Hackicti- & Crowley — Method of Separating Fats in Butter-Fat. 367 



first fraction, 37o; third fraction, 3-4-8; sixth fraction, 340; butter-fat 32. 

 This approximation towards the value for butter may be due in part to the 

 incomplete removal of the butter from the paper by scraping. Assuming the 

 presence of several substances with different rates of diffusion, it is also possible, 

 though not immediately evident, that the composition of the extract may depend 

 on the time of diffusion. These considerations suggested some modifications in 

 the method of working. 



The next series of experiments was accordingly arranged to obtain a more 

 uniform extract. The time for diffusion was now reduced to three hours. 

 The paper on which the cylinders rested was cut out of the filter-paper before 

 extraction, so as to obtain only the substance which had diffused. This extract 

 was a golden-brown liquid which set to a butter-yellow gel-like solid on standing 

 at 15° C, and gave a stationaiy temperature on the cooling curve at 11-7° C. 

 The substance was still moist in appearance below this temperature, and when 

 placed in a room at 10° C. it yielded an oil to filter-paper. The solidifying- 

 point of the residue removed from the filter-paper was found to be 11-6° C. 

 The other extracts obtained between 20° and 21° C. Avere also allowed to diffuse 

 into filter-paper at 10° C, so as to obtain these extracts as free as possible 

 from lower melting products previous to analysis. 



The iDresence of at least two diffusing oils with different solidifying-points 

 explains the obseiwations made at 16° C. At this temperature 3 grams 

 of extract were obtained in three days. It remained liquid at room temperature. 

 On taking a cooling curve it gave a pasty mass between 12° and 10° C, and 

 solidified completely about 5° C. The two oils were evidently present in nearly 

 equal proportions. No chemical analysis was made on this sample. 



An experiment on diffusion was also made in a cold room whose temperature 

 varied between the limits 9° and 10° C. The process was very slow, and not 

 more than 1 gram was obtained in three weeks from 74 grams of butter on 

 five filter-papers. To avoid the possibility of extracting any butter particles, 

 the paper on which the cylinders' rested was cut away before insertion in the 

 Soxlilet apparatus. A determination of the iodine value gave 82. The iodine 

 value for olein is 86. We may infer, therefore, as one might expect, that the 

 oil diffusing in the cold is olein, but that, in agreement with Amberger's 

 results, the quantity present is small. 



The authors are indebted to Mr. Brownlee, Agricultural Analyst of the 

 Department of Agriculture, for an examination of the butter used in these 

 experiments, and of the fat diffusing from it at 20° C. A sufficient quantity of 

 the latter was obtained by combining all the extracts together, and they were 

 partially purified by cooling them to 10° C, and absorbing the liquid present 

 at ■ that temperature in filter-paper. The comparative determinations include 

 the Reichert-Wollny number, indicating the volatile fatty acids soluble in water, 

 and the Ave-Lallemant number, which is calculated from the difference between 

 the insoluble and soluble barium salts of the acids present. Mr. Brownlee (3) 

 has shown that the Reichert-Wollny number has a seasonal variation for Irish 

 butter, reaching its peak value between March and June. As the sample of 

 butter examined was made in March, the figures obtained in the above two 

 determinations are consec[uently above the average. 



Reichert-Wollny number 

 Ave-Lallemant number 

 Koettstorfer's saponification value 



