DAIRY PRODUCTS. 69 



by adding water to the contents of the retort, again distilling, and repeating this 

 process several times, a very considerable proportion of volatile fatty acids can bo 

 obtained from cocoanut oil. In assaying butter, the appearance of insoluble acids in 

 the distillate would furnish a valuable indication of the presence of cocoanut oil, and 

 they should be removed by filtration, or the distillate will be found to neutralize so 

 large a volume of alkali as considerably to diminish the practical value of the process 

 as a means of distinguishing butter from butter substitutes, as has been pointed out 

 by Moore aud others. Latterly, I have adopted the plan of filtering the distillate in 

 all cases, washing the filter with cold water, and then immersing the filter, with any 

 adhering insoluble acids, in alcohol, which is then titrated with decinormal alkali 

 and phenol-phthaleiu. In the case of ordinary butters and butter substitutes the in- 

 soluble volatile acids only neutralize about 0.2cc. of decinormal alkali. 



The question having recently been raised, the following experiments were made at 

 my request by Mr. William Barraclough on a sample of butter fat, in order to ascer- 

 tain the variation in the results of Reichert's process produced by modifications in the 

 methods of conducting the saponification and distillation : 



(1) Two and a half cubic centimeters of butter fat was saponified by alcoholic potash 

 ' in an open basin, the alcohol evaporated off completely at a steam heat, the residual soap 



dissolved in water, the solution acidulated with' sulphuric acid in slight excess, di- 

 luted to 75cc. and distilled gently in a globular flask with side tubulure adapted to a 

 condenser until 50cc. had passed over. The flask held 460cc. up to the side tube, and 

 some fragments of pumice-stone coiled round with platinum wire were added to the 

 contents to promote evolution of vapor. 



(2) An exact repetition of No. 1 experiment. 



(3) Saponification was effected in a flask furnished with a long tube and heated by 

 steam. The subsequent manipulations were the same as in experiment 1. 



(4) Saponification was effected in a well-closed bottle placed in the water oven. 

 Other manipulations unchanged. 



(5) Manipulation exactly as in experiment 3, except that the distillation was con- 

 ducted in a flask fitted to the condenser by a cork and bent tube. 



(6) Conducted as in experiment 3, except that the distillation was conducted in a 

 retort. 



(7) Blank experiment with the alcoholic potash employed in the previous experi- 

 ments, the manipulation beiug that in experiment 3. The alcoholic potash was 

 brown and not very recently prepared. 



Experiments. 



These results show that a sensible loss occurs if the saponification be conducted in 

 an open basin, doubtless owing to the formation of butyric ether. On the other hand, 

 the exact nature of the distilling apparatus appears to be of little importance. This 

 latter conclusion is not in accordance with the experience of some other chemists. 



Zulkowsky and Groger 1 have studied and modified Haussinan's 

 method 2 of volumetric fat analysis. This method is based on the fact 



!Ber. Chem. Gesel., vol. 16, p. 1140. 



2 Ding. Poly., J., vol. 244, p. 303, and vol. 246, p. 286. 



