oe. 
CRHEAMERY PROBLEMS. 39 
One of the advantages of the separator over the Cooley 
creamer is in the more complete removal of the butter fat 
from the milk. Table 8 gives the Babcock tests of thirteen 
samples of skim milk from Cooley creamers and thirty-six 
samples of skim milk from-hand separators. These samples 
were taken at the farm by the dairymen and forwarded to the 
Station by mail. Instructions were given to thoroughly mix 
the skim milk of one skimming before taking the sample. The 
tests were made by the Babcock method in the usual way, and 
the readings were taken from the double neck Wagner skim 
milk test bottles. 
TABLE 8. 
Per Cent. of Fatin Skim Milk from Hand Separators 
and Cooley Creamers. 



| Sharples 




Cooley | De Laval ¥ Utes: National | _Empire Pee as tualay 
Creamer. | Separator. Separator. = Separator. | Separator. Separator. 
% | 7 qe Te | %o | % 
.09 .O5 .O4 .03 .04 .02 
EEs10 Ot Os 07 1060 al .08 
sep? 203 .O7 £05 .06 — 
.20 | .OI 215 elZ .08 — 
m2 .03 Jif .O4 — — 
16 | Ol FOR 15 — | — 
Gs. .05 .03 — — | — 
.26 04 502 —~ — — 
212 .O1 .05 — — — 
.2t 05 eo — _ | — 
.16 .06 .O4 — — — 
+30 .02 — — _ a 
50 i183 — — a a 
| Ea == = } = | — 
Ave. 27 . .038 064 .076 | .06 .05 

Average of 13 samples of Cooley skim milk, 0.27 4. 
Average of 36 samples of separator skim milk, 0.056 %. 
These results are no doubt too low. ‘There isa small amount 
of ‘“‘residual fat’’ in all milks, not recovered by the usual 
Babcock method. In full milk samples this error is compen- 
sated for by the method of reading the tests. In skim milk 
tests an addition of about one-tenth per cent. to the reading 
would give results more nearly agreeing with the gravimetric 
analysis. Even with this addition the results show a remark- 
ably efficient skimming by the hand separator in the hands of 
