340 
Tests were next made to determine how small a propor- 
tion of adulteration could be detected with certainty in this 
way, with the following results: 


Caustic potash used: grams. LO AS eG 
DUVUEL eo eta ee ore LL Reta eee gona TPH gir teas 2. 67 78 114 
50 per cent butter 50 per cent neutral lard ...... ...... .. 416 
66 “6 «sh Bt Pa: 66 66 6 9 
SUrhoeh edn GU liane Rie a ieck eae aie 
ORT elo Ee GB |S, ld Ue Seiler 
a BO ess pie kad wana hard zs Dali slg tat ina AM (2 a4 
ie ys ty ued oe 8 SOSAME OllL ake eee By dred 
BD ERS NS RE AERO ATS EU SOT SNE SSR IORETD BRET ae ee 
Qa is AY! Sy WO S GE " aR pore Reena eee .. 114 258 
1 Cee TE Leet eee Pane ee eee 1916 QBN as 
The apparent discrepancies in this table may be account- 
ed for by the fact that the butter fat used in these tests wag 
not all taken from the same sample. These figures like 
those in the table preceding were mostly obtained with a 
defective apparatus and must be considered only ina rela- 
tive sense. No tests were made with less than five per 
cent. of adulteration, but there is little doubt that by mak- 
ing the solution more alkaline the addition of one per cent, 
of adulteration to any given sample of butter would be 
readily shown. 
. The butters used in these experiments were mostly made 
at the Station, from milk of the Station herd consisting of 
both native and Jersey cows. In some cases the butter from 
individual cows was used. Two or three samples were ob- 
tained from groceries in Geneva and one sample of J ersey 
butter was nearly a year and a half old. There were in all 
nearly 20 samples of butter tested. When only 10 grams 
of caustic potash were used there was practically no differ- 
ence in the viscosity of the soaps obtained and with 15 
grams of potash the variation was small, not enough to 
prevent the detection of 5 per cent. of adulteration. 
In Ootober, some time after the above tests were complet- 
ed, a large number of butters were obtained from different 
sections of the country, which included samp'es made from 
milk of different breeds of cattle and from milk treated 
in different ways. Several of these butters were taken 
from the exhibit of the Bay State Fair, held in Boston early 
in October. These were tested with a new viscometer in 
which the faults of the first instrument were, as far as pos- 
sible, corrected. 
The variation in the last lot of butters was so great as to 
materially modify the conclusion reached above. The fig- 
ures obtained, together with those from many other fats and 
oils, will be found in the next table. In order to avoid the 
trouble of weighing the caustic potash for each test a very 
concentrated solution containing 2 grams of fused potassium 

