anal. Chem., 2, 223) describes a new method for the determination of soluble phosphoric acid in superphosphates. The solution of superphosphates, mixed with calcium chloride, is allowed to flow into a standardized solution of caustic soda, to which some pheuolphtalein has been added. The following solutions are required: (1) Sodium hydrate, of which 1 cubic centimeter represents about .005 P a 0> (2) Calcium chloride solution made by dissolving 200 grams C&Clg in one liter of water. (3) Pheuolphtalein, 1 gram in 500 cubic centimeters of alcohol. (4) Methyl orange in aqueous solution. The execution of an analysis is carried on as follows: Two hundred cubic centimeters of the solution of the sample prepared in the ordi- nary manner are well mixed with 50 cubic centimeters solution of calcium chloride. One burette is filled with this mixed solution and a second with the soda solution. Of this latter 20, 10, or 5 cubic centimeters (according to the strength of the super- phosphate) are measured into a beaker ; 2 cubic centimeters of the pheuolphtalein solution are added, and some water, and the superphosphate solution is then run rather rapidly. As soon as the color begins to grow faint the superphosphate liquor is added more gradually, and at last drop by drop, until the redness has entirely dis- appeared. The author next measures off again the same number of cubic centimeters of the mixed solution of superphosphate and calcium chloride as used above, dilutes with a little water and mixes with 4 to 6 drops of the methyl orange solution. The liquid is titrated cautiously with the soda solution, finally drop by drop until the reddish color changes to a yellow or orange yellow. The smaller number of cubic centimeters of soda solution, as consumed in the titra- tion with methyl-orange, is deducted from the number obtained on titrating with phenolphtalein. If we have dissolved 2<) grains superphosphate in 1,000 cubic cen- timeters of water, and make up 200 fco250 cubic centimeters by the addition of 50 cubic Centimeters of the calcium chloride solution, and if a is tin 1 measured quant it \ of soda, fcthe solution of superphosphate and calcium chloride consumed, the soda used for neutralizing the free acid, and / the sf andard of the soda expressed as 1'. < ) . t he percentage of soluble phosphoric acid in the sample is equal to ,250x1000x100 cX/X 200X0. "JO This method is applicable t<> ferruginous superphosphates. J. Ruffle (J. Soe. Chem. Ind., 6, 321 333; Abs. J. Chem., Boc., L888, p. 87; J. Anal Chem., l, p. 455) has contributed an interesting paper on moisture and free acids in superphosphate. The loss of moisture, dried to constant weight, at various temper- atures varied in one Sample from L2.92 per cent, at 38 to 50 ('., to 17.93 per cent, at L50 c. Drying at the same temperature, 100 C, for different periods, showed a vari- ation of from '.».'. 17 per cent, for thirty minutes, to 17.15 percent, for seven hours. The fertilizer dried in its nat oral state, at inn C. t showed a much greater Loss of moisture than when previously rubbed tOS line paste in a mortal. It i> also shown that the soluble phosphoric acid existing in superphosphates is not entirely pre* at •< • m >no-calcium phosphate, ami that t lie exposure of i"" ( '. drives 29 off more than the true moisture, that is, the adhering nncombined water. It is recom- mended to determine the moisture in the following manner: Weigh out 2 to 5 grams of the phosphate in its natural state on a double watch glass, place under an air pump over dry calcium chloride, exhaust, then leave for eighteen to twenty-four hours and weigh. John Clark (J. Soc. Chem. Ind., 7, 311-312) describes a modification of Perrot's (Compt. Eend, 03, 495) for the estimation of phosphoric acid as phosphate of silver. The objection to Perrot's process, when applied to manures and natural phosphates, are — (1) The chlorides which may be present will be estimated ns phosphates. (2) The method of separating the phosphates of iron and alumina from the phos- phates of lime and maguesia is inaccurate, as it is practically impossible to dissolve out the whole of the prospbate of lime in the manner indicated. The modifications which the author recommends are — (1) The solution of the phosphate of silver precipitate in nitric acid, and the titra- tion of the silver with sulpho-cyanide. (2) The neutralization of the acid solution with caustic soda instead of ammonia, to avoid the presence of an excessive quantity of ammoniacal salt, which affects the results. (3) The previous precipitation of the iron and alumina as phosphate with acetate of soda containing free acetic acid. The following is an outline of the process: The phosphate is dissolved either in water, nitric acid, or sulphuric acid, the greater portion of the free acid neutralized with caustic soda, and to the cold solution acetate of soda containing free acetic acid is added in excess. If the addition of the acetate of soda produces a precipitate, this must be filtered off, re-dissolved, and re-precipitated with acetate of soda, as before. The filtrate and washings are added to the previous filtrate, then excess of nitrate of silver, which will give an immediate precipitate of phosphate of silver, Ag 3 P0 4 , and the free acetic acid is neutralized witb caustic soda till there is only a faint acid re- action to litmus paper. The slightest excess of caustic soda will cause a brown precipitate of oxide of silver, but this oxide of silver dissolves easily on the addition of a few drops of dilute acetic acid. The precipitate of phosphate of silver, which will contain any chloride which may have been present, is thrown on a filter, thoroughly washed with water, t hen dissolved off the filter with hot dilute nitric acid, mixed with a little ferric sulphate, and the silver titrated with sulpho-cyanide, as described by Volhard. and calculated to phos- phoric acid. The precipitate of phosphate of iron and alumina is dried, ignited, and weighed, then dissolved in acid, the iron determined voluinetrically. calculated to phosphate of iron, and the balance assumed to be phosphate of alumina ; or the oxide of iron may parated with can-tie soda and the alumina in the filtrate weighed as phosphate of alumina, using the precautions recommended by Thomson (J. Soc. Chem. Ind., vol. 5). Comparative tests on a variety of materials show a close agreement with the mo- h bdic method. Dircks and Werenskiold (I.andw. Versuchs. Stat., 1887, 125-453; Ai>s. J. Chem 1 — , 628) have tested the various processes employed for the est i mat ion and separation of tri-caleium from mono- and di-calcium phosphate, namely, the various modifica- tions of the ammonium citrate method. Tiny find t hat although Done of the i net hods give a really satisfactory and exact result, Petermann's process is perhaps the most trust worthy. Sohindler (ZeitsAnal. ('hem. .-jr. l ij-i 16 ; J. chem. 8oc, 1888,753, J. Boo. chem. Ind., ?, 455) describes ■ volumetric method for the determination of phosphoric acid 30 based upon the format ion of a compound of phospho-molybdate of aminonia of definite composition. The following solutions are requisite : (1.) Mohjbdic solution. — To a liter of molyldic acid solution prepared in the usual manner, 30 cubic centimeters of a solution of citric acid containing 500 grams to the liter are added. (2) Concentrated ammonium nitrate solution containing 750 grains to the liter. (3) Dilute ammonium nitrate solution containing 100 grams to the liter and 10 cubic centimeters nitric acid. (4) Magnesia mixture prepared in the usual manner. (5) Lead solution, 1 cubic centimeter of which corresponds to .04 grams P a O . This is obtained by dissolving 55 grams lead acetate in one liter of water and a lit- tle acetic acid. (6) Molybdate of ammonium solution, 1 cubic centimeter of which corresponds to 1 cubic centimeter lead solution : 25 grams ammonium molybdate are dissolved in one liter and standardized against the lead solution. (7) Tannin solution, 1 gram tannin is dissolved in "20 to 30 cubic centimeters of water. The analysis is conducted as follows: 50 cubic centimeters of the nitric acid solu- tion of the phosphate (0.5 gram of substance) is mixed with so much of the concen- trated solution of ammonium nitrate that after the addition of the molybdic acid solution the mixture shall contain 25 grams ammonium nitrate per 100 cubic cen- timeters. Then for each 0.1 gram ofP-O.-,, 100 cubic centimeters of the molybdic acid solution is added, and the mixture heated in a water bath to about 58 C. The precipitate is allowed to settle for ten minutes, and the liquid decanted and passed through a tilter. The precipitate is washed with 50 cubic centimeters dilute ammonium nitrate solution by decantation, and dissolved in 3 per cent, ammonia solution. The solution, together with that dissolved oflf the filter, is brought into a quarter-liter flask, 10 to 20 cubic centimeters magnesia mixture added: after shaking, it is made up to 250 cubic centimeters and filtered. Fifty cubic centimeters of the filtrate are taken, acidified with acetic acid, and the liquid made up with boil- ing water to 300 cubic centimeters. Lead solution is now added from a burette until a small excess of lead goes into solution, and is titrated back with ammonium molyb- date solution, using the tannin solution as an indicator. A drop of tannin solution gives a red coloration with ammonium molybdate, which is visible in a solution of 1 in 400,000, whereas lead molybdate gives no coloration: and lead acetate only a greenish coloration. Comparative determinations OU a variety :>f materials show a close agreement with the magnesia method. At a meeting of the directors of the Belgian experiment stations, the following method for the estimation of the phosphoric acid soluble in water and in citrate of ammonia in manures was adopted : Two grams of ordinary superphosphate (or one gram of a rich sample, say over 90 ,1. » are triturated iu a mortar with water ami thrown upon a tilter. The filter is washed with water until about 200 cub-e oenti meters have passed through; bo the filtrate is added l oubio centimeter of nitric acid, and the volume of tin- solution is made ui> to 250 cubic centimeters. The tiller, and t he insoluble mat t«r contained in it. is now introduced into a 250 cubic centimeters flask with 50 oubio centimeters of l'e I erina nn's alkaline citrate of ammonia so hi I ion. and allowed to digesl for One hour on the water bath at a temperature of 3€ to 48 C. After rapidly cooling the fiask is tilled up to the mark with \\ ater. Of the above filtered liquid (the citrate of ammonia solution), take 50 cubic centimeters and mi\ it with 50 cubic centimeters of t in- aqueous solution, acidify with nitric acid and precipitate with molybdic acid solution. The remainder of the opera t ions remain as for the ordinary estimation of phosphoric acid. I '•■ rma tin's ;i I kali ue citrate of a Minion i;i is prepared by dlSSOh tag 500 -rams of cit- 31 ric acid crystals in water and mixing with TOO cubic centimeters of ammonia of 0.920 specific gravity. Bring the concentration of the liquid to 1.09 specific gravity at 15 c C, and add to each liter 50 cubic centimeters ammonia of .920 specific gravity. RESULTS OF WORK DOXE DURING THE TEAR. Last October your committee sent to nineteen official and seventeen commercial chemists live samples for phosphoric acid determination, with the request that the determinations be made within the week ending October 29. The samples were marked as follows: No. 1. Ground South Carolina phosphate. No. 2. Ground tankage. No. 3. Ammoniated superphosphate. No. 4. Dissolved South Carolina phosphate. No. 5. Dissolved Navassa phosphate The samples were carefully mixed, so as to secure uniformity. Returns have been received from twenty-seven laboratories, covering the work of thirty-two chemists. The results are given in the following tables : No. 1 and Xo. 2. Analyst. E. H. Farrington, Connecticut experiment stal ion T. B. Osborne, Connecticut experiment station A. L Winton, Connecticut experiment .station JT. P. Carson, Richmond, Va ('. Glaeer, Baltimore, Md \V. J. Gasooyne, Baltimore, Md Lehmann & Mager, Baltimore, Mil <\ C. Read, New Bedford, Mass. is. Terne, Philadelphia, Pa Basin Fertilize r Company, Baltimore, Md P. R Wilson, Baltimore, Md R. EL Gaines, Richmond, Va .1. M. liait lett, experiment station, Orono, Me s. II. Merrill, experiment station, Orono, Me W. Frear, Stat.- College, Pa Clifford Richardson, Washington, D.C St ill well & Gladding, New York C. B. Back, Wilmington, Del W. Robertson, Charleston, S C M. A. Scovcll. Kentucky experiment .station A. M. Peter, Kentucky experiment station 11. }',. Battle, North Carolina experiment station A. V Shireick, Wood's Boll, Mas-, N. W. Lord, Columbus, Ohio W. MoMurtrie, Champaign, 111 A. B. Knorr, Depart n on I of Agriculture, Washington. - . .!. M. MoCandlesa, Atlanta. Ga H. Proehling, Richmond, \'a P. R ChaaaC Columbia, S. C Bradley Fertilizer Company (Benson), Boston Bradley Fertilizer Company (Patrick), Boston No. 1 — South Caro- lina phosphate. No. 2— Tankage. Phos- Phos- Moisture. phorio Moisture, plioric acid. acid. .92 1.37 .83 1 02 1.00 1.04 97 .76 1. OS 27.67 27. «9 28.14 28.30 28.78 27 80 27 88 28. 20 28. 22 28. 30 28.00 27. 87 27.84 27.67 2a 83 27. 98 28.40 28.22 28. 34 27. 87 28 16 6.78 6.66 6.03 6.94 7. 14 7. 14 7. 64 7."66" <;. 55 6.16 6. 26 7.08 C. 4 t 6.70 6.70 14.39 14.30 14.33 13.95 14.01 14.20 14.52 14.91 14.24 14. 02 14.06 14. 71 13.89 14.00 it. to 14.87 14.59 14.50 14.71 14.41 14.15 it 32 14.08 14. 13 14.23 14.16 18.62 14 75 14.91 32 ►.— Ammoniated superphosphate. Analyst. K. II. Farrington, Connecticut experiment station.. T. B. Osborne, Connecticut experiment station A. L. Winton, Connecticut experiment station J. P. Carson, Richmond, Va ('. Glaser, Baltimore, lid W.J. G altimore, Md. Lehmann & Mager, Baltimore, Md nl. New Bedford, .Mass B. Teme, Philadelphia, Pa Rasin Fertilizer Company, Baltimore, Md J'. B. Wilson, Baltimore, Md K. II. Gaines, Richmond, Va. J. M. Bartlett, experiment station. Orono Me S. II. Merrill, experiment station. Orono, Me W. Frear, State College, Pa Clifford Richardson, Washington, D. C St ill well & Cladding, New York ('. i:. Buck, Wilmington, Del W. Robertson, Charleston, S. (' M. A. Sec veil. Kentucky experiment station A. M. Peter, Kentucky experiment station II. B. Battle, North Carolina experiment station .. A. !•'. Shireick, Wood's Holl, Mass N. W. Lord. Columbus, Ohio W. McMortrie, Champaign, 111 BLnorr, Department of Agriculture, Washington J M. MoCandless, Atlanta, Ga W. W. Cooke, Burlington, Vt II. Froebling, Richmond, Va P. E. Chazal, Columbia, S. C Bradley Fertilizer Company ( Benson), Boston Bradley Fertilizer Company (Patrick*, Boston.. 1-1.59 14. 82 17. 21 10. GO 17. If) 14.0.-, 12.00 14.81 16.40 16.38 12.87 12. 87 18.36 18.10 14.30 15. 70 13.91 13.88 14.3!) 16.50 19. 65 14.06 12. 95 15. 95 — - = *-* 30 7.02 7. L2 7.11 8. 75 7.08 7. Hi 7. 16 7.oo 7.19 7. 10 7.01 6.00 7.49 7.15 7.00 7. 29 6.97 7.09 7. 02 7. (10 7.02 7.07 6, 74 7.14 6.99 7 19 ;. 48 7.04 1. 52 . 05 1. 25 1.39 1.41 1.78 1.35 1.03 1. 55 1. 05 1.02 1. 14 1. 34 1.58 2. 27 .80 1.22 1.36 1.49 1.17 1. 27 . 93 1.27 1.42 1.7S 1.59 1. 20 . 93 1.26 1.26 .99 1.22 8.54 8. 86 8.16 - B4 8. : 1 9.60 8.71 8.05 - 8.33 8.44 8. 50 - 28 8.37 8.14 - 7. 95 8. 27 .v »! 8.85 8. 34 7. 02 8.74 1.40 1.84 1.36 1.64 1. 92 1.39 1.79 1.89 1.72 2. 20 2. 17 1.40 1.38 1.60 1.59 1.73 1.44 1.58 1.78 1.56 1.78 1.52 1.40 1.45 1.30 1.35 1. 82 1. 52 1. 56 ' - 2.00 No. 4. — Dissolved South Carolina phosphate. as T. 5 o . o • ^3 ©.2 ~rs .= — -~Z Analyst. g _ C5 s ■ /. z -: > — ,= "x'Z •a fk > ~ a """ H o £ 0U - >\el, Kentucky experiment station 2. 73 1. 07 i i ; A. M. Peter, Kentucky experiment station 2. 82 i I i 1. . > 11. B. Battle, North Carolina experiment station .- 10.94 •J. 2 1 13. 18 2. 10 1 5. 37 A 1- Shireick Wood's Roll Id 10.78 11. if. 10.68 8.01 1.81 8.81 it. 18 L3.74 18.94 1. M 1.72 2.81 N. W. Lord, Columbus Ohio L5 io w. Mr Mm tn.- i ihamp'aign 111 l i 1- A. i: Knorr, Departm 1 Agriculture, Washington 15 L'2 [( Candless, \t antn Ga 10.50 3.12 15.00 W. \v. Cooke, Bnrlingtou s'i 10.70 .' 1 15.63 ling, Rii hmoncl Vn 11.18 1.68 15.61 P. F. CbazaT, Col i n - « LI. 82 1'. Ml Bradley Fei i ilizei < pai ton ... 10 -0 13. lo 2.31 1 5. 4 1 1 ompanj | Pati ick I, Boston 11.00 16.80 33 No. 5. — Dissolved Navassa phosphate. 05 O . i 2_: 6 P.S -Z — - ~Z Analyst. -~ - 'Z ~ s s« 2 a ■p tt « IJ > — -- 1 a ~~'~ 1 M < q ~" H z c-i B. 11. Farrington, Connecticut experiment station... 9. 29 8.00 7.38 3.18 18.56 T B. Osborne, Connecticut experiment station 9.29 a oo 7.21 15.21 3. 22 1-.43 7 93 3. 02 IS. 37 J. P. Carson, Richmond, Va G.78 8.78 15.56 3.38 18.95 C. G-laser, Baltimore, Md 10. :;:» - _ 15.18 3. 4 1 18.59 9.15 7. 78 ]5. 10 3. 33 18. 19 8.72 6 78 8 89 15 07 3 52 19. 19 C.C. Read, New Bedford, Mass 8.09 7.90 7.27 15.17 3.20 B. Terne, Philadelphia, Pa 9. g:j G. 91 8.44 15. 35 3.77 19.12 10.40 6.05 8 02 14. G7 3. KG 18. 5 1 P B Wilson Baltimore Md 10.35 6. CO 8.62 14. 62 18.43 7.52 7.67 15. 19 3.52 18. 71 .). M. BarUett, expeiiraent station, Orono. Mc 8. 65 6.93 14.98 3.G1 s II Merrill, experiment station, Orono, Mo 7.68 7.36 15 nt 3.42 18.46 "W Frear State College, Pa 11.91 G. 92 7.82 14. 74 3.70 18. 14 15 80 8 03 G 66 14. G9 3. 39 8 92 G 93 8. 52 1"). 45 3. 2] 18.66 C. E. Duck, Wilmington, Del 6.87 8.16 15.03 3.26 18.29 W. Robertson, Charleston, & C 8. 55 7.55 8.01 15. 5G 3.31 18.87 M A. Scovell, Kentucky experiment station 9.09 8.06 6. 72 14.78 18.30 9.27 8.08 6. 52 14. GO 18.25 JI. is. I ', it tic, North Carolina experiment station 9. 7:, 7.46 7. 38 14.84 3.34 ia is 9. 10 7 33 7.89 15.22 2. 93 1 8 1 5 N \V Lord Columbus, Ohio 10. 83 7 4? 7. 61 15.08 3. 13 \V. MM in trie, Champaign, HI 10. 04 4.48 14. 52 3.58 18.10 A. K. Kuorr, Department of Agriculture, Washing- 8.25 9.50 6.78 7.04 5. 65 14.72 15.53 3.34 2. 95 18. 06 J. M. McCandless, Atlanta, Ga 18.48 W. W.Cooke, Burlington, Vt : 32 15. 16 18.84 7.39 8. 25 7.G7 7.G5 15.1 6 15.90 3.1U 18.36 izal, Colombia, S. (J 8. 95 Bi adley Fertilizer Company (Benson), Boston 7.11 7.95 15. 00 :; : \ 18.60 7.95 7.05 15. 00 3.70 18.70 Averages. Sample No. 1 : 1 1 1 1 e I'lc sphorie acid Bample No 2 : Moisture Phosphoi lo acid Bample No. 8: Moist are Soluble phosphoric acid i led phosphoric acid a mailable phosphoric acid Insoluble phosphoric acid Total phosphoric acid Sample No 1. Moisture Soluble phosphoric acid Kevei ted phosphoric acid A raUable phosphoi Ic acid Insoluble phosphoric acid Total phosphoric n Id. Sample ~ Moisture Soluble phosphoi ic acid I;. \ ei ted phosphoi i«- acid 1 ible phosphoi Ic acid — [nsolnble phosphoric acid Total phosphoi Ic acid Number Of determi- Average. Bighest ' Difference. nations. 17 . 95 1.37 .70 .61 31 28.07 28. 78 27. 47 1.81 17 G 7:) 7. Gt l.Gl 31 11.31 14.91 13. G2 1.29 24 15. 10 19.65 12 LO 7.65 82 7. 11 G. M> 2. 57 ::_' 1.31 •_'. 27 1.47 82 9. 60 7. 92 32 1.65 2. 20 1.80 32 1H.07 11. 19 1. 66 24 9.61 11. 14 2 1 1 82 10 88 11.51 83 2. 93 5.21 1.21 82 13 BI 14.20 12 37 82 1.96 2.81 32 18 76 L6.20 is. ia 84 B 08 7.71 82 7.:. j 10. (1 1 82 82 15. ll 3. 40 1 0.-19 1.13 7717— No. 1!»- 34 As further illustrating the differences in the results the following table is pre- sented : Percentage of results differing from the average not more than— .10 .25 .50 .75 1.00 L26 1.50 , -- 2.00 and over. Sample No. 1: 19 129 5!) 23 Gl 52 94 47 91 90 100 70 4 94 97 97 04 97 87 94 M 91 78 9G 29 47 53 8s 100 94 100 100 Sample No. 'J : Phosphoric acid Sample No. 3: 88 17 ill 97 97 1(10 97 9:2 97 ill 97 97 100 Gl 63 97 100 29 94 100 97 46 94 58 97 66 100 100 Soluble phosphoric acid 59 41 38 22 28 37 31 1G 22 13 9 12 12 16 34 28 34 78 6.1 83 GO 59 54 : : 44 37 5!) 53 17 28 34 50 GO 59 100 Total 97 95 97 91 91 100 97 96 97 94 97 97 96 97 94 100 100 100 97 97 Sample No. 4: 100 100 Total Sample No. 5: G7 84 7.'. 100 83 87 88 90 94 91 93 94 100 100 Reverted phosphoric acid 100 Insoluble phosphoric acid Total 100 The committee recommends the continuance of the presenl general method for the determination of phosphoric acid, subject to such modifications as the association may determine. Respectfully submitted, W. J. Gascoym:. Chairman. The report being open for discussion, the president called attention to the dependence of amount of soluble phosphoric acid on the size of filters used during its extraction, and thought it would be well for the committee to recommend a definite size of ( .> centimeters. Professor Frear spoke of differences due to variations in moisture in different laboratories. Mr. Gascoyne said the variations in moisture was not proportional to the variations in phosphoric acid, and that such differences as i- and 19 per cent, could not, be due to anything but carelessness, the highest determinations of moisture being accoiupanie 1 by highest per- centages Of phosphoric acid. Professor Lupton, OB examining the results, said he could see no way to harmonize variations. The president thought the publication Of these results would be most beneficial to the association, and that results anot her year would be sent in \\ Ufa greater care. The president then introduced the subject of the use of pumps for ex- tracting soluble phosphoric acid, considering it an advantage where much work was to be done. 35 Mr. Gascoyne spoke of quick work with molybdate and magnesia solu- tions, allowing them to stand but five to ten minutes for the yellow pre- cipitate and not more than fifteen for magnesia precipitate. Mr. Gaines's and Mr. Bichardson's experiences confirmed his observa- tions. On motion of Dr. Wiley, further consideration was postponed till after- noon, and then by unanimous consent he offered certain amendments to the constitution, reported later to the convention, and asked for the ap- pointment of a committee of tbree under the constitution to consider them. The president appointed Dr. Wiley, Professor Stubbs, and Professor Voorhees. On Professor Meyer's motion the convention then adjourned till 2 p. m. AFTERNOON SESSION, THURSDAY. The convention was called to order by the president at 2 p. m. The discussion of the report of the committee on phosphoric acid be- ing in order, it was directed that the method for the coming year pro- vide for the use of Schleicher and Schiill No. 589 filters, 9 centimeters iu diameter, in the determination of soluble phosphoric acid, and that the preliminary washing be by decantation from a beaker, using small quantities of water, as described by Messrs. Gascoyne and Voorhees, who considered the use of the pestle only necessary with the very high grade phosphates in New England. After some argument the use of a pump in determining soluble phos- phoric acid was left to the judgment of the analyst. Dr. Gascoyne then called attention to the method of Isbert & Stutzer for the estimation of phosphoric acid, and gave the following deter- minations, comparing it with the official method. The method was care- fully followed, except that half normal sulphuric acid and caustic soda were used, and cochineal as the indicator. Acid phosphate Booth Carolina phosphate ... Tankage Steamed li.nu> Ammoniated raporphosphate Aciii phosphate total Arid phosphate, Bolahle A< i. TS Molybdic precipitate ten minutes magnesia precipitate two hours 15. BO Molybdic precipitate one hour, magnesia precipitate fifteen minutes 15.80 Molybdic precipitate one hour, magnesia precipitate thirty minutes. 15 82 Molybdic precipitate live minutes, magnesia precipitate lii'tccn to twenty minutes 15.79 Acid phosphate, total Acid phosphate, soluble .. Acid phosphate, insoluble Ground bono Tankage. .' Molybdic pre- cipitate, live Official minutes, mag* method. nesia precipi- tate, 15 to 20 mm.; 1G. 10 16.07 10.22 10. 24 2, L2 ■_'. OS 22. 1 1 22.17 9.93 9.90 The report was then adopted and the committee directed to insert the modifications in the method. Dr. Wiley then introduced, by permission, Prof. F. W. Clarke, of the U. S. Geological Survey, who addressed the association on the de- sirability of forming a national chemical society, and suggested the ap- pointment of a committee to represent the association in any steps that might be taken, and this being agreed to, the president appointed Dr. II. W. Wiley, of Washington; Prof. W. O. Stubbs, of Louisiana; and Prof. E. A. von Schweinitz, of North Carolina. Dr. Wiley then exhibited one of Abbe's large refractometers and read Die following paper: EEFB L< FIVE iM>rx or BUTTER FAT. By ii. w. win v. Heretofore \ ei 5 Little attention hat been given t<> the refractive index of batter fat in studying the properties of thai Bubstanoe. II.t\ ing had oooasion this year to ex- amine the refraotn «• index of various Eats and <>iK used in the adulteration thereof, I extended the examination to twelve samples of batter fat from bul ter bonghl in the open market. The object of the investigation was twofold : First, to determine the mean refractive index; and, second, tin- rate of variation of the refraotive index for rise 01 foil "i" temperature, it is, of oonrse, obvious thai the refraotive index of but- 37 ter can not be read after it solidifies, and therefore trie minimum temperature at which the refractive index can be determined is not much below 30°, although the butter fat may be kept some time even at 25° without solidification. Avery warm room where the temperature was reasonably constant was selected in the determination of the refractive index at the lower temperature. For the higher temperature the hot room of a Turkish-bath establishment was used. The temperatures at which the reading was mace and the observed index of refraction, the rate of variation for each degree tor each sample, and the mean rate of variation for each degree of the twelve samples a c given in the following table: Vuniber. To. Index. Rate for each degree. Number. To. Index. Rate for each degree. 1 C 32 \ 54. 6 ( 31.9 I 50. 2 C 31.8 I 55.0 S :s(>. 8 \ 54. 8 ] 32. 3 1.4 5:; 5 i 1. 4497 5 1.4535 I 1. 4495 5 1. 4530 I 1.44855 1.4530 ) 1. 4495 i 1.4530 } . 0001G9 . 000165 .000194 .000146 1 . 000153 . 000203 C 32.2 \ 56 6 C 32.4 \ 54.8 5 32. 2 \ 55.0 ( 31.8 I 5(5. 2 S 31. 6 j C 31.8 \ 56. 2 1. 4540 ) 1. 4495 5 1.4540? 1. 4505 j 1. 4536 > 1. 4500 5 1. 4530 ? 1. 4490 5 1.4540 I 1.4493 5 1. 4485 5 . 000186 2 8... .000157 3 9 000158 4 10 . 000164 5 11 . 000186 6 I 55.8 1.44H9) < 33. 2 ' 1. 12 000171 1 55.4 1. 4febtt > The reading of the instrument with distilled water at 18° = 1.3305. Having deter- mined the rate of variation for butter fats, in order to reduce the reading to any standard temperature it is simply necessary to use the mean factor and add or sub- tract the result of the multiplication as the case may be. For instance, take the case of i he first sample. The refractive index at 32° is 1.4535; suppose it is wished to re- dace this to a temperature of 25° ; the difference in temperature between 32° and ' . the mean rate of variation for each degree is .000171, for 7° it would, bo .001197. Since the index of refraction would be higher at 25° than at 30° we add this number to the number obtained above. Then 1.4535 + .001197= 1.1547 — index of refraction of that sample at 25°. The index of refraction of butter fat is distinctly low.-i- than that of COtton-seed oil, lard oil, olive oil, and linseed oil. The rale of variation I of temperature is also less for that of butter fat than the sub- stances mentioned. On the conclusion of Dr. Wiley's paper a practical demonstration of the working of the lactocrite was given by Mr. A. B. Knorr, after which the convention adjourned until Friday at 10 o'clock. FRIDAY MORNING. The convention was called t<> order at 10 o'clock by the president Dr. FarringtOD called attention to the fact that water was as >nita- ble for washing the yellow molybdate precipitate as ammonic nitrate. [sberl and Stutzer bad observed this, and Dr. Gascoyne said that his experience confirmed it. On motion, the committee <>a phosphoric acid was directed to insert this as an alternative in the methods. 38 Dr. Crampton, in the absence of the chairman, Professor Eising, pre- sented the report of the committee on fermented liquors, as contained in the following letter: Office of State Analyst, Berkeley, July 18, 1888, Dear Sin : I hope you will not think that I have intentionally been -wanting in respect to the other members of the committee on " fermented liquors " because I have till now failed to communicate with you. Early in the year we began some original work upon artificial colors, etc., and it was my expectation to have had some results that I could have submitted to you for your approval before the close of the year, but our work was interrupted by other analyses, and at the last moment the re- sults were not sufficiently satisfactory to offer to the public. However, I believe we are on the right track, and that we may yet be able to recognize many of these artificial colors in a very short time. I had until a few weeks ago fully expected to have been present at the August meeting of the association. I had planned to have spent a few weeks in your laboratory before the annual meeting, and then and there we could have made up our report. This was my plan, but at the last moment I find myself unable to carry it out, and now am unable to make auy suggestion. If the associa- t ion would grant us another year I would promise to be on hand, and I think I could also promise to spend some time in your laboratory, when wo could perfect our re- port. However, if you can make a report I would not in the least hamper you, but wish you to feel at liberty to do so. I have written Professor Wiley and Mr. Clifford Richardson my views, which are intended as an explanation to the association for the failure on my part to do my duty in the matter. We have proved beyond all question, as I think, the presence of boracic acid in many unadulterated California wines. In addition to the test with tumeric paper we have obtained the flame-test so decided that there can bo no doubt. I have just received a new spectroscope, and I shall hope by its aid to settle this point beyond all question. Our method of work is as follows : 50 cubie centimeters to 100 eubic centi- meters of wine are evaporated in a platinum dish, ignited, and burned to ash. Part of this ash is transferred to a platinum spoon; such an one as is used for blow-pipe work answers very well. A few drops of the strongest H 2 S0 4 (I have used a 96-98 per cent, acid), then alcohol is added and then lighted and immediately blown out, and relighted and again extinguished. The first flash will show the acid very dis- tinctly if it is present. The adulterations that we have had to look for have been easily detected. The wine in all cases was taken direct from the wine-maker. It was taken upon its sale to the wine-dealer, who had required a certificate of purity before accepting it. Nat- urally, only pure wine or wine that was believed to be pure would l^ offered. The only kind of adulteration believed to have been practiced l»y wine-makers is, first, the use of ant iseptlCS, salicylic acid, sulphurous acid, and boracic acid, coloring matters, alkalis (K : CO t ) to neutralize exoess of acid, etc. The watering of wine (st retching) belongs to t he dealer's art, etc. We have not had occasion to investigate t his branch of the subject . We examine for plaster, bases added, antisept ics, artificial colors, copper, lead, zinc salt-, alcohol, free acid (volatile and non-volat ile), solid residue, ash (solu- ble and insoluble), alkalinity of the soluble portions, sugar, glucose, glycerine, etc., tannic acid, cream of tart a r, etc. Some samples of east en i wine were clearly "manu- factured." I have I >een obliged to go into the country OB account of mj family and am now in the midst of the Napa \ alley wine reglOO and am making the acquaintance of the wine men here, and hope to gain their confidence so that they will consult me in re- gard to their troubles, etO. There Is B BtrOUg .sentiment among the wine-makers against the use of any agent thai could be called an adulterant, ami I find them \'i\ anxious to maintain the reputation they already have for purity. Thewine- dealei la the man we have to fear; he is read j todoalmosi anything aud to get ahead 39 of the chemist if lie can. The Viticultural Commission has sent a special agent to Paris. They have instructed him to go about and get as many of the practices of wine makers and dealers as possible. I find great difficulty in getting a set of sam- ples of wine-coloriug substances. They are only sold in original packages, etc., under various names. I shall hope through our special agent to get a pretty complete set, and then if I can identify them and give them their true scientific names, we will have a surer base to stand upon. I regret chat I can not attend the meeting this year. If the convention will con- tinue the committee I think we can get to work very early the coming year, and pre- sent to the next convention a valuable contribution to the subject. May I ask you to present my apology to Professor Fellows. I do not know his address, nor can I get it till I return to Berkeley. Very truly, yours, W. B. Rising. Dr. C. A. Crampton, Agricultural Department, Washington, D. C. Dr. Crampton added that he thought it advisable to adopt as pro- visional methods those which he had collated from foreign sources and published in Part III of Bulletin No. 13 of the Division of Chemistry of the United States Department of Agriculture. Mr. Richards suggested that the committee examine into the subject of alcohol tables and recommend some standard that might be brought into common use, as the English, French, and American are now all different, and since the recent investigation of Squibb all shown to be more or less incorrect. The suggestions of the committee were then adopted. Prof. Myers, on behalf of the committee on potash, then presented their report as follows: REPORT OF THE COMMITTEE OX POTASH. The committee on the determination of potash would respectfully submit: That since t be List meel ing of this associal ion little has appeared in chemical literature re- lating to the determination of potash, so far as we know, which calls for notice here. The method of Undo as modified by Gladding appears to give universal satisfaction. The New Jersey Agricultural Experiment Station (Report for 1887, p. 175) has deter- mined the potash in forty-eighl mixed fertilizers by this method, with certain slight modifications, and also by the Btohmann method. The greatest difference in the re- sults in any ease was .'27 per cent. ; the average difference .08 per cent. In two-thirds of the determinations the differences were not over .10 per cent., and in only one case was the difference over .15 percent. Stohmann's method gave on the average slightly higher results than the other. The modified Lindo-Gladding method used at HieNe* Jersey station is as follows: '• The solutions were made as described in the [official method] hut previous to the addition of ammonium hydrate 1 cubic centimeter of ammonium oxalate was added; the sol nt inn was tlim idled to the mark, thoroughly mixed and filtered ; an aliquot part was transferred to a porcelain dish of approximately 60 cubic centimeters capac- ity. From this dish the material was not removed until tie double salt of potassio- platinic chloride was reads t<> be brought upon the Gtoooh Qlter. After the determi- nation was removed from the bath the precipitate was moistened with water, and then washed with 80 per cent, alcohol till all trace-, of platinic chloride wer moved. It was theu washed with the saturated ammonium ohloride solution reoom< mended, it was ion ml in this laboratory that better results could be secured by adding 40 tb is solution to the precipitate in the porcelain dish, and washing by decautation until all salts other than the potassio-platiuic chloride were removed. Since a rela- tively small amount only of the potassio-platiuic chloride is necessary to saturate the ammonium chloride solution, these washings were all rejected, the time saved and the greater accuracy secured compensating for the loss of larger amounts of the solu- tion. After this washing has been completed, the double salt was transferred to the Gooch filter, washed with pure alcohol, dried and weighed. The salt was then dissolved in hot watt r, and the final weight of the potassio-platiuic chloride secured by difference." SAMPLES FOR ANALYSIS. Following the custom adopted by the association, very carefully prepared samples were sent out by the committee with the request that they be analyzed and the re- sults reported to the committee. Twenty-one samples were sent, but reports were received from only nine laboratories or analysts, the most of the others informing the committee that, owing to reorganization and changes taking place due to the in- iluence of the ''Hatch bill,'' they could not find the time to complete the work. The samples sent out by the committee were numbered from one to five inclusive, and consisted of the following compounds: No. 1, chloride of potassium; No. 2, sul- phate of potassium; No. 3, an annuoniated superphosphate mixed with No. '2 in such proportions as to give theoretically 13.50 per cent, of K 2 O; No. 1 was calcined kainite; No. •">, an ammonia* ed superphosphate containing theoretically 5.2 per cent. 0fK*O. Nos. 1 and 2 proved not to be C. P., though bought for C. P. salts. Table giving the results of the determination of potash in committers sample. No. l. No. 2. No. 3. No. 4. No. 5. E. II. Jenkins, analyst, Connecticut Experiment Station. !17. 86 •17.80 2 17. 89 2 17.M Lost. 3 53. ::() 53. 78 53.84 13.51 13.51 63. Hi 53. 23 53.81 13.51 17.847 William Freer, analyst, Pennsj Ivania Experiment Station. 4 • 5t. 11 10.38 Dr. Caldwell's laboratory , Cornell University. 53. 70 13 15 1 17. 7S 5.21 5.18 53.047 :>:\. (15 L3.185 17.77 5. 195 Iii. Battle, analyst, North Carolina Bxporlmenl : Ion. 14.49 it. i • 19.78 8.00 5. 82 58. M 63. 80 14.805 'JO. 025 5.01 No. 1. No. 2. No. 3. No. 4. Dr. Lnpton, analyst, Alabama Experimi n1 Station. 54.27 54. 20 64.235 51.53 61.87 13.30 13.47 10.08 19.67 19.675 4.90 I 61 4. 755 Dr. Crampton, analyst, Department 01 Agricoll urc, Washington. L.O A.O. 5;{. GO 54. it 51 08 A.O.A.C. 64.02 4.0 4.0 13.90 4.0.4.0. 17,20 17. 20 17.20 I 7. 22 I ' L7.32 L7.82 13.78 17.32 4.0 4.0 i 98 L99 £~0T 5. 36 :. 21 .i..\. M.. n analj • J 4. and hi. College, L3.I I "17.30 17.21 17.27 1 and Differ* nl oh< 0. P. Contains n.i. big. s o. 4 Avei . ills. • Not C. P. ■• I > i i 1 1 • i .hi weighings. 7 Nol sat i ifaotorj to analyst, i>ut not time to repeat. 41 Table giving the results of Hie determination of potash in committee's sample — Continued. No.l. No. 2. No. 3. No. 4. No. 5. Dr. B. V. Herff, analyst, Mississippi A. and M. Col- lege. No. 1. No. 2. No. 3. No. 4. No. 5. Dr. E. Y. Jlerff. analyst, Mississippi A. and M. Col- lege— Cont in ned. A.O.A.a A.O.A.a. A.O.A.a 13.74 13.70 13. 72 A.O.A.a 17.01 17.74 17. 02 17.92 17.76 17.58 17,50 17.42 A.O.A.a 5.36 5.26 5.20 5.12 L. G. L. G. L. G. L. G. 17.32 17.28 L. G. 5.22 53. 386J 53. 433 13. 49 1 17. 352 5.32 15. \V. -fcLilgore, analyst, Mississippi A. and M. Col- lege. 13.72 17. 647 5.234 l. a. 63. 56 53. 4* 53.32 L. 0. L. G. 13 50 13.48 L. G. 17.40 17.40 17. 36 L.Q. 5.16 5.34 5.56 53.70 13.87 13.91 '17.80 17.08 5 96 53. 44 5 83 53.30 13.89 17.44 5 895 No. 1. No. 2. No. 3. No. 4. No. 5. 53. 404 53. 524 13. 605 18.078 5 295 54.27 53.10 54.11 51.53 14.49 13.15 20.26 17. 08 6 00 4 61 Lll 2.58 1.34 3.18 1 39 23 17 24 22 Not satisfactory to analyst, but not time to repeat. Table giving moistures of committee's samples of potash. Name of analyst. No. 1. No. 2. No. 3. No. 4. No. 5. Dr. Caldwell Per cent. 0.065 0.38 0.27 0.3 Per cent. 0.00 0.11 0.21 0.15 Per cent. 4.40 0. 29 4.19 4.3 Per cent. 1. 25 1.67 1.07 1.30 5 SO Dr. ETrear, Pennsj 1 ■ ania Experiment Station 7.01 7 50 ' (To. 5 broken while in Bhipment. Not b. — < Mhc •! analysts railed to report the moistun . i Taken at time of bottling samples, In giving tin' results of the analyses the usual custom of giving tin" averages of all of the results is followed, though the committee does not claim thai it represents fairly the work of the chemists of the country. It is hoped, however, tli.u the Btudy of the results of these analyses may lead to more careful work by the various analysts in determining thisverj difficult element. It is true thatamong the analyses may lie found some results which arise from inexperience, but m perusal ^i' the table shows tli;it some of the very besl known analysts in the country differ i>y more than 0.5 per cent. npoD the analysis of the same Bample, or as Dearly the same sample .-is it is pos- sible to have mixed and shipped. The most serious difficulty appears t>> exist with thus,, samples containing sulphate of potassium, i. «., Nos. 2 and I. In the ana of these compounds we bave sum,- reallj extraordinary results, ■ variation »>f more than 2.5 per cent, occurring i»«t ween the extn In the c as e s of at least t wo of these the attention of the analysts was called to the variation, and the work was at once reviewed, It could not be explained upon the 42 theory of variation of nioisturo in the samples, as the determinations of moisture re- ported by the analysts came almost within the ll working error. " In reviewing these, different chemists in the same laboratories were placed upon them, and the results arrived at, in the judgment of the chief analysts, justified the report made at first by the chief. Every precaution having been taken to have the samples the same, Ave are driven to the conclusion that there is some minor variation in the methods prac- ticed in the respective laboratories, or else the scheme now in use is inadequate to overcome the difficulties of the case. In the cases of the other samples the variations from the mean are not so striking, but are still too considerable to be satisfactory. An analyst finding tho minimum per • cut. and one finding tho maximum per cent., reporting upon any one of these samples would differ by more than 1 per cent., which in a large business transaction based upon such analyses would be quite a serious matter. The results in the judgment of your committee are highly unsatisfactory. Tho sources of error in the determination of potash seem not yet to bo fully understood by our analytical chemists, and demand the most serious study. This association has been working upon this problem for several years, and appears to have made little <>r no progress towards removing the difficulties. The various analysts in any one laboratory appear to agree reasonably well with one another in most cases, but their results frequently do not bear comparison with those obtained elsewhere. In some cases, perhaps, care has not been observed to ap- ply proper corrections for impure PtCb or other sources of error which might be cor- rected by cheeking the work of the laboratory by blank determinations with pare KC1. Your committee would recommend the continuation of the present schemes for do- termination of potash, with the modification proposed by the Now Jersey Experiment Station. Eespectfully submitted. John A. Myers, Chairman. William Frkak. The report being open to discussion, Dr. Gascoyne called attention to the recent experiments of Lindo, contained in Chemical News, 50, 103, approving the association's method. Professor Scovell said lie had some difficulties, and Mr. Richardson spoke of the presence at times in the case of superphosphates of sul- phate <>r lime in the final double salt. Mr. Voorhees then described the modification practiced al the New Jersey Agricultural Experiment Station, as given in the eighth annual report, p. L75, and said that they found the Lindo-Gladding method most desirable when a little oxalate of ammonia was used to remove lime, and the evaporation was not carried too low, and referred to the large series Of results which he had published in the report referred to. [f the evaporation were carried too Ear, solids mighl form which would be insoluble in alcohol, in which case a small quantity of water should first be added, and then the Btrong alcohol. The greatest error in inexperienced bands lies in this point. Mr. Chazal said he had also found error In not evaporatiog low enough. Mr. VooiImts called attention to the tact that t he presence of am- monia in the air of the laboratory produced no effect with the Lindo 43 method, as the ammonia double salt would afterward be dissolved out by the wash, and then moved the addition to the association method of the use of oxalate with the ammonia and evaporation to dryness with HOI, taking up with water, repeated evaporation, and taking ivp with alcohol. This was adopted. Professor Scovell then presented the report of the committee on nitro- gen, as follows : REPORT OF THE COMMITTEE OX XITROGEX. Your committee on nitrogen have the honor to present the following report of the work which has heen accomplished during the year : ABSTRACT OF ARTICLES WHICH HAVE APPEARED DURING THE YEAR OX THE DE- TERMINATION OF NITROGEN IN FERTILIZERS. [Compiled, by request, by Dr. E. H. Jenkins, Connecticut Agricultural Experiment Station, and A. M. Peter, Kentucky Agricultural Experiment Station.] The Kjehlahl method. Lenz (Fres. Zeitschr. Analyt. Chem., 26, p. 590) has tested the effects of omit ting oxidation with potassium permanganate after boiling the substance to be analyzed •with sulphuric acid. He used a flask of 100 cubic centimeters' capacity, .2-7 gram of substance, 10 cubic centimeters of oil of vitriol, and .5 gram metallic mercury. Tho boiling was continued until the substance was clear and colorless. In one scries of determinations no permanganate -was added; in another it was added as recom- mended by Kjeldahl. In all cases higher results were obtained when permanganate wus used ; generally the differences were small, but in some cases amounted to .33 per cent. Ulsch (Zeitschr. fur das gesammte Brauwesen, 18S7, p. 3 ; Fres. Zeitschr., 'J? calls attention to the tact that if platinum chloride is added to hasten oxidation when the substance is boiled with oil of vitriol, a3 recommended by him in a previous pa- per, an excess must be avoided. Too much platinum chloride retards instead of hastening oxidation, and if tho boiling is continued too long, platinum may destroy ammonia by "catalytic action," and so cause loss of nitrogen. lie also recommends the use of iron sulphate to destroy mercuro-ammonium compounds formed when mer- cury is used to hasten oxidal ion. It has the advantage over potassium sulphide that it can be added to the acid before neutralizing it, thus avoiding danger ot* losing nitrogen. Dafert (Landwirtscliaft. Vei 3 1, 311 ; als i l'n b. Zeitschr, fur Analyt. (/hem., Ji has studied the chemical reactions involved in the method, and the question how generally applicable the original method of Kjehlahl is, and whether it can be so modified as to make it available for determining nit rogen in all classes <>i com- pounds. His conclusions are briefly as follows: (1) Sulphuric acid withdraws from t he nitrogenous organic matter the elements of wat.r and of ammonia, and from them forms ammonia. (•J) The sulphurous acid which is evolved regularly reduces the nitrogenous com- pounds; hut this effi ^nitieant compared w it h that mentioned ah.. (3) The addition of Organic compound . ic ) to nitrogenous matter-, delays the formation of ammonia, exoept in oases where il changes a uiti inbstanoe, which is volatile or easily decomposed, into one which is less easily attacked bj sul- phuric acid. 44 (4) Potassium permanganate, when brought into the hot mixture, destroys tho or- ganic compounds still remaining. A part or all of the nitrogen in such compounds forms ammonia. In quantitative work, when the previous digestion with acid h;is been carried far enough, and the permanganate is added carefully, all the nitrogen is thus converted into ammonia. (5) His study of the effect of adding metalsor metallic oxides leads to the conclusion that the addition of a metal to a substance which i; easily oxidized or is easily decom- posed by sulphuric acid may cause loss of nitrogen by a too rapid oxidation during the formation of ammonia, and in this way the process is shortened by sacrificing accuracy. Only in special cases, where the substance is very stable, is ii sate to use platinum chloride to hasten oxidation, as recommended by Ulsch. (C) He finds that nitrogen can be accurately determined by the original Kjeldahl method in all amides and ammonium bases, pyridin and ehinolin bodies, alkaloids, bitter principles, albuminoids, and allied substances; also most likely in the indol derivative-. In the following substances nitrogeu can not be determined satisfactorily by Kjel- dahl's original method. All nitro-, nitroso-, azo-, diazo-, hydrazo-, and amidoazo- bodies, compounds of nitric and nitrous acids, the hydrazines, and probably cyan- compounds. For the determination of nitrogen in these classes of compounds probably no geuera] rule can be given, but tho peculiarities of each group must be studied as Jodlbauer has studied the special modifications required by nitrates. E. Waller and H. C. Bowen ( Jour. Analy. Chem., 11, 293) give a sketch of the his- tory of the method, and mention modifications which they have adopted, containing nothing essentially new, except that they evaporate off* most of the sulphuric acid used in the oxidation of the substance and heat some time alter adding the perman- ganate ; an operation which seems hazardous. To the paper is appended a list of the papers which have been published on the Kjeldahl method. The New Jersey Agricultural Experimemt Station (Report N. J. Expt. Sta., L887, )». 169) has compared, in tho case of nineteen fertilizers, the results obtained by the modification of Kjeldahl's method as described by Scovell with those obtained by the absolute method as described in the Report of tho Connecticut Experiment Station for ]-;:>. The results showed very satisfactory agreement. The largest difference was .11 percent.; the average difference, .0(1 per cent. ; in eight cases the absolute method gave on the average .05 per cent, more nitrogen; in eleven cases the modi lied Kjeldahl method gave on I he average .06 per cent, mort! nitrogen. E. II. Parrington (Report Conn. Expt. Sta.. 1887, p. 126) reports comparative de- terminations of nit rogeu in t hirty-four samples of mixed fert i I izers containing nitrates by the Jodlbauer-Kjeldahl method and the absolute method as described in the re- ports of the same station for (.878 and 1879. Bis conclusions are as follows: "An inspection of these results show s t hat i a 68 per cent . of t he cases t he difference bet ween I he. t wo ii let hods was not over 0.1 per cent., and in 88 per Cent, of t he oases not over ,16 pet cent. The greatest difference was .21 per cent." " The plus differences are ii in Dumber; the minus differences, 20. The average of the former is .066 ; of the latter, .085. It is evident, then, that the two methods are about equally accurate." Otto Shonherr (Chem. Zeit., 12, '-'17) applies the a/.ot ometcr to the Kjeldahl del. r initiation. The digest ion Masks are graduated by I mark on the QG< k to a convenient volume (150 cubic centimeters) and after the oxidation has beeu effected in the usual way the contents of t he Mask are diluted, neutralized Approximately, made up to tho mark, and mixed, and an aliquot part (50 cubic centimeters) decomposed in the azo- t ometcr with 60 cubic centimeters hypobromite solution. Results accurate. i.\ Mei.h.i.i and E. i,\ Horiti (Jour. Boo. Chim, in.!.. ;. 63) purifj sulphuric from ammonium sulphate by adding about .05 gram potassium nitrate per [0 cubic centimeters bo the acid and heating aboul two hours. 45 Other methods. Houzeau (Pharm. Centr. 23, b27, 628, also Jour. Analyt. Chem. 2, 354) proposes a mixture to bo used in determining nitrogen, for the conversion of nitrogen in any- combined form into ammonia. Equal weights of sodium acetate and thisulphate are melted in their water of crystallization aud allowed to solidify. The mixture is pow- dered aud kept in well-stopped bottles. The combustion is made as follows : In the posterior end of the combustion tube are placed 2 grams of the above mixture and 2 grams of coarsely-pulverized soda-lime, and then a layer of soda-lime a few cen- timeters long. The finely-powdered substance to be analyzed (.5 gram if rich in nitrogen, 10-25 if ;i soil) is well mixed with 10 grams of the salt mixture and 10 grams of soda-lime, and lilled into the tube, followed by soda-lime and an asbestus or glass-wool plug. The combustion is made as usual, and the mixture at the rear end of the tube is used at the close of the operation for aspirating. The ammonia is absorbed and titrated in the usual way. The New Jersey Agricultural Experiment Station (Rep. N. J. Expt. Sta., 1887, p. 1GD) has determined the nitrogen in 140 samples of complete fertilizers, 22 samples of nitrogenous matter of high grade, and 19 samples of ground bone, both by the soda- lime method and the Kjeldahl method. The latter gave on the average .04 per cent, more nitrogen incomplete fertilizers, .05 per cent, more in high-grade nitrogenous matter, and .10 per cent, in bone. Reference should also be made to an elaborate series of experiments and observa- tions on the soda-lime method made in the laboratory of the Wesleyan University by or under the direction of Prof. W. O. Atwater (Am. Chem. Jour. 9, '311 ; 10, 111, 113, 1D7, 262). A complete abstract of these papers is not deemed necessary here. Pro- fessor At. water says, inclosing his paper : " The perfection to which Kjeldahl's method has lately been brought, and its accuracy, convenience, aud inexpensiveness, have led us to its use in this as in many other laboratories. Our experience leads us to decidedly prefer it to the soda-lime method, though we find it advantageous to use both, making one check the other. Bat the danger of incomplete ammonification of some classes of compounds, e. g., alkaloids, makes us feel it necessary to control both by the absolute method for all classes of substances except those for which they have been most thoroughly tested." RESULTS OF WORK DONE. Last November the chairman of your commit tee sent out to thirty-one official chem- ists and others five samples for nitrogen determinations. Twelve have reported re- sults. Sample No, 1, for the purpose of having the various chemists test the accuracy of their standard solutions, pipettes, burettes, and manipulation. Samples Nb. "2 and 5, for comparison of soda-lime method with Kjeldahl. Samples 1, 3, and 1, to compare the Kjeldahl modified, the Ruffle, and the absolute methods. in sample No. 4, a chloride was put in for the purpose of seeing whether by the Kjeldahl modified method Loss of nitrogen would not follow upon putting sulphuric acid on the .substance. COMPOS] i [ON OF i m: SAMPIJ S. i\ r e< i!'. d I No. 1. Potassium nitrate, C. P. (dried at LOO degrees) 13.85 No. 2. ( lol ton-seed meal, i. Sodium nitrate, C. P. dried at 100 degrees) B Ammonium sulphate, ('. P. (dried at l Legrees) 8.90 Cotton-seed meal l : Acid phosphate C- Theoretical per oent. of nitrogen 46 Per cent, nitrogen. No. 4. Sodium nitrate, C. P. (dried at 100 degrees) 10. 00 Cotton-seed meal 20. 00 Muriate of potash 10. 00 Acid phosphate 60. 00 Theoretical per cent, of nitrogen 3. 1C No. 5. A mixed tankage of the trade. No. 6. Sample of commercial fertilizer sent by Dr. E. H. Jenkins, Connecticut Experiment Station. No. 7. Pure sodium nitrate 16.47 All the materials, except the tankage, were sifted through a 40-mesh sieve and -well mixed before weighing out. The tankage was sifted through a 20-mesh sieve. The constituents, after weighing out, were thoroughly mixed, first with the spatula, on a large sheet of paper, then by running through a drug-mill several times and silting through the 20-mesh sieve. The mixture was then bottled and sealed as quickly as possible. A determination by ScovelFs method on 2.8 grams acid phosphate gave .02 per cent, of nitrogen. A similar experiment on the muriato of potash gave no nitrogen. RESULTS OBTAINED. Soda-lime method. Analyst. No. 2. No. 5. N.W. Lord H. W. Wiley W. J. G-aacoyne William Froar Wilkinson (Alabama Experiment Station) ... Michigan Carbon Works (W. L. Snyder). ... National Fertilizer Company (J. II. Kelley). Average 7.14 7.49 6. 02 7.70 7.47 7.1G 7.2G 2.97 2.85 3.01 2. 83 2.84 2. 98 2.89 2.91 Kjeldahl method. Analyst. No. 2. No. 5. 7. 62 6.63 7. 50 7. lit 7.01 7.21 7.38 2.98 2.80 2.81 ;;.!>) 3.00 2, '.17 2. 84 2. 9G William Frea* II W. Wiley E H Jenkins . \. M. Peter m L Seovell II \ Weber X. w Lord lb80lute UK lluxl. Analyst X... 1. \... I. No 8. \ W Lord 7. 23 3.70 \. A. Bennett !•:. ii. .1. akini Wmlon 13. 8G 3. 06 " 13. 80 7. 23 4. 10 3. 30 3.14 Unknown. 47 Kjeldald method modified for nitrates. Analyst Ho. 1. Xo. 2. Xo. 3. Xo. 4. Xo. 5. Xo. 6. Xo. 7. W. J. Gascoyne H. AV. Wilev H A. Webor 13.80 13. 73 3.93 3.70 3.92 3.98 3.93 3.93 3.96 3.54 3.13 3.13 3.22 3.09 2.90 3.18 3.15 2.48 3.03 E. 11. Jenkins* William Frear M. A. Scovell A. M. Peter G. C.Caldwell B.t 13. 53 13.86 13.77 13.77 13.86 "7.45" 7.47 7.82 3.11 2. 98 3.00 3.19 3.18 3.17 16.41 16.36 3.45 3.16 3.16 C.t D.t Average 13. 76 7.58 3.. SO 3.03 3.05 3.20 16.39 * Phenol used in place of salicylic acid, and digested one-half hour in the cold hefore heating, t Unknown. Iiiifflc method. Analyst. Xo. 1. Xo. 2. Xo. 3. Xo. 4. Xo. 5. H. W. "Wilev 13.83 12.04 13.68 13.50 13.86 7.47 7.42 7.27 3. 86 3.48 3.92 3.87 4.00 3.16 2.8! 3.10 2.90 2. 22 2.97 3.05 2.92 Wilkinson (ALabamaExperimi Xatioual Fertilizer Company (J. II. Kelley) Average 13.38 7.38 3.83 2.85 2.98 * Professor Frear reports the following results on Xos. 1, 3, and 4 by using charcoal in the place of sugar: Xo. 1, 13.78; Xo. 3, 4.16; Xo. 4, 3.43. AVERAGES. Method. Soda lime J Kjeldahl J Kjeldahl modified for J nitrates. ) Ruffle Sample. Xo. of analysts. Average. 7.26 2,91 7.:i8 2.96 13. 76 3 86 3. 03 Highest. 7.70 3.01 7.61 3.04 13.86 3.98 13. 86 4.00 3.16 Lowest. 7.14 2.83 0.03 2. 80 La 65 8.54 12.04 Difference. 0. 56 0.18 0.98 6.24 0.31 0. H 0.74 1.82 0.91 REMARKS. In the soda- lime method the results on No. 5 are quite satisfactory, the average being about .1 lower than with the Kjeldahl method. On sample No. 2, however, the results are very unsatisfactory. Even by leaving din one result' so far away from the others as to Bhow an error of the chemist, the re- sults vary too much to draw any other oonolusion than that, if the soda-lime method 1, much more car.- must be taken on such samplei a-seed meal. In the Kjeldahl method all results agree olosely with two exceptions J ad one ex eeption in No. 5. Leaving out these exceptions, the results of which are so tar out of the way as to refleel inaccuracy on the pari oftheanalyel rather than the method, tin' results are highly Batisfaofe On Xo. 2 the average is 7. r.», the highest 7.61, the low. ; the difference Del wren kighesl and low i 48 On Xo. 5, average 3.00, highest 3.04, lowest 2.91, difference 0.13 per cent. The Kjeldahl shows far better agreeing results than the soda-lime method and gives an average of .23 per ceut. higher results on No. 2 and .09 on Xo. 5. The Baffle method shows by the results that it is capable of bringing accurate re- sults, but that it seems of difficult manipulation by many chemists. On sample No. 1, leaving out one result evidently an error, the average is 13.72, the highest 13.86, the lowest 13.50, difference 0.3G. Ou sample No. 3, the average is 3.91, highest 4.00, lowest 3.86, difference .11. 1 i -ample Xo. 4. average 2.85, highest 3. Id, lowest 2.81, difference 0.35. The Kjeldahl method modified for nitrates, although a new method and used by the chemists for tho first time as au official method, gives very concordant results. On sample No. 1 the average is 13.76, and leaving out one low result 13.80, theory highest 13.86, Lowest 13.73, difference 0.13. On sample No. 3, average 3.91, highest 3.98, lowest 3.70, difference 0.28, calculated result 3.93. On sample Xo. 4, average 3.12, highest 3.22, lowest 3.09, difference 0.13, calculated 3.1(5. As compared with the Ruffle method the modified Kjeldahl shows higher results on No. 1 by .08, on X.j. 3 the average is identical, on No, 4 the average 0.27 per cent, greater. The absolute method has been used only by two chemists and the results are ai EO great a variance that no comparison can be made. RECOMMENDATIONS. Your committee would recommend the continuance of the five methods adopted last year, viz: (1) The absolute method. (2) The Kjeldahl method. (3) The Kjeldahl method modified for nitrates. (4) The Ruffle method. (.">) The soda-lime method. lint after a careful consideration of the results obtained above, they would recom- mend : (1) That the soda-lime method he given more in detail, in the hopes of obtaining more concordant results. (2) That detail changes lie made in the Kjeldahl methods. (3) That at tent ion be called to the use of zinc-dust, in the Kjeldahl modified method. Prof. H. A. Weber calls attention to the fact that unless the zinc-dust he in an im- palpable powder ami well mixed into the solution, 1 he nascent hydrogen tails to come in contact with all the uitro-oompounds and therefore reduction is incomplete. As an example he gives I he follow ing results: Sample. I " i r i . ly pulverized zinc. . Zim; Impalpable powder N... :;. No. 4. :;. 82 'J. Ml tespect full v submit ted. M. A. SCOVKLL, Chairman, V I. I.t PTON, Discussion of the report being in order, .Messrs. Fivar, Gascoyne, and Voorbees stated that they had found trouble with zinc dust, blanks BUOWinc sonic nil itreti. 49 Dr. Voorhees said that at the New Jersey Station they always made blanks for error, aud questioned whether commercial chemists did so. Professor Scovell said he had found that sulphuric acid often con- tained nitrogen, but the strictly C. P. from Eimer & Amend was quite free. Dr. Voorhees then gave an account of the experiments at his station with the Scovell modification of the Kjeldahl method, as published in their last annual report, and described a convenient form of iron gas- pipe condenser. Of the soda-lime method, Dr. Gascoyne said that the necessity for absence of a channel, as shown by Atwater, was an impor- tant factor. Dr. Voorhees pointed to the necessity of fine division of the sub- stance to at least GO mesh, while coarse samples could be used with the Kjeldahl method. Professor Scovell called attention to the generally bad results on cot- ton-seed and blood with soda lime, in the tables given in his report, and a general opinion that a careful understanding of the details of work- ing this method were very necessary, and without this results would easily fall too low. With the Scovell modification of the Kjeldahl method general satis- faction was expressed by those who had tried it carefully. The report was then adopted. Professor Stubbs then presented the report of the committee on sugar, as follows : REPORT OF THE COMMITTEE OX SUGAR. The increasing interest, in sugar-growing in this country is manifested by the large governmental aid recently given to the manufacture of sugar from sorghum and sugar- cane. This interest has created a demaud for chemical information relative to all sugar matters, from the raw juice to the manufactured articles, and to meet this de- mand is one of the increasing duties of this association. A Manual of Sugar Analy- sis, by J. EL Tucker, published by D. Van Nostrand, New York, in 1881$ contains all of the approved methods of analyses up to that date, many of which are still used without modifications in the laboratories of the world. Others have been slightly modified, while a few have been discarded as totally without merit. In all sugars and sugar products, including raw juices, the following determina- tions are needed for industrial work: (1) Total solids. (2) Percentage of sucrose. (:}) Percentage of invert SUgan. (4) Ash. /// rawjuioe$.—(l) Syrups and molasses: The total solids may be obtained in tin- ibi • owing manner: (1) By accurately graduated spindles, preferably Baix. (2) By taking specific gravit \ . (3) By evaporating to dryness a vreighed quantity mixed with ignited sand in ■ tailed dish. I A method (Wiley's) of drying and weighing a piece of filter paper, MJ L9 bj 2 pches, satnrating with juice or with mola i> disseminated over the pap means of a few drops of alcohol), drying and reweighing. 7717— No. 19 1 50 Of tho above methods only the third is absolutely correct. The fourth in our hands gives results slightly low, while second and first are only, approximate. However, the first method, when accurately graduated spindles are used, corrections made for temperature are sufficiently approximate for industrial work. A sufficient time should elapse between the pressing of the cane and the reading of the spindle to permit all the air bubbles to escape ; one-half hour is time enough. Sucrose. — For sugars, raw juices, and sirups sucrose is best determined by a simple polariscopic test, with tho following precautious: (1) Proper sampling of substance. (2) Careful weighing of sample. (3) Avoid excess of subacetate of lead. (4) Dissipate bubbles, which prevent accurate gauging, by a drop or two of ether. (5) Perfect filling of polariscope tube. (G) Careful readiug of polariscope. (7) Making all tests in duplicate. (8) Testing the accuracy of polariscope daily. Instead of direct weighing of substance the Ventzke's process may be used. Following the above (see Tucker, pages 261 and 262), all ripe sorghum juices, all sugars, and sirups of above grade give accurate results. With immature sorghum juices double polarizations show increased results. (See Table No. 1.) Table No. 1. — Analyses of varieties of sorghum from experimental plats of sugar experi- ment station ift Louisiana, made July 12 and 13; all immature. Number of variety from experimental plats. Total solids by 15 ux. * Per cent. sucrose, polarization. Tor cent, glucose. Single. Double. 1 0.8 7.0 9.8 <;. 5 1(1.9 9.0 11.7 10. G 13. 3 11.1 8.9 13.2 13.6 13.3 11.8 ll.fi 9.8 1.0 2.0 4.8 0.4 5.4 2. :; ('.. 4.8 6. •J.o 8. -J 0.0 7.0 3. ! 8. :; 1.0 2. 2 1.81 2. 9G 5. 1 8 2.00 <;. 9Q :; OS 7. 50 6. 07 8.81 4.18 LB3 7.78 0. 20 a OS B 12 6. •-'- 3.-J2 3.40 1.90 1.59 :;. 20 1 . 82 2.12 2, 13 2. 68 4.25 3.40 1.70 6.34 3.71 2. 75 8.40 8.30 2.95 3 4 5 . <; 7 8 9 10 11 12 H. ] ") 10 17 18 19 With molasses of all kinds, a1 Least from Louisiana cane, simple polarization gives results always too low, especially if during the pr< cess of manufacture much sugar lias !„.,,, inverted. Double polarization by Clerget'a method of inversion is then usually ,,.„.,! f or determining the Buorose In these bodies, and when invert sngar is the only optically active body, is approximately accurate. When optically active bodies other than invert Bugar are present, recourse mual be had to inversion and estimations by Fehling'a or Violette'a methods, Tuchamidt has confirmed the accuracy of Clerget'i process, and Bernhardt and Bellman have perfected it, giving us what is now known to as set llerget'a met hod. Creydt, on the other band, has corrected the constant 144 of Clerget and substi- tuted therefore L42. Bo further preaoxpbea pClof L.1888 specific gravity, equal to :;- ,„., cent ofacid, and washes with thisacid the bone charcoal need for filter, drioa, ftn d pulverizes it. Hewfeld thinka the exact truth as tothia constant ia cotyel known on account of the difficulties, first, of maintaining the temperature of observe 51 tion ; second, the nature of the so-called "pure sugar" used. Landolt recently adds other errors, among others, that 1° Ventzke (white light) = 0. 34455 circular degrees (sodium light) is too high, and that diverse active suhstauces give numbers sensibly different. He hag fouud with Dr. Rallyan for 1° Ventzke (white light). Sucrose = 0.3465, dextrose =0.3448, invert sugar =0.3433 circular degrees. Landolt finds the constant of Clerget to bo 142.4, ail expression believed to-day to be the most exact. Hcrzfeld gives a perfected Clerget method which has the following differences : Water of local temperature surrounds the tube, which cools the solution, at once permit- ting accurate reading of temperature. The use of washed and dried bone-black after Eeichart and Bellman, and calculation of results according to the formula of Clerget, modified by Landolt, viz: 100S R: 142.4— vr The following table of work, carefully done by my assistant, Prof. W. L. Hutchinson, with molasses from different plantations in Louisiana, shows results by double polarization and inversion and estimation with Folding's solution. Table No. 2. "Where from— plantation. Belle Chasec John Crossly &. Sons Joseph (lair Boujere Belle Terre Woodlawn Crescent tfary Perseverance David Orange- Grove Palo Alto Hester B 3 © u to . o o C CO o rctnl. Mi- cro s o by double- po- larisoope. « O 8 f- '-Z CD to n H * fc O P4 41.4 1. 4039 78.2 39.59 24.32 15.03 39. 65 41.8 1.4092 79.0 41. 68 22.69 14.44 40. 61 40.6 1.3933 76.6 34.91 29.19 9.05 32. 25 40.7 1.3940 76.7 37. 62 24. 77 15.32 37. 53 41.6 1. 4005 78.6 42. 57 14.63 41.30 42. n 1.4165 80.1 33.42 33.75 14.49 32. 79 42.3 1.4165 80.1 36.44 28. 42 14. 92 37. 67 42.3 1.4165 80.1 38.82 26. FG 10. 72 39. 08 4:?. 2 1.4287 81.9 40. 31 20. 00 22.93 43.51 42.3 1.4165 80.1 37. 62 26. 92 16. 72 36. 97 42. 4 1.4179 80.3 42. 25 26. 73 13 < 5 39. 94 42. G 1.4205 80.7 38.14 2ft. 96 16.43 42.5 1.4192 80.5 34.72 30.17 1&52 33.14 ■ J3 — To 65. 06 65. 4 4 63.14 61.28 66. 26 68. 22 67.81 65 K 65.84 65. 06 While the concurrence of results in the main are satisfactory, there are Borne which conld not be made to agree, suggesting strongly either a fault in the method or other optically active bodies than invert sugar, perhaps both; and while for the present wo recommend Clerget's method we earnestly request an investigation of Herzfeld's modifi- cation of Clerget's method during the coming year by all sugar chemists. INVERT SUGAR. The exact estimation of invert sugar in sngar matters has been recently ably in- vestigated and discussed. Heretofore invert sogar has been determined by a simple titration with the cupric tartrate Liquor <>f Fehling, Violette, et< . The formulas foi the pre pa rat ion of these solutions arc uumeroas. They all contain sulphate of copper, rochelle salts, and caustic alkali and present little real difference, sfeisaland Hei fold have described a process based opon the weight of copper reduced from the Fehling solution when accompanied by particular conditions. In spite oJ ion it has fonnd only limited application, especially in France, where its nu- merous difficulties have prevented its ii- 1 ' on an i scale. Beggar! employs Fehling's solution, which he keeps in bottles I bouch< es ;i L'emeri) and which are placed in a dark place. To titrate this liquid they dissolvi mof pure sugar in a little water with 2 cubic centimeters of hydrochloric acid, and a tier adding •'■' cubic « en- 52 tiineters of water it is gently heated for ten minutes at 70° C. This solution is cooled and saturated with bicarbouate of soda and raised to one liter; .'J5 cubic centi- meters of this liquid will reduce 5 cubic centimeters of Fehliug's solution, the end reaction being determined as usual by means of ferrocyanide of potash and aeetieacid, using Wiley's tubes. To preserve a standard solution of invert sugar he prepares a to per oent. solution and make strongly alkaline. In this way it will be preserved many months. "With this solution he daily tests his Fehliug's solution. This constitutes the only merit of this process, a merit which should be observed in every process using the oaprio tartrate process. Patterson's process is based upon another principle. To 100 cubic centimeters of Fehling's solution he adds a quantity of sugar insufficient to precipitate all the copper, and determines excess of copper by a titration with a normal invert sugar solution containing .002 grams per cubic centimeter. Ilerzfeld rightly condemns the process, since no correct conclusions from the use of the cupric tartrate solution can he drawn without considering the strength of ebullition and influence of crystallizable sugar upon the reducing power of invert sugar. All the copper liquids so far proposed, although different in composition, have one common property, viz, they all contain a salt of copper, a double tartrate of potash and soda, and caustic or carbonated alkali. They are all reduced not only by invert sugar but also by dextrine and sometimes even by sucrose. After a certain time t bey undergo alteration, deposit the oxide of copper, especially if exposed to light. Soldaiui's solution does not possess these objectionable qualities, at least so says Degener, who has studied carefully its properties. This liquid is prepared, accordiugto Degener, in following manner: (1) 40 grams of sulphate of copper is dissolved in water, and. in another vessel, 40 grams of carbonate o f soda is also dissolved in water. The two solutions are mixed and the copper pre- cipitated in the state of hydrobasic carbonate according to following equation: 2 Cu S0 4 +2Na 2 CO a -l-H 2 0=CuCO :i CuOH i O-fCO,+2Na 2 S0 4 . The precipitate is washed with cold water and dried. This precipitate (15 grams) is added to a very con- centrated and boiling solution of bicarbonate of potash (about 415 grams) and agitated until the whole is completely, or nearly so, dissolved, water is added to form a \ olume of 1,400 cubic centimeters, and the whole mass heated for two hours upon a water bath. The insoluble matter is filtered and the filtrate, alter cooling, Is of a deep blue color and has a density of about 22 .5 Baume* (1.165). The sensibility of this liquid is so great that it gives a decided reaction with .0014 grams of in vert sugar. The presence of sucrose in the 'solution increases still this sensibility. Ammoniaca] salts are not hurt I'ul to t he read ion, but they do sometimes retain in solution a slight quant ity of copper. It is then indispensable to boil the solution for five minutes at least in order to drive off ammonia. The quantitative test for sugar is made in the following manner : fifty cubic centimeters of Soldaiui's sol tit ion is boiled for five minutes upon a water bath, and L5 cubic centimeters of the sugar Liquid (containing 100 grama of matter in 100 cubic centimeters of water) added, and boiled again for live minutes. After a rapid cooling the Liquid is thrown upon a Swedish filter and washed with water until every t race of a blue color disappears. Then the paper is renio\ ed and the pre- cipitate examined, which should be of a (dear red color. Bodenbaender and Schiller (Zeitschrifl dee Vereina Ettr Riibenzucker-Industrie, 1 387, L38) have determined the following quantitative process with this Liquid: one hundred to 150 cubic ecu 1 i meters of Soldaiui's sol ui ion is placed in an Krlenme\ er Mask ami heated for five minutes over a gas j''t. A solution containing LO grams of sugar (clarified with subacetate of lead if necessary I Is added, and heated again for five min- ntes, always over a direct dame. This precipitates all the copper, to which is added LOO cubic centimeters of distilled' water, in order to cool the mas-. \ ery quickly. The t roil 1)1 e dliquid is lilt end | preferably through a Bohxlol or 1 1 civ field filter) and washed 53 with distilled water. This oxide of copper is reduced in a current of hydrogen, and the copper is weighed in a metallic form, and this weight multiplied by .3546 gives the weight of invert sugar. By this method .01 per cent, of invert sugar can he accu- rately determined. Siderslcy has recently offered a new volumetric method based upon the use of Sol- daini's solution. With sugars the same method as is now in use with Fehling's solu- tion can easily be followed, watching the disappearauce of the blue color, aud testing the end with ferrocyauide of potash and acetic acid. This process oilers no serious objections common to Fehling's solution, but is inapplicable to colored sugar solutions, such as molasses, etc. For the last the following is recommended : t went y-li ve grams of molasses is dissolved in 100 cubic centimeters of water and subacetate of lead added in sufficient quantities to precipitate the impurities, and the volume raised to 200 cubic centimeters and filtered. To 100 cubic centimeters of the filtrate are added 25 cubic centimeters of concentrated solution of carbonate of soda, agitated and filtered again. One hundred cubic centimeters of the second filtrate with excess of lead removed is taken for analysis. On the other hand, 100 cubic centimeters of Sol- daini's solution is added to a flask of Bohemian glass and heated five minutes over an open flame. The sugar solution is now added a little by little, and the heating continued for five minutes. Finally, the heat is withdrawn and cooled by turning in 100 cubic centimeters of cold water, and filtered through a Swedish filter, washed with hot water, letting each washing run off before another addition. Three or four washings will generally remove completely the alkaline reaction. The precipitate is then washed through a hole in the filter into a flask, removing the last trace of eopper. Twenty-five cubic centimeters of normal sulphuric acid is added, with two or three crystals of chlorate of potash, and the whole gently heated to dissolve completely the oxide of copper, which is transformed into copper sulphate, according toequafcion,3Ca J 0+6H 9 S0 4 +KC10 3 = 6CHS04+6H 4 0+KCl. The excess of sulphu- ric acid is determined by a standard ammonia solution (semi-normal) of which the best indicator is the sulphate of copper itself. When the deep blue eolor gives place to a greenish tinge the titration is completed. The method of titration is performed as follows: Having cooled the contents of the flask, a quantity of ammonia equiva- lent to 2.") cubie centimeters of normal sulphuric acid added. From a burette gradu- ated into one-tenth cubic centimeters standard sulphuric acid is dropped in drop by drop, agitating after eaeli addition. The blue color disappears with each addition to reappear after -baking. When the last trace of ammonia is saturated the titration is complete, which is known by a very feeble greenish tinge. The number of cubic centimeters is read from the burette, which is equivalent to the copper precipitated. The equivalent of copper is 31.7, and the normal acid equivalent is ,0317 of copper. Multiplying the copper found by 3546 we have the invert sugar. A blank titration is n< •.■ded to accurately determine the slight exoess which gives the pale green 1 tnge. This process is to be highly recommended if experiments show it to be as accurate in our hands a- it ha- in Prance. (4) Ank.—Vov tke determination of ash, Seheibler's process oi ignition with 11 s< >, and deducting one-tenth is recommended. Fox details, see Tucker's Manual. pag< Your committee, in conclusion, recommends thai for technical work the following is admissible: i Taking the total solids with an accurately graduating Brix spindle and cor- rect ing for temporal ore. (2) A polariscopio test for raw juices, low sirups and sugars, using the precautions men! ioned in t Ins report . (:{) Foi heavy simps, molasses, and tank bottoms, double polarixation with Cl< tables. (4) For invert sugar, use Fehling's or Violettfa solutions, preferably the last, as an end reaction ferrocyanide of potash and aosiioaeii; bast accomplished by dm of Wiley's tubes.' 54 (5) Ignition with sulphuric acid at a low red heat and deducting ono-tenth for ash. For laboratory work the above may bo substituted by some of the more accurate methods given above. Respectfully submitted. W. C. Stubbs, Chairman. HE MARKS ON PROFESSOR STUBBS' S REPORT. By H. W. Wiley. Mr. Pabsident. Id connection with the report of Professor Stubbs ou sugar analy- ':icli has just been presented to the association, I beg leave to call the attention of the members to a standard polariscope which I have had constructed by Schmfdt ensoh in Berlin. This instrument is the double compensating apparatus manu- factured by that firm, which has already gained great favor among chemists. Tho apparatus which I show you was specially constructed for the use of this Department, and is capable of receiving an observation tube 600 millimeters in length. By means of the double compensating apparatus any error which may be due to an inaccuracy in the quartz wedge is reduced to a minimum. Tho apparatus has tho advantage of using ordinary lamp-light, which is of great importance when outsido- station work is to be considered, where tho difficulty of securing a monochromatic} flame is great. This instrument has lately been examined by Professor Andrews, the expert employed by the United States Treasury to investigate the methods of sugar analysis employed in the United States custom-houses. He proved it to be as nearly exact as any instrument can be made. It is not necessary to mention that for general purposes of research a rotation instrument like the large model Laurent is preferable to any form of compensating apparatus; but in sugar work, where an arbitrary scale is ii-t [ess than 250 cubic centimeters. Mix the washings. Take an aliquot (cor- responding to \ gram) and determine phosphoric acid, as under total phosphoric acid. (1) Citrate insoluble phosphoric acid. — Wash the residue of die treat- ment with water into a 200 cubic centimeter flask with 100 cubic centi- meters of strictly neutral ammonium citrate solution of L.09 density. prepared as hereafter directed. Cork the flask securely and place it in a water bath, the water of which stands at G5° C. (The water bath should be <>f such a size that the introduction of the cold flask or flasks shall not cause a reduction of the temperature of the bath of more than 2 0.) Raising the temperature ad rapidly as practicable to 65 < '., which is subsequently maintained, digest with frequent shakings for thirty minutes from flic instant of insertion, filter the warm solution quickly (best with lilter-pump), and wash with water of ordinary tem- perature. Transfer the filter and its contents to a capsule, ignite until the organic matter is destroyed, treat with lo to 15 oabto centimeters 58 of concentrated hydrochloric or nitric acid, digest over a low Ha me until the phosphate is dissolved, dilute to -200 cubic centimeters, mix, pass through a dry filter, take an aliquot and determine phosphoric acid as under total. In case a determination of citrate-insoluble phosphoric acid is re- quired in non-acidulated goods, it is to be made by treating 2 grams of the phosphatic material, without previous washing with water, precisely in the way above described, except that in case the substance contains much animal matter (bone, fish, etc.), the residue insoluble in ammon- ium citrate is to be treated by one of the processes described below. (5) Total phosphoric acid. — Weigh 2 grams and treat by one of the following methods: (1) Evaporation with 5 cubic centimeters magne- sium nitrate, ignition, and solution in acid. (2) Solution in 30 cubic centimeters concentrated nitric acid with a small quantity of hydro- chloric acid. (3) Add 30 cubic centimeters concentrated hydrochloric acid, heat, and add cautiously and in small quantities at a time about 0.5 gram of finely pulverized potassium chlorate. Boil gently until all phosphates are dissolved and all organic matter destroyed; dilute to 200 cubic centimeters; mix and pass through a dry filter; take 50 cubic centimeters of filtrate ; neutralize with am- monia (in case hydrochloric acid has been used as a solvent add about 15 grams dry ammonium nitrate or its equivalent). To the hot solu- tions for every decigram of P 2 5 that is present, add 50 cubic centi- meters of molybdic solution. Digest at about 65° C. for one hour, filter, and wash with water or ammonium nitrate solution. (Test the filtrate by renewed digestion and addition of more molybdic solution.) Dis- solve the precipitate on the filter with ammonia and hot water and wash into a beaker to a bulk of not more than 100 cubic centimeters. Nearly neutralize with hydrochloric acid, cool, and add magnesia mix- ture from a burette; add slowly (one drop per second), stirring vigor- ously. After fifteen minutes add 30 cubic centimeters of ammonia solution of density 0.95. Let stand several hours (two hours is usually enough). Filter, wash with dilute ammonia, ignite intensely lor ten minutes, and weigh. (G) Citrate-soluble phosphoric acid. The sum of the water soluble and citric insoluble subtracted from the total gives the citrate soluble. PREPARATION OF REAGENTS. (1) To prepare ammonium citrate solution. — Mix 370 grams of com- mercial citric acid with l,5()() cubic centimeters of water j nearly neu- tralize with crushed commercial carbonate of ammonia ; heat to < the carbonic acid ; cool; add ammonia until exactly neutral (testing by Saturated alcoholic, solution of coralline) and bring to volume of two liters. Test the gravity, which should be L.09 at 20° 0., before using. (2) To prepare molybdic solution. — Dissolve loo grams of molybdic acid in inn grams <>r 1 1 7 cubic cent [meters of ammonia of specific grav- 59 ity 0.06, and pour the solution thus obtained into 1,500 grams or 1,250 cubic centimeters of nitric acid of specific gravity 1.20. Keep the mix- ture in a warm place for several days, or until a portion heated to 40° O. deposits no yellow precipitate of ammonium phospho-molybdate. Decant the solution from any sediment, and preserve in glass-stoppered vessels. (3) To prepare ammonium nitrate solution. — Dissolve 200 grams of commercial ammonium nitrate in water and bring to a volume of two liters. (4) To prepare magnesia mixture. — Dissolve 22 grams of recently ig- nited calcined magnesia in dilute hydrochloric acid, avoiding excess of the latter. Add a little calcined magnesia in excess, and boil a few minutes to precipitate iron, alumina, and phosphoric acid; filter, add 280 grams of ammonium chloride, 700 cubic centimeters of ammonia of specific gravity 0.96, and water enough to mate the volume of two liters. Instead of the solution of 22 grams of calcined magnesia 110 grams of crystallized magnesium chloride (MgCl 2 , GILO) may be used. (5) Dilute ammonia for washing. — One volume ammonia of specific gravity 0.96 mixed with three volumes of water, or usually one volume of concentrated ammonia with six volumes of water. (6) Nitrate of magnesia. — Dissolve 320 grams of calcined magnesia in nitric acid, avoiding an excess of the latter; then add a little cal- cined magnesia in excess; boil; filter from excess of magnesia, ferric oxide, etc., and bring to volume of two liters. METHODS OF DETERMINING POTASH. METHOD OF LINDO AS MODIFIED BY GLADDING. (1) Superphosphates. — Boil 10 grams of the fertilizer with 300 cubic centimeters of water for ten minutes. Cool the solution ; add a little oxalate of ammonia and then ammonia in slight excess, thus precipi- tating all phosphate and sulphate of lime, oxide of iron, and alumina, etc.; makeup to 500 cubic centimeters, mix thoroughly and filter through a dry filter; take 50 cubic centimeters corresponding to 1 gram, evap- orate nearly to dryness, add 1 cubic centimeter of dilute H 2 SO< (I to 1), and evaporate to dryness and ignite to whiteness. As all the potash is in form of sulphate, no loss need be apprehended by volatilization of potash, and a full red heat must be used until the residue is perfectly white. This residue is dissolved in hot water plus a tew drops of IIC1; 5 cubio centimeters of a solution of pure NaGl (containing L'O grams NaCi to the liter) and an excess of platinum solution (l cubic centi- meters) are now added. This solutionis then evaporated to dryness in a small dish, the residue taken up with a little water, sullieient to dissolve it, and st rong alcohol added. The precipitate IS washed thor- oughly with alcohol by deeantation and on filter as usual. The wa>h- 60 ing should be continued even after the filtrate is colorless. Ten cubic centimeters of the NH 4 01 solution prepared as directed are now run through the filter, or the washing may be performed in the dish. These 10 cubic centimeters will contain the bulk of the impurities, and are thrown away. Fresh portions of 10 cubic centimeters NH 4 Cl are now run through the filter several times (five or six). The filter is then washed thoroughly with pure alcohol, dried, and weighed as usual. The platinum solution used contains 1 gram metallic platinum in every L0 cubic centimeters. (2) Muriates of potash. — In the analysis of these salts an aliquot portion, containing .500 gram, is evaporated with 10 cubic centimeters platinum solution plus a few drops of HC1, and washed as before. (.':>) Sulphate of potash, lainite, etc.— In the analysis of these salts an aliquot portion containing .500 gram is taken, .250 gram of XaCl added, plus a few drops of IIC1, and the whole evaporated with 15 cubic cen- timeters platinum solution. In this case special care must be taken, in the washing with alcohol, to remove all the double chloride of plat- inum and sodium. The washing should be continued for some time after the filtrate is colorless. Twenty-five cubic centimeters of the XIJ 4 C1 solution are employed, instead of 10 cubic centimeters, and the 25 cubic centimeters poured through at least six times to remove all sulphates and chlorides. Wash finally with alcohol, dry, and weigh as usual. To prepare the washing solution of NH 4 C1, place in a bottle 600 cubic centimeters H 2 0, 100 grams of NH 4 C1; shake till dissolved. Now pul- verize 5 or 10 grams of K 2 PtCl c , put in the bottle, and shake at inter- vals for six or eight hours; let settle over night; then filter oil* liquid into a second bottle. The first bottle is then ready for a preparation of a fresh supply when needed. ALTERNATE METHOD. In case the potash is contained in organic compounds, like tobacco stems, cotton-seed hulls, etc., it is to be saturated with strong sulphuric acid and ignited in a muffle to destroy organic matter. Pulverize the fertilizer (200 or 300 grams) in a mortar; take 10 grams, boil for teu minutes with 200 cubic centimeters water, and after cooling, and with- out filtering, make up to 1,000 cubic; centimeters, and filter through a dry paper, [f the sample have L0 to L5 per cent. K»0 (kainite), take 50 cubic centimeters of the filtrate; if from 2 to :\ per cent. K.o (ordinary potash fertilizers), take IOO cubic centimeters of the filtrate. In each case make the volume up to L50 cubic centimeters, heat to 100°, and add, drop by drop, with constant stirring, slight excess of barium chloride; without filtering, in the same manner, add barium hydrate in slight excess. Meat, filter, and wash until precipitate is free of chlo- rides. Add to filtrate I cubic centimeter strong ammonium hydrate, and then a saturated solution of ammonium carbonate until excess of 61 barium is precipitated. Heat. Add now, in fine powder, 0.5 gram pure oxalic acid or 0.75 gram ammonium oxalate. Filter, wash free of chlorides, evaporate filtrate to dryness in a platiuum dish, and, holding dish with crucible tongs, ignite carefully over the free flame below red heat until all volatile matter is driven off. The residue is now digested with hot water, filtered through a small filter, and washed with successive small portions of water until the filtrate amounts to 30 cubic centimeters or more. To this filtrate, after adding two drops of strong hydrochloric acid, is added, in a porcelain dish, 5 to 10 cubic centimeters of a solution of 10 grams of platinic chlo- ride in 100 cubic centimeters of water. The mixture is now evaporated on the water-bath to a thick sirup, or further, as above treated with strong alcohol, wasbed by decantation, collected in a Gooch crucible or other form of filter, washed with strong alcohol, afterwards with 5 cubic centimeters ether, dried for thirty minutes at 100° C, and weighed. It is desirable, if there is an appearance of foreign matter in the double salt, that it should be washed, according to the previous method, with 10 cubic centimeters of the half-concentrated solution of NH 4 C1, which has been saturated by shaking with K 2 PtCl G , as recommended by Gladding. The use of the factor 0.305G for converting K 2 PtCl G to KC1 and 0.19308 for converting to KjO are continued. METHODS FOR THE DETERMINATION OF NITROGEN. THE ABSOLUTE OR CUPRIC OXIDE METHOD. (Applicable to all nitrogeo determinations.) The apparatus and re-agents needed are as follows: APPARATUS. Combustion tube of best hard Bohemian glass, about 2G inches long and one-half inch Internal diameter. .1 zotometi r of at least 100 cubic centimeters capacity, accurately cali- brated. Sprengel mercury air pump. Small paper scoop, easily made from stiff writing-paper. RE-AGKN i 9. Ouprie oxide (coarse). — Wire form; to be ignited and cooled before using. Fine ouprie oxide. — Prepared by pounding ordinary cupric oxide in mortar. Metallic ooppi r, — ( Granulated copper or line copper gauze reduced and cooled in stream of hydrogen. 62 Sodium bicarbonate. — Free from organic matter. Caustic potash solution. — Dissolve commercial stick potash in less than its weight of water so that crystals are deposited on cooling. When absorption of carbonic acid ceases to be prompt solution must be discarded. LOADING TUBE. Of ordinary commercial fertilizers takel to 2 grams for analysis. In the case of highly nitrogenous substances the amount to be taken must be regulated by the amount of nitrogen estimated to be present. Fill tube as follows: (1) About 2 inches of coarse cupric oxide. (2) Place on the small paper scoop enough of the fine cupric oxide to fill, after having been mixed with the substance to be analyzed, about 1 inches of the tube; pour on this the substance, rinsing watch glass with a little of the fine oxide, and mix thoroughly with spatula; pour into tube, rinsing the scoop with a little fine oxide. (3) About 12 inches of coarse cupric oxide. (4) About 3 inches of metallic copper. (5) About 2.] inches of coarse cupric oxide (anterior layer). (0) Small plug of asbestos. (7) Eight-tenths to 1 gram of sodium bicarbonate. (8) Large, loose plug of asbestos; place tube in furnace, leaving about 1 inch of it pro- jecting; connect with pump by rubber stopper smeared with glycerine, taking care to make conuection perfectly tight. OPERATION. Exhaust air from tube by means of pump. When a vacuum has been obtained allow flow of mercury to continue, light gas under that part of tube containing metallic copper, anterior layer of cupric oxide (see 5th above) and bicarbonate of soda. As soon as vacuum is destroyed and apparatus filled with carbonic acid gas, shut off the tlow of mer- cury and at once introduce the delivery tube of the pump into the re- ceiving arm of the azotometer, and just below the surface of the mercury seal of the azotometer, so that the escaping bubbles will pass into the air ami not into the azotometer, thus avoiding the useless saturation of the caustic; potash solution. When the How of carbonic acid has very nearly or completely ceased, the delivery tube down into the receiving arm, so that the bubbles will escape into the azotometer. Light the jets under the 12-mch layer of oxide, heal gently tor a few moments to drive out any moisture that may be present, and bring to red heat. Meat gradually mixture of substance and oxide, lighting one jet at a time. Avoid too rapid evolution of bub- bles, which should be' allowed to escape at rate of about one per second or a little faster. When the jets under mixture have all been turned on, light jets under layer Of oxide at end of tube. When evolution of gas has ceased turn out all the lights except those under the metallic copper and anterior layer ot oxide, and allow to cool fora few moments. Exhaust with pump ami remove azotometer before Bow of mercury is stopped. Break eon- 63 nection of tube with pump, stop flow of mercury, and extinguish lights Allow azotometer to stand for at least an hour or cool with stream of water until permanent volume and temperature are reached. Adjust accurately the level of the KOII solution in bulb to that in azotometer, note volume of gas, temperature, and height of barometer; make calculations as usual. The labor of calculation may be much diminished by the use of the tables prepared by Messrs. Battle aud Dancy, of the North Carolina Experiment Station (Raleigh, N. C). The above details are, with some modifications, those given in the re- port of the Connecticut Station for 1870 (p. 124), which may be consulted for details of apparatus, should such details be desired. THE KJELDAHL 3IETHOD. (Not applicable in presence of nitrates.) APPARATUS AND REAGBNT8. (1) Hydrochloric acid, whose absolute strength has been determined (a) by precipitating with silver nitrate and weighing the silver chloride, (b) by sodium carbonate, as described in Fresenius's Quantitative Analysis, second American edition, page G80, and (c) by determining the amount neutralized by the distillate from a weighed quantity of pure ammonium chloride boiled with an excess of sodium hydrate. Half normal acid, ?. ) Granulated /inc. (7) A solution of 10 grams of commercial potassium sulphide in one liter of water. (5) A saturated solution of sodium hydrate 1 tree from nit rales, which ometimes added in the process of manufacture to destroy organic mailer ami improve the color of the product. That of the (ireenbank Alkali Company is of good quality. (!>) Solution of cochineal prepared according to Fresenius's Quanti- tative Analysis, second American edition, page 679, (1(>) Kjeldahl digestion flasks of hard, moderately thick, well an nealed glass. These flasks are about 9 inches long,, with a round, pear shaped bottom, having a maximum diameter of 2j inches, and tapering 64 out gradually in a long neck, which is three-fourths of an inch in diam- eter at the narrowest part, and flared a little at the edge. The total capacity is 225 to UoO cubic centimeters. (11) Distillation flasks of ordinary shape, 550 cubic centimeters ca- pacity, and fitted with a rubber stopper and a bulb tube above to pre- vent the possibility of sodium hydrate being carried over mechanically during distillation. The bulbs are about 1A inches in diameter, the tubes being the same diameter as the condenser and cut off obliquely at the lower end. This is adjusted to the tube of the condenser by a rubber tube. (12) A condenser.— Several forms have been described, no one of which is equally convenient for all laboratories. The essential thing is that the tube which carries the steam to be condensed shall be of block-tin. The upper ends of the tin tubes should be bent so that the glass con- nections may have a slope toward the distilling flasks. All kinds of glass are decomposed by steam and ammonia vapor, and will give up alkali enough to impair accuracy. (See Kreussler and llenzold, Ber. Berichte, XVI 1, 34.) The condenser in use in the laboratory of the Connecticut Experiment Station, devised by Professor Johnson, con- sists of a copper tank, supported by a wooden frame, so that its bottom is 11 inches above the work-bench on which it stands. This tank is 1(> inches high, 32 inches long, and 3 inches wide from front to back, widening above to G inches. It is provided with a water supply tube, which goes to the bottom, and a larger overflow pipe above. The block- tin condensing tubes, whose external diameter is three-eighths of an inch, seven in number, enter the tank through holes in the front side of it near the top, above the level of the overflow, and pass down per- pendicularly through the lank and out through rubber stoppers tightly fitted into holes in the bottom. They project about 1J inches below the bottom of the tank, and are connected by short rubber tubes with glass bulb tubes of the usual shape, which dip into precipitating beakers or Brlenmeyer flasks of about 300 cubic centimeters capacity. The titra- tion can be made directly in them. The distillation flasks are supported on a sheet-iron shelf, attached to the wooden frame that supports the tank 8 in front Of the latter. Where each flask is to stand a circular hole is cut, with three projecting lips, which support the wire gauze or asbostus under the flask, and three other lips, which hold the flask in place and prevent its moving laterally out of place while distillation is going on. Below this sheet-iron shelf is a metal tube carrying seven Bun sen burners, each with a stop-cock like those of a gas-eombustiou furnace. These burners are of a larger diameter at the top, Which pre- vents smoking when covered with line gauze to prevent the llame from striking back. I.; Tin- stand for holding the digestion flasks consists of a pan of sheet iron '2\) inches long by 8 inches wide, on the front of which is fastened a shelf of sheet iron as long as the pan,. ~> inches wide and 1 inches 65 high. In this are cut six holes 1| inches in diameter. At the back of the pan is a stout wire running lengthwise of the stand, 8 inches high, with a bend or depression opposite each hole in the shelf. The diges- tion flask rests with its lower part over a hole in tbe shelf and its neck in one of the depressions in the wire frame, which holds it securely in position. Heat is supplied by low Bunsen burners below the shelf. With a little care the naked flame can be applied directly to the flask without danger. THE DETERMINATION. (1) The digestion. — Seven-tenths to 2.8 grams of the substance to be analyzed, according to its proportion of nitrogen, is brought into a di- gestion flask with approximately .7 gram of mercuric oxide and 20 cubic centimeters of sulphuric acid. Tbe flask is placed on the frame above described in an inclined position, and heated below the boiling point of the acid for from 5 to 15 minutes, or until frothing has ceased. If the mixture froths badly, a small piece of paraffine may be added to prevent it. Tbe heat is then raised until the acid boils briskly. Ko further attention is required till the contents of tbe flask have become a clear liquid, which is colorless, or at least has only a very pale straw color. Tbe flask is then removed from the frame, held upright, and, while still hot, potassium permanganate is dropped in carefully and in small quantity at a time till, after shaking, the liquid remains of a green or purple color. (2) The distillation. — After cooling, the contents of tbe flask are trans- ferred to the distilling flask with about 200 cubic centimeters of water, and to this a few pieces of granulated zinc and 25 cubic centimeters of potassium sulphide solution are added, shaking the flask to mix its contents. Next add 50 cubic centimeters of the soda solution, or suf- ficient to make the reaction strongly alkaline, pouring it down the side of tbe flask so that it does not mix at once with the acid solution. Connect tbe flask with the condenser, mix the contents by shaking, and distill until all ammonia has passed over into tbe standard acid. The liist 150 cubic centimeters of the distillate will generally contain all of the ammonia. This operation usually requires from forty minutes to one hour and a half. The distillate is then titrated with standard ammonia. Tbe use of mercuric oxide in this operation greatly shortens the time necessary for digestion, which Is rarely over an hour and a half in cases of substances most difficult to oxidize, and is more commonly less than an hour. In most cases the use of potassium permanganate is quite unnecessary, but it is believed that in exceptional cases it is required loi com pie te oxidation, and in view of the uncertainty it is always used. Potassium sulphide removes all mercury from solution, and so prevents the formation of mercuro-anunonium compounds which are not com pletcly decomposed by soda solution. The addition of zinc gives rise, to an evolution of hydrogen and prevent^ \ ml.Mit bumpm-. Prevjoqg 7717— No. 1!) 5 66 to use the reagents should be tested by a blank experiment with sugar, which will partially reduce any nitrates that are present which might otherwise escape notice. KJELDAIIL METHOD MODIFIED TO INCLUDE THE NITROGEN OF NI- TBATES.* (Applicable to all fertilizers containing nitrates.) Besides the reagents and apparatus given under the Kjeldahl method, there will be needed — (1) Zinc dust. This should be an impalpable powder: granulated zinc or zinc filings will not answer. (2) Commercial salicylic acid. THE DETERMINATION. Bring from .7 to IA grains of the substance to be analyzed into a Kjeldahl digesting flask, add to this 30 cubic centimeters of sulphuric acid containing 2 grams of salicylic acid, and shake thoroughly; then add gradually 3 grams of zinc dust, shaking the contents of the flask at the same time. Finally place the flask on the stand for holding the digestion flasks, where it is heated over a low flame until all danger from frothing has passed. The heat is then raised until the acid b »ils briskly, and the boiling continued until white fumes no longer pour out of the flask. This requires about live or ten minutes. Add now ap- proximately .7 gram mercuric oxide, and continue the boiling until the liquid in the flask is colorless, or nearly so. (In case tin 1 contents of the flask are likely to become solid before this point is reached, add 10 cubic centimeters more of sulphuric acid.) Complete the oxidation with a little permanganate of potash in the usual way, and proceed with the distillation as described in the Kjeldahl method. The reagents should be tested by blank experiments. THE RUFFLE METHOD, APPARATUS AND RB- \oi \ rs. (1) Standard solutions and indioator the same as tor the Kjeldahl method. (2) A mixture of equal parts by weight of tine soda-lime and finely powdered, crystallized sodium hyposulphite. (:j) A mixture of equal parts by weight of finely powdered granulated sugar and flowers of sulphur. (4) Granulated soda-lime, as described under the soda lime method. (5) Combustion tubes of hard Bohemian glass, 20 Inches long and j inch in diameter. ((>) Bulbed "f ' lubes or Wills's bulbs, as described under t he soda lime method. Described by Prof, M. A. 8< ovell. 67 PREPARATION. (1) Clean aud fill the 0" tube with 10 cubic centimeters of standard acid. (2) Fit cork and glass connecting tube. Fill the tube as follows : (1) A loosely fitting plug of asbestus, previously ignited, and then 1 to 1 J inches of the hyposulphite mixture. (2) The weighed portion of the substance to be analyzed is intimately mixed with from 5 to 10 grams of the sugar and sulphur mixture. (3) Pour on a piece of glazed paper, or porcelain mortar, a sufficient quantity of the hyposulphite mixture to fill about 10 inches of the tube ; then add the substance to be analyzed, as previously prepared; mix carefully and pour into the tube; shake down the contents of the tube; rinse off the paper or mortar with a small quantity of the hyposulphite mixture and pour into the tube; then fill up with soda-lime to within 2 inches of the end of the tube. (4) Place another plug of ignited asbestus at the end of the tube, and close with a cork. (5) Hold the tube in a horizontal position, aud tap on the table until there is a gas channel all along the top of the tube. Make connection with the U tube containing the acid, aspirate, and see that the apparatus is tight. The combustion.— Place the prepared combustion tube in the furnace, letting the open end project a little so as not to burn the cork. Com- mence by heating the soda-lime portion until it is brought to a full red heat. Then turn on slowly jet after jet toward the outer end of the tube, so that the bubbles come off two or three a second. When the whole tube is red hot and the evolution of the gas has ceased and the Liquid in the U tube begins to recede towards the furnace, attach the aspirator to the other limb of the U tube, break off the end of the tube, and draw a current of air through for a few minutes. Detach the U tube and wash the contents into a beaker or porcelain basin, add a few drops of the cochineal solution, and titrate. THE SODA-LIME METIIol*. (Not applicable in presence of nitrates.) ArPARATls AND BB-AGKNT& (1) Standard solutions and indicator the same as for tin* Kjeldahl method. (ii) Granulated soda-lime, line enongb to pass a 10 mesh seive, and thoroughly dry. (.») Pine soda lime, line enough ti) pass a 20-mesh seive, also thor OUghly dry. To prepare soda-lime of the required fineness, the coarse granulated article of the trade may be ground until it will pass through ;i seive of aboul .10 inch mesh. It is then sifted on a seive of about .20-inch mesh to separate tlie line from the coarse. The two portions are then 68 thoroughly dried ill the air-bath, or by heating in a porcelain dish or iron pan over a lamp, stirring constantly. Excellent soda-lime may be easily and cheaply prepared according to the directions of Professor Atwater (Am. Chem. Journal, vol. 9, p. 312), by slaking 2.! parts of quick-lime with a strong solution of 1 part commercial caustic soda (such soda as is used in the Kjeldahl process), care being taken that there is enough water in the solution to slake the lime. The mixture is then dried aud heated iu an iron pot to in- cipient fusion, aud when cold, ground and sifted as above. Instead of soda-lime, Johnson's mixture of carbonate of soda aud lime or slaked lime may be used. Slaked lime may be granulated by mixing it with a little water to form a thick mass, which is dried in the water-oven until hard and brittle. It is then ground and sifted as abore. Slaked lime is much easier to work witli than soda-lime and gives excellent results, though it is proba- ble that more of it should be used, in proportion to the substance to bo analyzed, than is the case with soda-lime. (4) Asbestus which has been ignited and kept in a glass-stoppered bottle. (5) Cumbustion tubes about 40 centimeters long aud about 12 milli- meters internal diameter, drawn out to a point and closed at one end. (0) Large bulbed U U" tubes with glass stop-cock, or Wills tubes with four bulbs. THE DETERMINATION. The substance to bo analyzed should be powdered fine enough to pass through a seive of 1 millimeter mesh ; 0.7 to 1.4 grams, according to the amount of nitrogen present, is taken for the determination. Into the closed end of the combustion tube put a small, loose plug of asbes- tos, and upon it about 4 centimeters of fine soda-lime. Iu a porcelain disli or mortar mix the substance to be analyzed, thoroughly but quickly, with enough line soda-lime to fill about 1G centimeters of the tube, or about 40 times as much soda-lime as substance, and put the mixture into the combustion tube as quickly as possible by means of a wide- necked funnel, rinsing out the dish and funnel with a little more line Soda lime, which is to be put in on top of the mixture. Fill the rest of the tube to about 5 centimeters of the end with a granulated lime. makingit as compact as possible by tapping the tube gently, while held iu a nearly upright positiou during the Ailing. The layer of gran- ulated soda-lime should be not Less than L2 centimeters Jong. Lastly, j, ul in ;i pin- of asbestus about L* eenl [meters long, pressed rather 1 ightly, and wipe OUl the end Of the tube to t'nn^ it from adhering soda limr. Conned the tube by means of a well lining rubber stopper or cork vvill, the '• , '" tube Or Wills bulbs containing L0 eubie cent im< i tors 69 standard acid, and adjust it in the combustion furnace so that the end projects about 4 centimeters from the furnace, supporting the " U "' tube or Wills tube suitably. Heat the portion of the tube containing the granulated soda lime to moderate redness, and when this is attained, extend the heat gradually through the portion containing the substance so as to keep up a moderate and regular flow of gases through the bulbs, maintaining the heat of the first part until the whole tube is healed uniformly to the same degree. Keep up the heat until gases have ceased bubbling through the acid in the bulbs and the mixture of sub- stance and soda-lime has become white or nearly so, which shows that tbe combustion is finished. The combustion should occupy about three- quarters of an hour, or not more than one hour. Remove the heat, and when the tube has cooled below redness, break off the closed tip and aspirate air slowly through the apparatus for two or three minutes to bring all the ammonia into the acid. Disconnect, wash tbe acid into a beaker or flask, and titrate with the standard alkali. During the combustion, the end of the tube projecting from the fur- nace must be kept heated sufficiently to x^revent the condensation of moisture, yet not enough to char the stopper, f he heat may be regu- lated by a shield of tiu slipped over the projecting end of the combustion tube. It is found very advantageous to attach a Bunseu valve to the exit tube, allowing the evolved gases to pass out freely, but preventing a violent "sucking back" in case of a sudden condensation of ste^m in the bulbs. METHOD FOR THE ANALYSIS OF CATTLE FOOD. Hygroscopic moisture.— Dry 2 to 3 grams at 100° C. to constant weight. Ash. — Char the substauce at a low red heat, exhaust the charred mass with water, burn the insoluble residue, add the ash to the residue from the evaporation of the aqueous extract obtained above, dry the whole at 110°, and weigh. Determine carbonic acid and insoluble matter (sand and charcoal) in the product for the estimation of the pure ash. Ether extract. — Pulverize the air-dry substance till all of it will pass through a sieve with less than one-sixtieth inch mesh ; dry about 5 grams at L00°, and take about 2 grama for each extraction. Exhaust with anhydrous ether of specific gravity .Tl'o to .725, not less than eight hours continuously. Dry ether extract at L00° in a current of dry hydrogen to a constant weight. Crude protein. — Determine nitrogen by the Kjeldahl method in the manner recommended by the nitrogen committee, and multiply the result by 6.25 for the crude protein. 70 Albuminoid nitrogen, Stuteer's method. — Prepare cupric hydrate as follows: Dissolve 100 grama of pure capric sulphate in 5 liters of water and add 2.5 cubic centimeters of glycerine; add dilute solution of sodium hydrate till the liquid is alkaline, filter, rub the precipitate up with water containing 5 cubic centimeters of glycerine per liter, and then wash by decantation or filtration till the washings are no longer alkaline. Then rub the precipitate up again in a mortar with water containing 10 per cent, of glycerine, thus preparing a uniform gelatin- ous mass that can be measured out with a pipette. Determine the quantity of cupric hydrate per 1 cubic centimeter of this mixture. To 1 gram of the substance, pulverized as for the determination of the ether extract, add 100 cubic centimeters of water in a beaker, heat to boiling, or in case of substances rich in starch heat on the water bath ten minutes, add a quantity of the cupric hydrate mixture contain- ing 0.7 to 0.8 gram of the hydrate, stir thoroughly, filter when cold, wash with cold water, and without drying put the filter and its con- tents into the concentrated sulphuric acid for the determination of the nitrogen after Kjeldahl. For filtering use either Schleicher and Shull's No. 589 filter paper or Swedish paper, either of which contains so little nitrogen that it can be left out of account. If the substance examined consists of seed of any kind or residues of seeds, such as oil-cake or anything rich in alkaline phosphate, add a few cubic centimeters of a concentrated solution of alum just before adding the cupric hydrate, and mix well by stirring. This serves to decompose the alkaline phosphate, precipitating aluminum phosphate; if this is not done cupric phosphate and pure alkali may be formed, and the protein copper may be partially dissolved in this alkaline liquid. ('rude fiber, Weende method. — Pulverize as for the ether extract, and extract the fat, at least nearly completely. To 2 grams of the substance in an Krlenmeyer flask add 200 cubic centimeters of boiling 1.25 per cent, sulphuric acid, boil immediately and for thirty minutes continu- ously with as much rotary motion as may be necessary to keep all the substances that may tend to adhere to the sides of the llask in contact with the liquid, throw on a filter of finest linen, and rinse the llask and wash the contents of the filter thoroughly with boiling water. Wash the contents of the filter back into the llask with 200 cubic centimeters of a L.25 per cent, solution of sodium hydrate, at once raise to boiling, and boil continuously for thirty minutes with rotary motion as above described, filter through a Q-OOCD crucible or according to some similar device, and wash very thoroughly with boiling water. Such thorough washing has been found to be essential to success. Wash then with alcohol and finally with ether, dry at 1 10 J , weigh, incinerate, and give the loss of weight for crude liber. 71 Statement of result*. Iu tlir SuWanct- in its natural condition (or as received). In the dry substance. Ether extract Albuminoid METHODS OF ANALYSES OF DAIRY PRODUCTS. BUTTER. MICROSCOPIC EXAMINATION. Place a small portion of the fresh sample, taken from the inside of the mass, on a slide, add a drop of pure sweet-oil, cover with gentle press- are, and examine with a one-half to one eighth inch objective for crys- tals of lard, etc. Examine same specimen with polarized light and selenite plate with- out the use of oil. Pare fresh butter will neither show crystals nor a parti-colored held with selenite. Other fats, melted and cooled and mixed with butter, will usually present crystals and variegated colors with the selenite plate. srECiric GRAVITY. Take a speci lie-gravity flask, not graduated, the stopper of which car- ries a delicate thermometer, whose bulb occupies sensibly the center of the flask. The thermometer should be first compared with a standard instrument. Clean the picnometer with water, alcohol, and ether; dry, be careful to remove all the ether vapor, and weigh. Fill with distilled water, insert Stopper, and place in vessel contain- ing water which comes nearly to top of picnometer (without stopper ; gradually raise the temperature of water in vessel containing llask until it is 10..V to U°0. The moment the thermometer of the picnometer marks i<> remove the llask, dry carefully, cover capillary safety-tube, and set in balance or desiccator to cnol. Weigh when temperature falls nearly to that of balance room. Having determined the tit re of the tl ask, it is to be care fully dried, filled with the melted and filtered fat at a temperature below 40°, and the weight of the fat determined ;it i<» '. as directed above. 72 MELTING POINT. Apparatus. — The apparatus consists of (1) an accurate thermometer, for reading easily tenths of a degree; (2) a less accurate thermometer for measuring the temperature of water in the large beaker glass; (3) a tall beaker glass, 35 centimeters high and 10 centimeters in diameter; (4) a tost tube 30 centimeters high and 3.5 centimeters in diameter; (5) a stand for supporting the apparatus; ((») some method of stirring the water in the beaker — for example, a blowing bulb of rubber and a bent glass tube extending to ne ir the bottom of the beaker; (7) a mixture of alcohol and water of the same specific gravity as the fat to be examined. Manipulation. — The disks of the fat are prepared as follows: The melted and filtered fat is allowed to fall from a dropping tube from a height of 15 to 20 centimeters on to a smooth piece of ice floating in water. The disks thus formed are from 1 to l.J centimeters in diameter and weigh about 200 milligrams. By pressing the ice under the water the disks are made to float on the surface, whence they are easily re- moved with a steel spatula. The mixture of alcohol and water is prepared by boiling distilled water and 05 per cent, alcohol for ten minutes to remove the gases which they may hold in solution. While still hot the water is poured into the test-tube already described until it is nearly half full. The test-tube is then filled with the hot alcohol. It should be poured in gently down the side of the inclined tube to avoid too much mixing. If the tube is not filled until the water has cooled the mixture will con- tain so many air bubbles as to be unfit for use. These bubbles will gather on the disk of fat as the temperature rises and finally force it to the top of the mixture. The test tube containing the alcohol and water is placed in a vessel containing cold water, and the whole cooled to below 1C° O. The disk of fat is dropped into the tube from the spatula, and at once sinks until it reache8 a part of the tube where the density of the alcohol-water is exactly equivalent to its own. Here it remains at rest and tree from the action of any force save that inherent in its own molecules. The delicate thermometer is placed in the test-tube and lowered until the bulb is just above the disk. In order to secure an even tempera tare in all parts of the alcohol mixture in the vicinity of the disk the thermometer LS moved from time to time in a circularly pendulous man- ner. A tube prepared in this way wiil be suitable for use for several days; in fact, until the air bubbles begin to attach themselves to the
  • U of fat. In no case did the two liquids become so thoroughly mixed as to lose the property of holding the disk at a fixed point, even when they were kept lor several weeks. In practice, owing to the absorption of air, il has been found neces sary to prepare n«\\ solutions every third or fourth day. The disk having been placed in position, the water in t he beaker-glass 73 is slowly heated and kept constantly stirred by means of the blowing apparatus already described. When the temperature of the alcohol- water mixture rises to about G° below the melting-point, the disk of fat begins to shrivel, and gradually rolls up into an irregular mass. The thermometer is now lowered until the fat particle is even with the center of the bulb. The bulb of the thermometer should be small, so as to indicate only the temperature of the mixture near the fat. .V gentle rotary movement should be given to the thermometer bulb, which might be done with a kind of clockwork. The rise of tempera- ture should be so regulated that the last 2° of increment require about ten minutes. The mass of fat gradually approaches the form of a sphere, and when it is sensibly so the reading of the thermometer is to be made. As soon as the temperature is taken the test-tube is re- moved from the bath and placed again in the cooler. A second tube, containing alcohol and water, is at once placed in the bath. It is not necessary to cool the water in the bath. The test-tube (ice water being used as a cooler) is of low enough temperature to cool the bath sufti- ciently. After the first determination, which should be only a trial, the temperature of the bath should be so regulated as to reach a maximum about l ..~> above the melting-point of the fat under examination. Working thus with two tubes about three determinations can be made in an hour. After the test- tube has been cooled the globule of fat is removed with a small spoon attached to a wire before another disk of fat is pat in. VOLATILE ACIDS. Take about 2.5 grams filtered fat in 200 cubic centimeters flask, to be used in subsequent distillation; add 2 cubic centimeters aqueous solu- tion potassium hydrate containing .5 gram KOH per cubic centimeter. Add 2 cubic centimeters 95 per cent, alcohol and heat on water bath, with occasional agitation. Saponification is complete in a few minutes Evaporate until alcohol is all driven off, using a current of air to re- move alcoholic vapor. The flask is fitted with a delivery tube and con denser. The delivery tube 18 carried up about 8 inches before it is bent to enter the condenser, and a bulb is blown in it just below the elbow. and this is tilled with broken glass or glass-wool. The fatty acids are then set free with 20 cubic centimeters of a solu- tion of phosphoric acid made by dissolving 200 grams glacial phos- phoric acid in water and making up to 1 liter. Enough water is added to make the volume of liquid in the llask TO cubic centimeters. Distillation is continued until the distillate measures 50 cubic centi- N meters, w Inch is titrated with NaOU. using phenol-phthalciu as indi- cator. (('are should be taken to have the potash used free from carbonate. 74 A potash containing carbonate saponifies a fat with great difficulty. In an instance where the saponification was taking place very slowly, the potash was found to contain 0.80 per cent. CO., equal to 21.17 per cent. K 2 C0 3 .) SOLUBLE ACIO-. The washings obtained in the process of estimating the insoluble acids, the description of which follows, are made np to 1 liter and an aliquot part taken for titration with one-tenth X alkali, using phenol phthalein as indicator. METIIOD FOR THE DETERMINATION OF SOLUBLE AND INSOLUBLE ACIDS. HE- AGENTS. 1. A standard semi-normal hydrochloric-acid solution, accurately prepared. 2. A standard decinormal soda solution, accurately prepared; each 1 cubic centimeter, contains .0040 gram of NaOH, and neutralizes .0088 gram of butyric acid, C 4 H a 2 . 3. An approximately semi-normal alcoholic potash. Dissolve 40 grams of good stick potash in 1 liter of 95 per ceut. alcohol, redistilled. The solution must be clear and the KOtI free from carbonates. 4. A 1 per cent solution of phenol-pthalein in 95 per cent, alcohol. Saponification is carried out in rubber-stoppered beer-bottles holding about 250 cubic centimeters. About 5 grains of the melted butter fat, filtered and treed from water and salt, arc weighed out by means of a small pipette and beaker, which are rcweighed alter the sample has been taken out and run into a saponification bottle; 5 ) cubic centimeters of the semi normal potash is added, the bottle closed and placed on the steam bath until the con- tents arc entirely saponified, facilitating the operation by occasional agitation. The alcoholic; potash is measured always in the same pipe! tc and uniformity further insured by always allowing it to drain the same length of time, viz, thirty seconds. Two or three blanks arc also meas- ured out at the same time and treated in (lie same way. In from five to thirty minutes, according to the nature of the f.it. the liquid will appear perfectly homogeneous, and when this is the case the saponification is complete, and the battle may be removed and cooled. When sufficiently cool the stopper is removed and the con tents of the llask rinsed with a little 95 per cent, alcohol into an Krlcn- mcaT flank of about 2 j0 cubic cen ti meters capacity, which is placed on the steam bath, together wi h the blanks, until the alcohol has evap- orated. Titrate the blanks with semi-normal ilCl. using phenol-phthaloin as an indicator Then run into each of the ll,i>ks containing the fat acids l cubic centimeter more semi-normal BOl than is required to neutral- 75 ize the potash in the blanks. The flask is then connected with a con- densing tube 3 feet long, made of small glass tubing, and placed on the steam bath until the separated fatty acids form a clear stratum on the surface of the liquid. The flask and contents are then allowed to become thoroughly cold, ice water being used for cooling. The fatty acids having quite solidified, the contents of the flask are filtered through a dry filter paper into a liter flask, care being taken not to break the cake. Two hundred to three hundred cubic centime- ters of hot water is next poured on the contents of the flask, the cork with its condenser tube re-inserted, and heated on the steam bath until the cake of acids is thoroughly melted, the flask being occasionally agitated with a circular motion, so that none of its contents are brought on the cork. When the fatty acids have again separated as an oily layer the flask and its contents are cooled in ice water, and the liquid filtered through the same filter into the same liter flask. This treat- ment with hot water, followed by cooling and filtration of the wash water, is repeated three times, the washings being added to the first filtrate. The mixed washings and filtrate are next made up to 1 liter, and 100 cubic centimeters, in duplicate, are taken and titrated with decinormal NaOH. The volume required is calculated to the whole liquid. The number so obtained represents the measure of decinormal NaOH neutralized by the soluble fatty acids of the butter fat taken, plus that corresponding to the excess of the standard acid used, viz, 1 cubic centimeter. The amount of soda employed tor the neutraliza- tion is to be diminished, for the 1 liter, by 5 cubic centimeters, corre- sponding to the excess of 1 cubic centimeter J N. acid. This corrected volume, multiplied by the factor 0.0088, gives the bu- tyric acid in the weight of butter fat employed. (See table.) The flask containing the cake of insoluble fat acids is inverted and allowed to draiu and dry for twelve hours, together with the filter paper through which its soluble fatty acids have been filtered. When dry the cake is broken up and transferred to a weighed glass evaporating dish. Remove from the dried filter paper as much of the adhering fat acids as possible and add them to the contents of the dish. The funnel, with the filter paper, is then placed in an Erlenmeyer flask, a hole is made in the bottom of the filter paper, and it is thoroughly washed with absolute alcohol from a wash-bottle. The flask is rinsed with the wash- ings from the filter paperaod pure alcohol, and these transferred to the evaporating dish. The dish is placed on the Bteam bath and the alcohol driven off. It is then transferred to the air bath and dried at 100° C. for two hours, taken out. cooled in a desiccator, and weighed. It is then again placed in the air bath and dried for another two hours, cooled as before, and weighed. If there is no considerable decrease in weight the first weight will do; otherwise, reheat two hours and weigh. This gives the weight Of insoluble fat acids in khe quantity taken, from which the percentage is easily calculated. 76 Table for the calculation of soluble fatty acids. No. cc. KOII sol. Equiv. 10 11 ] 2 Grams. .088 .0968 . 1056 .1144 . 1232 . 1320 . 1408 .1496 .1584 . 1672 . 1760 .1848 .1936 . 2024 .2112 . 2200 . 2288 . 2376 .2464 . 2...V2 .2640 13 14 15 16 17 18 19 20 L' 1 22 23 24 25 •jti 27 28 '-"J 30 N 10 XaOlI. Equiv. Grams. . 0902 .0090 .1078 . 1168 . 1254 . 1342 .1430 . 1518 .1606 .1694 . 1782 .1870 . 1958 .2010 .2134 . 2222 .2310 . 2398 .2486 .2571 .2662 X 10 xaon. Equiv. + .50 50 50 50 I 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 Grams. . 0924 .1012 .1100 .1188 . 1276 .1364 . 1452 .1540 . 1628 .1716 .1804 .1892 .1980 .2068 . 2156 .2244 .2332 . 2420 .2508 . 2596 . 2684 N 10 XaOlI. +.75 Equh Grams. .9946 .1034 .1122 . 1210 . 1298 .1386 .1474 . 1562 . 1650 .1738 .1826 . 1D14 . 2002 . 2090 .2178 . 2266 . 2354 .2442 .2536 .2618 .2706 The table gives the weight of soluble fatty acids (butyric, etc.) for each quarter of a cubic centimeter of deci normal alkali from 10 to 30. Example: Weight fal taken grams.. 4.967 N No. cubic centimeters 1(J alkal: used 25. 50 N cubic centimeters due to 1 cubic centimeter k acid 20.50 Weight s oluble 1'at acids *.. 1804 Per cent, soluble, fat acids :>.(*>:? SAPONIFICATION EQUIVALENT. About 2.5 grams butter fat (filtered and free from water) are weighed into a patent rubber stoppered bottle, and 25 cubic centimeters approxi- mately semi-normal alcoholic potasb added. The exact amount taken is determined by weighing a small pipette with the beaker of tat, run- ning the fat into the bottle from the pipette and weighing beaker and pipette again. The alcoholic potash is measured always in the same pipette, and uniformity further insured by always allowing il to drain the same length of time (thirty seconds). The bottle is then placed in the steam bath, together with a blank, containing no fat After sapon- ification Is complete and the bottles cooled, the contents are titrated with accurately semi-normal hydrochloric acid, using phenol phthalein as an Indicator. The number of cubic centimeters of tin 4 acid used for the sample deducted from the number required for the blank gives the number of cubic centimeters which combines with the fat, and the sa- ponification equivalent is calculated by the following formula, in which \Y equals the weight of fat taken in milligrams and N the number of cubic centimeters which have combined with the fat. 2 W Sap. Equiv. =-^ . 77 If it is desirable to express the number of milligrams of potash for each gram of fat employed it can be clone by dividing 5,G10 by the saponification equivalent and multiplying the quotient by ten. WATER. Place about 2 grams in flat-bottomed platinum dish two-thirds full of clean, dry sand and heat for two hours in air bath at 105° C. (Eor alternate method of estimation of water, see page SI.) ESTIMATION OF SALT. Volumetric method. — The amount of the butter or butter substitute to be taken is ."3-10 grams; weigh in a counterpoised beaker-glass. The butter (fresh from the refrigerator) is placed in portions of about 1 gram at a time in the beaker, these portions being taken from different parts of the sample. By this means a reasonably fair sample of the whole is obtained. The given quantity having been weighed out, it is removed from the pan. Hot water is now added (about 20 cubic centimeters) to the beaker, containing the butter, and after it has melted the liquid is poured into the bulb of the separating apparatus. The stopper is now inserted and the contents shaken for a few moments. After standing until the fat has all collected on top of the wafer, the stop-cock is opened and the water, containing most of the salt, is allowed to run into an Erlenmeyer flask, being careful to let none of the fat globules pass. Hot water is again added to the beaker and thence poured into the separator? apparatus, the bottle well shaken, and the foregoing process is repeated ten to fifteen times, using each time L0to20 cubic centime- ters of water. The resulting washings contain all but a mere trace of the NaCl, originally present in the butter. Estimation of NaCl infiltrate. — The chloride of sodium is now deter- mined in the filtrate by a standard solution of AgXO . using a few drops of a saturated solution of potassium chromate as indicator. TOTAL MINERAL MATTER, (ASH) AND CURD. The methods of estimating curd depend on the principle of first diving a weighed portion of the butter, and afterwards extracting the tat with ether or petroleum. The residual mass is then weighed and the (and determined by loss on ignition. This process is carried on as follows: Five to ten grams of butter are dried, at 1(MP (\, for a few hours in a porcelain dish. The dried fit . v^c, is filtered through a Gooch cruci- ble, the contents of the dish all brought into the crucible and well washed with ether or light petroleum. The filter crncibleis dried for two hours in air bath and weighed. The (ami is then determined by loss of weight <>n ignition. 78 The total mineral matter is represented by the residue left after [g. nition. Curd and insoluble substances may be estimated somewhat more ac- curately as follows : Place the butter (5 to 10 grains) in a Gooch crucible; set the crucible on a pad of blotting-paper in an air bath (100° C.) for an hour, wash two or three rimes with ether or light petroleum. Dissolve the salt with hot water, dry at 100° C. and weigh. Any insoluble mineral matter and the ash of the curd is weighed with it in this method. CASEIN. Method of Babcock. — Ten grams of the dried butter are treated with light petroleum until all fat is removed. The residue is then ignited with soda-lime or treated by the Kjeldahl method. METHODS FOR MILK ANALYSIS. ESTIMATION <>K WATER. From a weighing bottle take 5 cubic centimeters milk, put in a weighed thin glass dish or dish lined with tinfoil one-third full of pow- dered asbestos ; dry for two hours at 100° C. The temperature obtained in a boiling-water bath does not reach 100° O. The milk should be dried in an air bath, the temperature of which is carefully controlled. ALTERNATE METHOD OF ESTIMATING WATER. Evaporate 1 to L* grams of milk in shallow watch glass or platinum dish on the water bath for thirty minutes. Dry for an hour at 100° C. and weigh. ESTIMATION OF CASEIN. Take 5 grams of milk, digest in Kjeldahl apparatus with lit) cubic centimeters n 2 S0 4 , and estimate ammonia in the usual way. ALTERNATE METHOD OF KSTIMATION OF CASEIN. Bab up >'i a mortar the thin disk containing the dried residue from the above process or remove the foil containing it and transfer to soda- lime combustion tube in the usual way. The mortar and pest 1 !' must bo well cleaned with the soda-lime and these cleanings placed in the tube. Or the dish or tinfoil and its contents may be transferred to a diges- tion flask and the casein estimated by the method of Kjeldahl. I 5 MM \TK.\ OF i \ I. Method of Adams. — The kind of paper and t lie method of using it first proposed by Adams are as follows: Aa for material, the only extra article is some stout white blotting-paper, known in the n;nk .-is "whin- cjemj Moiling mill r>," weighing 38 ponntU per ream, This 79 should be in unfolded sheets, machine-cut into strips 2$ inches wide and 22 inches long; each sheet in this manner cuts into seven strips. I have tried other papers, but none have answered so well as this; it is very porous and just thick enough. Each of these strips is carefully relied into a helical coil, for which purpose I use a little machine, made by myself, consisting of a stout double wire, cranked twice at right angles, and mounted in a simple frame. One end of the strip being thrust between the two wires, the handle is turned and the coil made with great facility. This may be done, for the nonce, on a glass rod, the size of a cedar pencil. Two points have to be carefully attended to; the paper must not be broken, and the coil must be somewhat loose, the finished diameter being a little under an inch. I am in the habit of rolling up a considerable number at a time and placing each within a brass ring as it is rolled, inscribing on one corner with a lead- pencil its own proper number. These coils are next thoroughly dried, and I need hardly say the accuracy of the process depends upon this drying. This can be satisfactorily done in an ordinary air bath at 100° C, providing the bath be heated properly and the paper kept in it long enough. I found the common way of heating the thin bottom of the bath with a single jet not to answer. My bath is placed upon a stout iron surface, which is heated by a large ring of jets ; in this way the heat is evenly distributed over the whole of the bottom of the bath, aud the papers, which are put in a cage frame of tinned-iron win- 5 by 2^ inches and divided into eight partitions, get evenly aud completely dried, if allowed to remain in the bath all night, aud weighed in a weighing tube next morning, and their weights having been registered according to their numbers, stored away ready for use, as follows: The milk to be examined is shaken, and with a pipette 5 cubic centimeters are dis- charged into a small beaker 2 inches high by 1£ iuches in diameter, of a capacity of about 30 cubic centimeters, weighing about 12 grams. This charged beaker is first weighed and then a paper coil gently thrust into the milk very nearly to the bottom. In a few minutes the paper sucks up nearly the whole of the milk. The paper is then carefully withdrawn by the dry extremity of the coil aud gentiy reversed and stood, dry end downwards, on a clean sheet of glass. With a little dexterity all but the la-t fracl ion of a drop can be removed from the beaker and got on the paper. The beaker is again weighed and the milk taken got by difference. It is of importance to take up the whole of the milk from the beaker, as I am disposed to consider the paper has a selective action, removing the watery constituents of the milk by preference over the fat. The charged paper is next placed in the water oven on the glass plate, milk-end upwards, aud rough-dried. Mismanagement may possibly cause a drop to pass down through the coil onto the glass. This accident ought never to occur ; but if it does it is revealed in a moment by inspect ion of the surface of the glass, and the experiment is thereby lost. In about an hour it is rough-dried and in a suitable condition for the extract ion of the fat. The method of Adams has been thoroughly tried by the English chem- ists and lias received the approval of the English Society of Public Analysis. It gives uniformly about .2 per cent, more fat in normal milk than the ordinary gravimetric methods. The following modification of the process may be nsed: The blotting-paper is replaced by thick altering paper cut into strips 2 feel long and 2.5 inches wide. These are thoroughly extracted by ether or petroleum Or alco- hol. One end of the atrip of paper being held horizontally by a clamp or by an I ant, 5 cubic centimeters milk is run out by a pipette from a weighing bottle along the middle of the stripof altering paper, being careful not to let the milk gel too near 80 the ends of the paper aud to secure an even distribution of it over the whole length of the slip. The pipette is replaced in the weighing bottle and the whole re weighed, and thus the quantity of milk taken is aocorately determined. The strip of paper is now hung up over a sand bath in an inclosed space high enough to receive it where the air has a temperature of 100 c C. (circa). In two or three minutes the paper is thoroughly dry. It is at once, while still hot, rolled into a (oil and placed, before cooling, in the extraction apparatus already described. The fat is dissolved by ether or petroleum, collected in a weighed flask, and, after thorough drying, weighed. The fat after extraction may also be estimated volumetrically. as de- scribed in the method of Morse. From data which have been collected, it appears that the estimation of the fat in milk by the lactocrite is strictly comparable with the re- salts of the Adams method. Those who have this instrument, therefore, can use it instead of the method given. ALTERNATE METHOD OF ESTIMATING THE FAT IX MILK. Method of Morse, Piggot, and Burton. — This method consists in the de- hydration of the milk by means of anhydrous sulphate of copper; the extraction of the fat by meausof the low-boiling products of petroleum ; the saponification of the butter by means of an excess of a standard so- lution of potassium hydroxide in alcohol: and the determination of the excess of the alkali by means of a solution of hydrochloric acid. The following apparatus and re-agents are required : (1) A porcelain mortar and pestle. (2) An extraction tube, 14 or 15 millimeters in diameter, 220 milli- meters in length, with funnel-shaped top. A straight chloride of cal- cium tube may be used for this. (3) A 200 cubic centimeter Krlenmeyer flask, strong euoogh to be used with a filter pomp. (1) A suitable stand for holding the flask and extraction tube. (5) Ten cnbic centimeter pipettes. (G) Weighing glasses with ground glass stoppers. (7) A low boiling gasoline, distilling between 30° and 60° 0. (8) Dehydrated sulphate of copper. (9) Semi-normal solution of potash in 95 per cent, alcohol. (10) A semi-normal solution of hydrochloric acid. Manipulation. — Place about 20 grains of the anhydrous copper Sul- phate, roughly measured in Q copper spoon of the size to hold about thai amount, in a porcelain mortar; make a cavity in the center < f the mass with the DCS tie. Allow 10 cubic centimeters of the milk to nm on to the copper Sulphate, being careful that none of it touches the sides of the mortal'. When the milk is nearly dry, grind the mass up with a Little clean sum I. transfer t<> the extraction tube, gently pressing it down in the tube by means of H glass rod. The lower portion of the c\trac (ion tube to be packed with clean cotton wool. The fit is extracted in the following way : (.5 cubic centimeters of benzine are p< mrcd over the 81 material in the extraction tube and drawn down, with the aid of the filter pump, until the whole of the mass to be extracted has become wet with the liquid, when the connection with the pump is closed; after about five minutes another portion of 15 cubic centimeters of benzine is poured into the tube and the whole of the liquid slowly drawn through with aid of the pump into the flask. Usually one extraction of this kind is sufficient to withdraw the whole of the butter, but for the sake of greater accuracy the process may be repeated two or three times. Titration. — The benzine may be evaporated and the residual butter fat saponified with about 25 cubic centimeters of the approximately semi-normal potash. The residual alkali is determined by means of the semi-normal hydrochloric acid, using phenol-phthalein as indicator. The difference between the amount required in this process and the f mount necessary to neutralize the quantity of alkali taken gives the amount of alkali required for the saponification. The number of milli- grams of potash required for one gram of the fat is taken at 230. The fat may also be accurately titrated without evaporating the benzine. ALTERNATE METHOD OF ESTIMATING WATER AND FAT IX MILK. Method ofBabcoclc. — In the bottom of a perforated test-tube is placed a clump of cleau cotton; the tube is then filled three-quarters full of ignited asbestos, lightly packed, and a plug of cotton inserted over it. The tube and contents are weighed and the plug of cotton carefully re- moved and 5 grams of milk from a weighed pipette run into it, and the ping of cotton replaced. The tube connected at its lower end by a rub- ber tube and adapter with a filter pump is placed in a drying oven of a 100° C, and a slow current of dry air drawn through it until the water is completely expelled, which in no case requires more than two hours. The tube containing the solids from the above operation is placed in an extraction apparatus and exhausted with ether in the usual way. ALTKKXATK METHOD OF ESTIMATING WATER AND FAT. Method of Professor Macfarlane. — A glass tube 4 to 5 cubic centimeters in length and 2 centimeters in diameter, open at one end, drawn out to a tube 5 millimeters in diameter at the other end, is two-thirds filled with asbestos fiber, such as is used in manufacturing packing. It is dried in the water bath for several hours, cooled in the desiccator, and weighed. Ten cubic centimeters of the milk is then added from a pipette, which is completely absorbed by the asbestos. It is (hen weighed, the addi- tional weight of the milk representing the amount taken. The tube, along with many others, is placed in a water bath with constant level and dried for ten or twelve hours (during the night) at a temperature of 90°. Next morning the tubes are cooled in the desiccator and weighed, the loss in weight being the moisture. The tubes are then placed in the Soxhlet extraction apparatus and exhausted with petroleum ether for 7717— No. 1!) 6 82 four boors. They are then removed and dried in a steam bath, cooled in desiccator, and weighed. The loss represents the butter fat THE ESTIMATION OF SUGAR. The re-agents, apparatus, and manipulation necessary to give the most reliable results in milk sugar estimation are as follows : Re-agents. — (1) Basic plumbic acetate, specific gravity 1.97. Boil a sat- urated solution of sugar of lead with an excess of litharge, and make it of the strength indicated above. One cubic centimeter of this will pre- cipitate the albumens in 50 to GO cubic centimeters of milk. (2) Acid mercuric nitrate. Dissolve mercury in double its weight of nitric acid, specific gravity 1.42. Add to the solution an equal volume of water. One cubic centimeter of this re-agent is sufficient for the quantity of milk mentioned above. Larger quantities can be used without affecting the results of polarization. (.*>) Mercuric iodide with acetic acid. KI 33.2 grams, HgCl 2 13.5 grams, II, C2II3O 20 cubic centimeters, H t O 64 cubic centimeters. Apparatus. — (1) Pipettes marked at 59.5, CO, and (30.5 cubic centime- ters. (2) Sugar flasks marked at 102.4 cubic centimeters. (3) Filters, observation tubes, and polariscope. (4) Specific gravity spindle and cylinder. (5) Thermometers. Manipulation. — (1) The room and milk should be kept at a constant temperature. It is not important that the temperature should be any given degree. The work can be carried on equally well at 15° C, 20° C, or 25° C. The slight variations in rotary power within the above limits will not affect the result for analytical purposes. The tempera- ture selected should be the one which is most easily kept constant. (2) The specific gravity of the milk is determined. For general work this is done by a delicate specific gravity spindle. Where greater ac- curacy is required, use specific-gravity flask. (3) If the specific gravity be 1.020, or nearly so, measure out 00.5 cubic centimeters into the sugar-flask. Add 1 cubic centimeter of mer- curic nitrate solution, or 30 cubic centimeters mercuric iodide solution, and till to 102.4 cubic-centimeter mark. The precipitated albumen oc- cupies a volume of about 2. II cubic centimeters. Hence the milk solu- tion is really 100 cubic centimeters. If the specific gravity is L.030, use GO cubic centimeters of milk, [f specific gravity is 1.034, use 59.5 cubic centimeters of milk. (4) Fill up to mark in 102.1 cubic cent imeter tlask, shake well, filter, and polarize. NOTES. — In the above method of analysis the specific rotatory power of milk sngar is taken ;it 52.5, and the weigh! of it in LOO cubic cent] meter solution to read 100 degrees in the cane-sugar scale at 20.56 grams. This is for instruments requiring 10. l!) grams sucrose to produce a 10 tut ion of loo Bugar degrees, it will be easj to calculate the number for milk-sugar, whatever instrument is employed. 83 Since the quality of milk taken is three times L'0.56 grams, the polar- iscopie readings divided by 3 give at once the percentage of milk sugar when a 200-inillimeter tube is used. If a 400-millimeter tube is employed, divide reading by G; if a 500- millimeter tube is used, divide by 7.5. Since it requires but little more time, it is advisable to make the analysis in duplicate and take four readings for each tube. By follow* ing this method gross errors of observation are detected and avoided. By using a flask graduated at 102.4 for GO cubic centimeters no cor- rection for volume of precipitated caseine need be made. In no case is it necessary to heat the sample before polarizing. In the above method no account is taken of the fat which is retained on the filter with the caseine. It is worth while to inquire if a correc- tion similar to that made for the albuminoids should not also be made for the fat ? (See page 10, Vieth's Investigation of this subject from Analyst 13, 63.) ESTIMATION OF ASH. Evaporate to dryness in a weighed platinum dish 20 cubic centimeters of milk from a weighing-bottle, to which G cubic centimeters of UNO; has been added, and burn in muffle at low red heat until ash is free from carbon. METHODS FOR THE ANALYSIS OF FERMENTED LIQTORS. A com mission of experts, appointed in the year 1884 by the chancellor of the Empire, to which was intrusted the establishment of uniform methods for the chemical investigation of wine, adopted the following resolutions, which were made public by the Prussian minister for com- merce and trade by a decree of the 12th August, 1884, which provides that they shall be rigidly adhered to in public institutions for the ex- amination of food-stuffs, and are recommended to the representatives of like private concerns : RESOLUTIONS or Till: COMMISSION FOR ESTABLISHING UNIFORM METHODS FOR Till: ANALYSIS or WINES ' since, in consequence of improper manner of taking, keeping, and Bending in of samples of n ine for investigal ion by the authorities, a decomposition or change in the latter ofteu occurs, the commission considers it advisable to give the following in- structions : INSTRUCTIONS FOB SAMPLING, PRESERVING, AND SENDING IN "i SAMPLES O] WINE I OB i:\.\mi\ \ i [OH HY nn: \i'i BOB! i K8. (1) Of each simple, at least one bottle (| liter), M well tilled as possible, must l.v taken. (-2) The bottles and corks need musl be perfectly clean; the best ate new bottles and corks. Pitchers or opaqne bottles in which the presence of impurities can not be seen ale not to lie iise.l. Pas Geset? betreffenq* den Verkehr nut Nahrungsniittel, u. s. w., p. 184, 84 (3) Each bottle shall be provided with a label, gammed (not tied) on, upon which shall be given the index number of the sample corresponding to a description of it. (4) The samples are to be sent to the chemical laboratory as soon as possible to avoid any chance of alteration which, uudsr some circumstances, can take place in a short time. If they are, for some special reason retailed in any other place for any length of time, the bottles are to be placed in a cellar and kept lying on their .sides. •(5) Jf in samples of wine taken from any business concern adulteration is shown, a bottle of the water is to be taken which was presumably used in the adulteration. (G) It is advisable, in many cases necessary, that, together with the wine, a copy of these resolutions be sent to the chemist. A. — Analytical methods. itlc gravity. — In this determination use is to be made of a pionometer, or a Westphal balance controlled by a picnometer. Temperature, 15° C. Alcohol. — The alcohol is estimated in 50 to 100 cubic centimeters of the wine by the distillation method. The amount of alcohol is to be given in the following way : In 100 cubic centimeters wine at 15° C. are contained n grams alcohol. For the calcula- tion the tables of Baumhauer or Hebnerare used. (The amounts of all the other constituents are also to be given in this way : in lid cubic centimeters wine at 15° C. are contained n grams.) Extract. — For this estimation 50 cubic centimeters of wine, measured out at 1."' ( '.. are evaporated on the water bath in a platinum dish (85 millimeters in diameter, 20 millimeters in height, and 75 cubic centimeters capacity, weight about "JO grains), and the residue heated for two and one-half hours in a water jacket. Of wines rich in sugar (that is, wines containing over 0.5 grams of sugar in 100 cubic centimeters) a smaller quantity, with corresponding dilution, is taken, so that 1 or at the mosl 1.5 grams extract are weighed. (ilijceriiic. —Oiiu hundred cubic centimeters of wine (for sweet wines see below) are evaporated in a roomy, not too shallow, porcelaiu dish to about 10 cubic centimeters, a little saud added, aud milk of lime to a Strong alkaline react ion, and the whole brought nearly to dryness. The residue is extracted with 50 cubic centimeters of 9C percent, alcohol on the water bath, with frequent stirring. The solution is poured off through a filter, and the residue exhausted by treatment with small quantities of alcohol. For this 50 to 100 cubic centimeters are generally sufficient, so that the en tin: filtrate measures 100 to 200 cubic cenl imeters. The alcoholic solul ion is evapo- rated on the water bath to a sirupy consistence. (The principal part of the alcohol may be distilled oil' if desired.) The residue is taken np by 10 cubic centimeters of absolute alcohol, mixed in a stoppered flask with 1"> cubic centimeters of ether and allowed to stand nut il clear, when t he (dear Liquid is poured off into ;. glass-stoppered weighing-glass, filtering the last portions of the solution. The solution is then evap- orated iu tin; weighing-glass until the residue do longer flows readily, after which it is dried an hour longer in a water jacket. After cooling it is weighed. In the case of sweet wines (over <>.:, grams sugar in 100 oubic centimeters) 50 oubic centimeters are taken in a good-sized llask, some sand added, ami a sufficient quan- tity of powdered slack lime, and heated wit h frequent shaking in t he water ba I h. Alter cooling, inn cubic centimeters of 96 per cent, alcohol are added, the precipitate which forms allowed to separate, the solution Altered, and the residue washed with alcohol of the same strength. The alcoholic solution is evaporated and the residue t idled as above. Free aeids (total quantity of tin' acid reacting constituents of the wine).— These are to he estimated with a sutliciently dilute normal solution of alkali (atleasl one-third normal alkali) in 10 t" 20 cubic centimeters wine. ]| one-tenth normal alkali is used ,,{ least 1" OUbic centimeters Of wine should he taken for titration ; it one-third nor- ■in cubic centimeters of win.'. The drop method ( Zfc»/< I method*), with delicate 85 re-agent paper, is recommended for the establishment of the neutral point. Any con- siderable quantities of carbonic acid in the wine are to be previously removed by skaking. These "free acids" are to be reckoned and reported as tartaric acid (C 4 H 6 6 ). Volatile acids. — These are to be estimated by 'distillation in a current of steam, and not indirectly, and reported as acetic acid (C2H4O2). The amount of the "fixed acids" is found by subtracting from the amount of "free acids" found, the amount of tartaric acid corresponding to the "volatile acids" found. Bitartrale of potash and free tartaric acid. — (a) Qualitative detection of free tartaric acid: 20 to 30 cubic centimeters of the wine are treated with precipitated and finely- powdered bitartrate of potash, shaken repeatedly, filtered off after an hour, and 2 to 3 drops of a 20 per cent, solution of acetate of potash added to the clear filtrate, and the solution allowed to stand twelve hours. The shaking and standing of the solu- tion must take place at as nearly as possible the same temperature. If any consider- able precipitato forms during this time free tartaric acid is present, and the estima- tion of it and of the bitartrate of potash may be necessary. (&) Quantitative estimation of the bitartrate of x>otash and free tartaric acid : In two stoppered flasks two samples of 20 cubic centimeters of wine each are treated with 200 cubic centimeters ether-alcohol (equal volumes), after adding to the one flask 2 to 3 drops of a 20 per cent, solution of acetate of potash. The mixtures are well shaken, and allowed to stand sixteen to eighteen hours at a low temperature (0 — 10° C), the precipitato filtered off, washed with ether-alcohol, and titrated. (The solution of acetate of potash must be neutral or acid. The addition of too much acetate may cause the retention of some bitartrate in solution.) It is best on the score of safety to add to the filtrate from the estimation of tho total tartaric acid a further portion of 2 drops of acetate of potash, to see if a further precipitation takes place. In special cases the following procedure of Nessler and Barth may be used as a con- trol : Fifty cubic centimeters of wine are evaporated to the consistency of a thin sirup (best with the addition of quartz sand), tho residue brought into a flask by means of small washings of 9o per cent, alcohol, and with continual shaking more alcohol is gradually added, uutil the entire quantity of alcohol is about 100 cubic centimeters. The flask and contents are corked and allowed to stand four hours in a cool place, then filtered, and tho precipitate washed with 90 per cent, alcohol ; tho filter paper, together with the partly flocculent, partly crystalline precipitate, is returned to tne flask, treated with 30 cubic centimeters warm water, titrated after cooling, and the acidity reckoned as bitartrate. Tho result is sometimes too high if pectinous bodies separate out in small lumps, inclosing a small portion of free acids. In the alcoholic filtrate the alcohol is evaporated, 0.5 cubic centimeters of a 20 per cent, potassic acetate solution added, which has been acidified by a slight ex- 1 acetic acid, and thus tin- formation of bitartrate from the free tartaric acid in the wine facilitated. The whole is now, like the first residue of evaporation, treated with (sand and) 96 per cent, alcohol, and carefully brought Lnto a flask, the volume of al- cohol increased to 100 cubic centimeters, well shaken, corked, allowed to stand in a cold place four hours, filtered, the precipitate washed, dissolved in warm water, ti- trated, and for one equivalent of alkili two equivalents of tartaric acid are reckoned. This method for the estimation of the free tartaric acid has the advantage over the former of being free from all errors of estimation by difference. The presence of con- siderable quantities of sulphates impairs the accuracy of the method. Malic acid, §uceini poured Off or filtered into a Mask, and the alcohol distilled or evaporated off down toabonl 5 cubic cent [meters. The residue is I rea ted with aboul 15 cubic cen- timeters water and a little I tone- Mack, filtered into a graduated cylinder, and w ashed with water until the lilt rate measures :'•" en i »ic cent i nut eis. if this shows on polari- zation a rotation of more than -f<».."> . Wild, the wine contains the iintermeulahle matter of commercial potato sugar (amylin). It' in the estimation of the sugar by 1'eh line's solution moil- than agar iii inn en hie centimeters was found, the original righl rotation caused i».\ the amylin may be diminished i».\ the left-rotating sugar; the above precipitation with alcohol is in this case to be undertaken, even 87 when the right-rotation is less than 0.3°, Wild. The sugar is, however, first fer- mented by the addition of pure yeast. With very considerable content in (Fehling's solution) reducing sugar and proportionally small left-rotation, the diminishing of the left-rotation inav be brought about by cane sugar or dextrin or amylin. For the detection of the first the wine is inverted by heating with hydrochloric acid (to 50 cubic centimeters wine, 5 cubic centimeters dilute hydrochloric acid of speciiic grav- ity 1.10), and again polarized. If the left-rotation has increased, the presence of cane sugar is demonstrated. The presence of dextrin is shown as given in the section on "gum." In case cane sugar is present well washed yeast, as pure as possible, should be added, and the wine polarized after fermentation is complete. The con- clusions are then the same as with the wines poor in sugar. For polarization only large, exact instruments are to be used. The rotation is to be calculated in degrees, Wild, according to Landolt (Zeitschr. f. analyt. Chemie, 7. 9): lo Wild = 4.G043 C Soleil. l°Soleil =0.217189° Wild. 1° Wild = 2.89005 c Ventzke. 1° Ventzke = 0.346015° Wild. Gum (arabic). — For establishing the addition of any considerable quantities of gum 4 cubic centimeters wine are treated with 10 cubic centimeters of 90 percent, alcohol. If gum is present, the mixture becomes milky, and only clears up again after several hours. The precipitate which occurs adheres partly to the sides of the tube, and forms hard lumps. In genuine wine. Hakes appear after a short time, which soon set- tle, and remain somewhat loose. For a more exact test it is recommended to evapo- rate the wine to the consistency of a sirup, extract with alcohol, of the strength given above, and dissolve the insoluble residue in water. This solution is treated with some hydrochloric acid (of specific gravity 1.10) heated under pressure two hours, and the reducing power ascertained with Fehling's solution, and calculated to dextrose. In genuine wines no considerable reduction is obtained in this way. (Dextrin is to be detected in the same way.) Mannitc. — As the presence of maunite in wines has been observed in a lew eases, it should be considered wben pointed crystals make their appearance in the extract or the glycerine. Nitrogen. — In the estimation of nitrogen the soda-lime method is to be used. Mineral m a fit rx. — For their estimation 50 cubic cent [meters of wine are used. Jf the incineration is incomplete, the charcoal is leached with some water, and burned by itself. The solution is evaporated in the same dish, ami the entire ash gently i united. Chlorine estimation. — The wine is saturated with sodic carbonate, evaporated, the residue gently ignited and exhausted with water. In this solution the chlorine is to be estimated vol u metrically according lo Vol hard, or gravi metrically. Wines whose ashes do not burn while by gent le ignij ion usually contain considerable quant it ies of chlorine (salt). Sulphuric add. — This is to be estimated directly in the wine by the addition ^A' barium chloride. The quantitative estimation Of the BUlphnric acid is to be carried out only in cases where the qualitative tesl indicates the presence of abnormally large (piant i i ies. (in the case of viscous or very rauddj wines a previous clarifica- tion with Spanish earth is to be rec mended.) If in a special case it is necessary t.> investigate whether free Bulphuric acid or potassium bisulphafe are present, it must be proved that nunc sulphuric acid is pres- ent than is necessary to form neutral Baits with all tin- bases. Phosphoric acid. — Iu the case of wines whose ashes do ool react Btrongly alkaline the estimation is made by evaporating the wine with sodic carbonate and potassic nitrate, the residue gently ignited and taken up with dilute nitric acid: then the molybdenum method is t«» be used. If the ash reacts stronglj alkaline the nitric- acid solution of it can be used dircctl) for the phosphoric-acid determination! 88 The other mineral constituents of wine (also alum) are to be determined in the ash or residue of incineration. Sulphurous acid.— One hundred cubic centimeters wine are distilled in a current of carbonic acid gas after the addition of phosphoric acid. For receiving the dis- tillate 5 cubic centimeters of normal iodine solution are used. After the first third has distilled off, the distillate, which must still contain an excess of fret- iodine, is acidi- fied with hydrochloric acid, heated and treated with barium chloride. Adulteration of grape wine with fruit trine. — The detection of this adulteration can only exceptionally be carried out with certainty by means of the methods that have so far been offered. Especially are all methods untrustworthy which rely upon a single reaction to distinguish grape from fruit wine; neither is it always possible to decide with certainty from the absence of tartaric acid, or from the presence of only very small quantities, that a wine is not made from grapes. In the manufacture of artificial wine together with water the following articles are known to be sometimes used: Alcohol (direct or in the shape of fortified wine), cane sugar, starch sugar, and substances rich in sugar (honey), glycerine, bitartrate of potash, tartaric acid, other vegetable acids, and substances rich in such acids, sali- cylic acid, mineral matters, gum arabic, tannic acid, and substances rich in the same (e. g., kino, catechu), foreign coloriug matters, various ethers and aromas. The estimation, or rather the means of detecting the most of these substauces has already been given above, with the exception of the aromas and then, for which no method can as yet be recommended. The following substances may be mentioned here in particular, which serve for in- creasing the sugar, extract and free acid: Dried fruit, tamarinds, St. John's bread, dates, figs. B. — Rules for judging of the purity of ivinv. I. (a) Tests and determinations which are, as a rule, to be performed in judging of the purity of wines: Extract, alcohol, sugar, free acids as a whole, free tartaric acid qualitative, sulphuric acid, total ash, polarization, gum, foreign coloring mat teig in red wines. (&) Tests and determinations which are also to be carried out under special circumstances: Specific gravity, volatile acids, bitartrate of potash, and free tartaric acid quantitative, succinic acid, malic acid, citric acid, salicylic acid, sulphurous acid, tannin, mannite, special ash constituents, nitrogen. The commission considers it desirable, in giving the estimations generally per- formed, to adhere to the order of succession given above, (under («) ). II. The commission can not regard it as their province to give a guide forjudging of the purity of wine, but thinks it advisable, in the light of its experience, to call at- tontion to the following points : Wines which are made wholly from pure grape juice very seldom contain a less quant ity of extract than 1..") grama in 100 cubic centimeters wine, if wines poorer in extract occur they should be condemned, unless it can be proven that natural wines of the same district and vintage occur with a similar low content of extract. After subtracting the" fixed adds" the remaining extraot (esiraelrwf) in pare wines, according to previous experience, amounts to at least LI grams in bill cubic centimeters, and after substrooting the "free acids," al least l gram. Wines which show Less trtrwin.it are to be condemned, En case it can not be shown thai natural wines of the same districl and vintage contain as small an extmoirett. A wine which contains appreciably more ash than HI per cent, of its extract eou- teut must contain, correspondingly, more extract than would other wis. 1 be accepted as a minimum limit. In natural wines the relation of ash to extract approaches very olosely 1 to 10 parts by weight, still a considerable deviation from this relation does not entirely justify the conclusion that the w ine is alult .rated. The amount of free tartaric acid in pure wines, according to previous experience, does not exceed one-sixth of the entire "fixed acids." 89 The relation between alcohol and glycerine can vary in pure wines between 100 parts by weight of alcohol to 7 parts by weight of glycerine, and 100 parts by weight of alcohol to 14 parts by weight of glycerine. In case of wines showing a different glycerine relation an addition of alcohol or glycerine can be inferred. As sometimes during its handling in cellars small quantities of alcohol (at most 1 per cent, by volume) may find their way into wine, this fact must be borne in mind in judging of its purity. These proportions are not always applicable to sweet wines. For the individual ash constituents no generally applicable limits can be given. The opinion that the better kinds of wine always contain more phosphoric acid than others is unfounded. Wines that contain less than 0.14 gram of mineral matter in 100 cubic centimeters are to be condemned, if it can not be shown that natural wines of the same kind and the same vintage, which have been subject to like treatment, have an eqnally small content of mineral matter. Wines which contain more than 0.05 gram of salt in 100 cubic centimeters are to be condemned. Wines that contain more than 0.092 gram sulphuric acid (S0 3 ) corresponding to 0.20 grams potassic sulphate (K3SO4) in 100 cubic centimeters, are to be designated as wines containing too much sulphuric acid, either from the use of gypsum or in some other way. Through various causes wines may become viscous, black, brown, cloudy, or bitter; they may otherwise change essentially in color, taste, and odor. The color of red wines may also separate in a solid form ; still all these phenomena in and of themselves would not justify the condemnation of the wine as not genuine. If during the summer time an energetic fermentation ■commences in a wine, this does not justify the conclusion that an addition of sugar or substances rich in sugar, e. g., honey, etc., has taken place, for the first fermentation may have been hindered in various ways or the wine may have had an addition of a wine rich in sugar. The methods adopted by the "Umon of Bavarian Chemists" differ considerably from the above in many particulars, so they are given also, together with the methods adopted by the same bodv for the examination of beer 1 in somewhat condensed form. WIN E 8. METHODS OF INVESTIGATION. I. Determination of specific gravity. — This is to be done by means of a Westphal's balance or a pienometer, and always at 15° C. II. Determination of extract— Ten to 50 cubic centimeters wine at 15° C. are evapo- rated in a platinum dish on the water bath to the proper consistence, and then dried in a drying oven at 100° C. to constant loss of weight. Constant loss of weighl sumed when three weighings, with equal intervals between the first and second and second and third give equal differences between the successive weigh] Weighings are to be made at intervals of fifteen minutes. III. Inorganio matter. — This is the incombustible ash obtained by burning in- fract. Repeated moistening, diving, and heating to redness are advisable to entirely get rid of all organic constituents. IV. Acidity. —After shaking vigorously, to drive off carbonic acid, the wine le to be titrated will, an alkali solution and the acidity expressed in terms of tartaric acid. V. Glycerine.— (I) This is determined in dry wines as follows: The alcohol is driven off from I00< ui. ie centimeters wine. Lime or magnesia added, and the mass evapo- rated todrynoss. The residue is boiled with 90 per cent, alcohol, filtered, and the 1 Ililger, Vereinbarungen 11. s. w., p, IS4. 90 filtrate evaporated to dryness. This residue is dissolved in 10 to 20 cubic centimeters alcohol, 15 to 30 cubic centimeters etbei added, and the mixture allowed to stand until it is clear. It is then decanted from the sticky precipitate into a glass-stoppered weighing-bottle, evaporated to constant Loss of weight, and weighed. (2) The following method is employed lor sweet wines : < >ne hundred cubic centi- meters wine are measured into a porcelain dish and evaporated on the water hath to a Birnpy consistence, mixed with 100 to 150 cubic centimeters absolute alcohol, poured into a flask, ether added in the proportion of 1+ volumes to each volume of alcohol need, the flask well shaken, and allowed to stand until the liquid becomes clear. This is then poured off and the residue again treated with a mixture of alcohol and ether. The liquids are mixed, the alcohol and ether driven off, the residue dissolved in water, and treated as in (1). (3) In all glycerine determinations it is necessary to take into consideration the loss of glycerine due to its volatility with water and alcohol vapor, and accordingly to add to thr glycerine found O.100 gram for each 100 cubic centimeters of liquid evaporated. (4) It is necessary to test the glycerine from sweet wines for sugar, and it' any is present it must be estimated by Soxhlet's or Knapp's method ami its weight sub- tracted from that of the glycerine. VI. Ahohol. — The determination must be made by distillation in glass vessels, and the results stated as follows : One hundred cubic centimeters wine at 15 ('.contain ams "i cubic cent [meters alcohol. VII. Polarization. — (1) The wine is decolorized with plumbic subacetate. (2) A slight excess of sodic carbonate is added to the filtrate from (1). Two cubic centimeters of a solution ofplnmbic subacetate are added t<>40 cubic centimeters white wine and 5 cubic centimeters to 10 cubic centimeters red wine, the solution is Altered and 1 cubic centimeter of a saturated solution of sodic carbonate added to SI or 22.5 cubic centimeters of the filtrate. (3) The kind of apparatus used and the length of the tube are to be given, and re- sults estimated in equivalents of Wild's polaristrobometer with 200-millimeter tubes. (1) All samples rotating more than 0.5 degrees to the right | in 220-millimeter tubes, alter t real ing as above), and showing no change, or but little change, in their rotatory power after inversion, are to be considered as containing unteriiiented glucose (starch BUgar) residue. (5) Rotatory power of less than 0.3 degrees to the right shows that impure glucose has not been added. (6) Wines rotat ing bet with 0.3 degrees and <»..""> degrees to the right must be t reated by t he alcohol method. (7) Wines rotating strongly to the left must be fermented and their optical proper- ties then examined. \ 111. Suijar. — This is to be determined by Soxhlet's or Knapp's met hod. The pres- et ne of ii n tei men ted cane sugar is to be show n by inversion, etc. IX. Pota88ic bitartrale. — The determination of potassic bi tartrate as such is to be omitted. X. Tartaric, malic, and succinic acids. — (1) According to Schmidt and 1 Liepe'q method. (•>) Determination of tartaric acid according to the modified Berthelot-Fleury method. (3) If the addition of 1 -ram finely-powdered tartaric acid to LOO grams wine pro- duces no precipitate of potassic bitartrale, (he modi lied I'.eri helot -Fleurv method must be employed to detenu i ne free tartaric acid. XI. Coloring matter. (1) Only aniline dyes are bo be looked for. ial attention is to be paid to the spectroscopic behavior of rosaniline dyes, as obtained by shaking wines with amyl alcohol before and after saturation with ammonia. A qualitative test for alumina is not sufficient evidence of the addition of alum. 91 XII. Nitrogen. — To be determined according to the ordinary method. XIII. Citric acid. — Presence to be shown by a qualitative test, as baric citrate. XIV. Sulphuric arid. — To he determined in the wine after adding hydrochloric acid. XV. Chlorine.— To he determined in the nitric-acid solution of the burnt residue by Volhard's method. XVI. Lime, magnesia, and phosphoric acid. — These are determined in the ash fused with sodic hydrate and potassic nitrate, the phosphoric acid by the molybdenum method. XVII. Potash. Either in the wine ash, as the platinum double salt, or in the wine itself, hy Kayser's method. XVIII. Gums. — Presence shown by precipitation by alcohol ; 4 cubic centimeters wine and 10 cubic centimeters 96 per cent, alcohol are mixed. If gum arabic has beeu added, a lumpy, thick, stringy precipitate is produced; whereas pure wine be- comes at first opalescent and then flocculent. methods of jubgikg pubity — (Beurtheilung). Part I. I. Commercial wines may be defined as follows: (a) The product obtained by the fermentation of grape juice with or without grape skins and stems, (h) The prod- uct obtained by the fermentation of pure must, to which pure sugar, water, or infu- sion of grape skins has beeu added. It must contain not more than i) per cent. alcohol and 0.3 per cent, sugar, and not less than 0.7 per cent, acid, estimated as tar- taric, (c) The product obtained in southern countries by the addition of alcohol to fermented or partly fermented grape juice. French wines are not included, however. {d) The product obtained by fermenting the expressed juice of more or less completely dried wine grapes. II. The above definitions do not apply to champagnes. III. The following include the operations undergone by wines in cellars (KeUer- wtftesige Bekandlnng) : (a) Drawing and filling. (6) Filtration, (c) Clarification by (he use of kaolin, isinglass, gelatine or albumin, with or without tannin, (d) Sul- phuring. Only minute traces of sulphurous acid may be contained in wine for con- sumption, (e) Adulteration of wine. (/) Addition of alcohol to wine intended for export. IV. Wines, even if plastered, must not contain more sulphuric acid than that cor- responding to 2 grams potassic sulphate (KaS0 4 ) per liter. V. Medicinal wines are those mentioned in Parts I and IV, with the following re- strictions : (a ) They must not contain more sulphuric acid than corresponds to 1 gram potassic sulphate per liter, {h) They must contain no sulphurous acid, (c) The per- centage of alcohol and sugar to be given on the label, (d) These restrictions apply only to wines expressly recommended or sold for medicinal use. Pari IT. I. Improperly galli/cd wines are preparations of grape juice, pure sugar and water, or grape-Skin infusion, that contain more than 9 per cent, alcohol or less than 0.7 pel cent, acid, or both, and preparations in which impure glucose has been us.d. The following facts enable us to detect t hem : Small quantity of inorganic matter (phos- phoric acid and magnesia), and right rotation if impure glucose is used. If t he rota- tion exceeds ii. -j t«> the right, the wine is to l,e eoiicei 1 1 ia t ed , freed from tartaric acid as far as possible, and again polarized. II. Add it ion of alcohol is to bo assumed if the rat ioof alcohol to glycerine is greater than 10 to 1 by weight. III. Addition of water and alcohol is recognised by I he diminution in the quantity of inorganic matter, especially magnesia, phosphoi ic acid, and usually potash. Addi- tion of water alone is recognised in the same way. 92 IV. Scheelization, i. a., addition of glycerine, is assumed if the ratio of glycerine to alcohol exceeds 1 to C by weight. V. The presence of cane sugar is ascertained by a determination of sugar (by Sox- h let's or Knapp's method), before and after inversion. BEER. A. — METHODS OF INVESTIGATION. By beer is to be understood a fermented and still fermenting drink, made from bar- ley (or wheat) malt, hops, and water, and which was fermented by yeast. I. Determination of specific gravity. — For this as well as all other determinations the beer is freed from carbonic acid, as far as possible, by half-filling bottles with it and shaking vigorously. It is then filtered. The specific gravity is then determined cither by Westphafs balance or by a picnometer at 15° C. II. Determination of extract. — Seventy-five cubic centimeters of beer are carefully weighed and evaporated in a suitable vessel to 25 cubic centimeters, care being taken to prevent boiling. After cooling water is added until the original weight is reached, and the specific gravity of the liquid taken as in I. The per cent, of extract is ob- tained from this specific gravity by the use of a table constructed by Dr. Schultz, and is given as "per cent, extract, Schultz." 1 III. Alcohol is determined by distilling the beer. A picnometer of about 50 cubic centimeters capacity and with a graduated neck is used as a receiver. The picnometer is carefully calibrated. Seventy-five cubic centimeters of beer are distilled until the distillate reaches about the center of the scale on the neck of the picnometer. This is then cooled to 15° C, dried, and weighed, and the alcohol determined by means of Baumber's table. D .d A = 9 The percentage of alcohol by weight is to be given. In very acid beers it is neces- sary to neutralize before distilling. IV. Original gravity of wort. — This may be ascertained, approximately, by doubling the per cent, by weight of alcohol found as above and adding the per cent, of extract. As this procedure is not exact, it may be made more nearly so by using the formula 100 (E-f 2.06G5 A) 100+1.0665 A V. Degree of fermentation. This is estimated by using the formula v l = ioo(i-|) VI. Sugar deU rmination.— This is to be determined directly, in the beer previously freed from oarbonic acid, by Soxhlet's method of weighing the reduced copper; 1.13 parts of copper correspond to l part anhydrous maltose, \ 1 1. Determination of dextrin is seldom returned, and if required is to be performed by Sachsse'a method. VIII. Nitrogen. Twenty («> thirty cubic centimeters are evaporated in a Hofmeister "sohalchen" oron warm mercury, and aheextrael burned with soda-lime. Theni- n may also be determined by Kjeldahl's method. IX. Adds.- (. Rising, Haven, Conn. of Berkeley, Cal. Nitrogen. — Prof. It. A. 800 veil, of Lex- Sugar.— Prof. W.C. Stubbs, of Kennor, ington, Ky. La. Cattle foods. — Prof. G. C. Caldwell, of Ithiea, N. V. UNIVERSITY OF FLORIDA 3 1262 09216 7815 CONSTITUTION U* THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS. (1) This association shall be known as the Association of Official Agricultural Chemists in the United States. The objects shall be (1) to secure uniformity and accuracy in the methods, results, and modes of statements of analysis of fertilizers soils, cattle foods, dairy products, and other materials connected with agricultural industry; (2) to afford opportunity for the discussion of matters of interest to agri- cultural chemists. (2) Analytical chemists connected with the United States Department of Agriculture, or with any State or national agricultural experiment station or agricultural college, or with any State or national institution or body charged with official control of the materials named in section 1, shall alone be eligible to membership, and one such representative for each of these institutions or boards, when properly accredited, shall be entitled to a vote in the association. Only such chemists as are connected with institutions exercising official fertilizer control shall vote on questions involving methods of analyzing fertilizers. Any person eligible to membership may become a member at any meeting of the association by presenting proper credentials and sign- ing this constitution. All members of the association who lose their right to such membership by retiring from positions indicated as requisite for membership shall be entitled to become honorary members and to all privileges of membership save the right to hold office and vote. All analytical chemists and others interested in the objects of the association may attend its meetings and take part in its discussions, but shall have no vote in the association. (3) The officers of the association shall consist of a president, vice-president, and bary, who shall also officiate as treasurer; and these officers, together with two other members to be elected by the association, shall constitute the executive com- mittee. When any officer ceases to be a member by reason of withdrawing from a department or board whose members are eligible to membership, his offico shall be considered vacant, and a successor may be appointed by the executive*committee to continue in office till the annual meeting next following. (4) There shall be appointed by the president, at the regular annual meeting, a reporter for each of the subjects to bo considered by the association. It shall be the duty of these reporters to prepare and distribute samples and stand- aid reagents to members of the association and others desiring the same; to furnish, blanks I'm- tabulating analyses, and to present at the annual meeting the results of work done, discussion thereof, and recommendations of methods to be followe 1. (.">) The special duties of the officers of the association shall be further defined, when .uy, by the executive committee. (6) The annual meeting of this association shall be held at Buch place as shall be decided by the association, and at snob time as shall be decided by the executive committee, and announced at Least three months before the time of meeting. (7) Special meetings shall be called by tin- executive Committee when in its judg- ment it shall be necessary, or on the written request of five memberej and at any meeting, regular or special, seven enrolled members entitled i<> vote shall constitute a 1 1 nor, mi foE t he t ran sac t ion of hii-di- ■ tive committee shall confer w ith the official boards represented with reference to the payment of expenses connected with the i itings and publication of the proceedings of the SSSOOiaJ ion. (D) All proposed alterations or amendments to this constitution shall be referred Led committee of three at a regular meeting, and after report from such com- mittee may b- adopted by a voto of two-thirds of the mombers present and entitled to rote.