544 
butter with 4 drops of color added. These tests showed no 
marked difference between butters colored and uncolcred. 
The viscosity is practically the same for all of ‘the butter 
soaps when 15 c.c. of potash solution is used, the average 
for the tests in this table being 78, the maximum 84 
and the minimum 65. There seems to be no difference 
which can be attributed either to breed of cow or to methods 
of manufacture. In the second column with 20 c.c. of pot- 
ash solution the difference in the butters is very marked, 
the average being 145, with a range of from 72 to 405. The 
difference with 25 c.c. is still more decided, the average 
being 289, with a range from 124 to a point where the so- 
tion was too viscous to be determined by the apparatus as 
arranged. 
+ With this amount of alkali the Holstein butters are de- 
cidedly more viscous than any others, but the number of 
samples examined was not sufficient to establish this as a 
breed characteristic. The two samples of Holstein butter 
tested were of prime quality and would compare favorably 
both in appearance and flavor, with the best samples tested. 
The cause of this difference is difficult to trace. I supposed 
at first that it could be attribnted to varying quantities of ° 
butyric acid, this being the one constituent which pre-emi- 
nently distinguishes butter from the other animal fats. 
This, however, can scarcely be the case, as examinations of 
the following butters by Reichert’s test show. The amount 
of decinormal soda required by this test was, for No. 11= ° 
13.9 c.c., No. 18=13.6 c.c., No. 23=13.2 c.c., and for No. 25 
=14.6 cc. These nnmbers should be proportional to the 
butyric acid in the butters, but they evidently have no re- 
lation to the viscosities found. 
Contrary to the prevailing opinion of chemists, it seems 
certain that the fatty acids of butter have properties differ- 
ent from the corresponding acids derived from other fats. 
Stearic acid from tallow when treated with 15 cc. of KOH 
solution gave a soap which was a thick jelly at 20°C. Oleic 
acid gavea very viscous soap. Twenty grams of butyric acid 
required more than 15 ¢.c. of potash solution to neutralize 
it. With 30 c.c. it showed about the same viscosity as but- 
ter with 15 c.c. and was not materially increased by addition 
of more. Palmitic acid was not examined, but there is 
every reason to think that it would behave in nearly the 
same way as stearic. A mixture of these acids, excepting 
the last, made in nearly the proportion occurring in 
butter gave a soap that was extremely viscous, see test 
63. On the other hand the insoluble acids from. several of 
the butters gave soaps less viscous than the butters from 
which they were obtained, although they differed in the 
same way as the butters. It is evident that the stearic and 

