156 



dried in an electric oven for two hours at 100 C. or in a vacuum oven below 70 C. 

 (gauge reading, 29 inches). 



Evaporation in a flat-bottomed dish without sand yielded low results, probably due 

 to a retention of moisture by the nitrogenous film which formed on the surface. -Simi- 

 lar treatment in a vacuum oven gave higher results, but below that obtained on sand. 

 These figures are not reported. 



Ash: Twenty-five cc were evaporated in a platinum dish with 5 cc of concentrated 

 nitric acid and burned to a white ash. 



Protein: Ten cc were treated by the Kjeldahl-Gunning method. 



Fat: Our work of last season indicated that a single extraction gave higher results 

 than the double and saved time and work. The dried solids on the sand were pul- 

 verized, washed with water, dried, and extracted with dry ethyl ether in a continuous 

 extractor. Long heating of the residue containing fat (at 100 C.) should be avoided, 

 as it appears to reduce the amount of fat that can be extracted. 



C. H. Jones, Vermont station: The result reported on fat by the modified Babcock 

 method is the average of ten determinations on three distinct 40 percent solutions. 

 The individual readings were 2.60, 2.65, 2.80, and seven readings of 2.70. Lactose 

 determinations were made on distinct portions of the original sample. In the sucrose 

 determination it was necessary to dilute the solution after inversion to 200 cc in order 

 to have an excess of copper in the Fehling solution. 



The- result reported by the Leach method is the average of two determinations, 

 70.70 and 70.58 per cent* respectively. The platinum dish used did not have an ab- 

 solutely flat bottom, but it was the nearest approach to anything of the kind available. 

 I am at present unable to explain the difference obtained by the two methods, unless 

 a too complete drying and consequent breaking down is obtained with Method A. 



The result reported on the double-extraction method for fat is the average of two 

 determinations from distinct solutions. They were, respectively, 8. 20 and 8.12 per 

 cent. The only awkward feature is the size of the filter paper used. The following 

 modification of the Babcock method described was suggested by the use of the hard- 

 ened filter on other laboratory determinations: 



Method: Place 15 cc of the 40 per cent solution (6 grams) in a small-lipped beaker, 

 diameter 1.5 inches, height 2 inches. Dilute with an equal amount of water; add 

 4 cc Fehling's copper solution; stir with glass rod. Filter through a 12.5 cm C. S. and 

 S. 575 hardened filter. Wash thoroughly with water; stir on the filter with glass rod 

 (100 cc is usually enough, though 160 cc had no lowering effect on the result). Return 

 precipitate to original beaker, removing any remaining particles by washing with hot 

 water through a fine-jet wash bottle. (The bulk is easily kept below 17 cc.) Stir 

 with a glass rod. Pour into Babcock milk bottle. Add a portion of the acid to the resi- 

 due in the beaker. Mix thoroughly, using stirring rod. Transfer to Babcock bottle. 

 Repeat with remainder of acid. Shake milk bottle thoroughly, and then rinse beaker 

 with a little hot water from the wash bottle and put into test bottle. Run as usual. 

 The individual results obtained by this procedure were 2.70 for five readings and 2.65 

 on the sixth, three different 40 per cent solutions used. 



I find it desirable, both with this method and with Method B, to use rather more 

 sulphuric acid than is specified ; often 18.5 cc. While the results by this procedure are 

 not different from those obtained by Method B, yet when a number of samples are run 

 a considerable gain is made in actual working time. 



DISCUSSION OF RESULTS. 



The results obtained this year are quite satisfactory on all determinations except 

 fat. The lactose results, with three exceptions, are probably as good as one could 

 hope to get from a number of chemists working independently and not making a spe- 

 cialty of sugar determinations. There are some variations in total solids for which it 

 seems difficult to account. One might think it due to variations in the different cans 

 of milk were it not for the fact that in some cases when the solids were as much as 2 

 per cent low, the other determinations, such as proteids, ash, etc., were as high or 

 higher than the average. The referee can only account for these discrepancies in 

 one of two ways, faulty sampling, or that the sugars were allowed to ferment and 

 cause loss before the determinations were made. It is believed that all determina- 

 tions should be made as soon as the solutions are made up, and no solution which has 

 stood in a warm laboratory twenty-four hours should be used for the determinations 

 of solids or sugars. Leach's method appears to give high results, probably because of 

 the large amount of sugar present to hold the water. 



