CHEMISTRY. 109 



This method was compared with the official method on 24 samples of 

 butter. The diiiereuces iu the case of some samples were noticeable, 

 an)ouuting- in 9 cases to over 0.5 per cent. Duplicate determinations 

 by the proposed method agreed rather more closely than by the official 

 method. 



"The mean of the differences between the two methods is — 0.13 per cent, wliich 

 indicates that by tlie new method nothing l)iit the fat was in solution. Or it might 

 indicate that by the official method the fat was not completely dried." 



Determination of fat in cheese, S. Bondzynski [ZUclir. anal. 

 Chem., 33, JSTo. 1:~\ pp. 186-189). — The author has applied the principle 

 of the Schmidt-Bondzyuski method for milk analysis (E. S. R., 5, p. 

 801) to the analysis of cheese, changing- only the proportions. In detail 

 the method is as follows: A weighed quantity (amount not stated) of 

 finely ground cheese is placed iu the calibrated tube used in the above 

 method and heated with 20 cc. of nCl of 1.1 specific gravity (about 19 

 per cent) until the cheese is dissolved and the melted fat rises to the 

 surface. The tube is then cooled and about 30 cc. of ether added, 

 which dissolves the fat readily without shaking. The ether-fat layer 

 is separated from the HCl by heating for a time at about 40° C, and 

 this separation is aided by centrifugal treatment. The volume of the 

 ether-fat layer is read off on the scale, 20 cc. of this evaporated in an 

 Erlenmeyer flask, and the result calculated for the whole layer. 



The results of parallel determinations agreed well with each other and 

 with the results obtained by extraction in a Soxhlet extractor. 



Method of sampling milk for analysis, W. Freak, and J. W. 

 Fields {Pennsylvania Sta. Bpt. 1892., pp. 146, 147). — In analyzing com- 

 posite milk samples for nitrogen by the Kjeldahl method, in one case the 

 composite samples were made rather small and the whole quantity 

 analyzed, and in another they were made larger and a subsample taken 

 for analysis, the milk being preserved with corrosive subhmate. The 

 two methods gave equally accurate results, but the analysis was more 

 easily executed with the latter method, which is preferred. 



Report of the Chemical Control Station at Alnarp, Sweden, 

 for 1893, M. Weibull {pp. 9). — The report shows that 755 samples of 

 soils, fertilizers, feeding stuffs, butter, water, etc., and 2,579 samples of 

 milk were analyzed during the year. Of the latter number 2,204 samples 

 were examined for their fat content by the lactocrite method. T'he new 

 milk analyzed usually contained between 3 and 4 per cent of fat; 92 

 samples contained 2.0 per cent or less; and 81 samples contained 4 per 

 cent or more of fat. Of the 153 samples of skim milk examined, about 

 80 were separator skim milk, containing from 0.19 to 0.43 per cent of 

 fat. The skim milk produced by gravity setting gave higher results, 

 the maximum tat content found being 1.12 per cent. The 10 samples 

 of buttermilk analyzed contained from 0.2 to 0.68 per cent of fat. 



Two dairy investigations are reported, viz: (1) the application of 

 potassium permanganate for the preservation of samples of milk for 



