New York AGRICULTURAL EXPERIMENT STATION. 89 
out on pages 81-84 a test was made of the effect of filling only 
the suction filter cup with cotton and contrasting this with 
the results where this cotton was not used. In this test the 
rubber parts were kept in 10 per ct. brine and the machine, 
No. 2716, was handled with care in every respect, the only 
difference being in the presence and absence of cotton in the 
suction filter. The results of these tests are given in Table X. 
TABLE X.— EFFECT OF PLACING COTTON IN THE SUCTION FILTER ONLY. 










Cotton. No cotton. 
1908. Nora. | Bess. 1908. Nora. | Bess. 
Germs per cc. of Germs per cc. of 
milk. milk. 
BOTCON ONES: sek ess doc: 1,800 Ge EON OCCy A tetera a: 1,400 8,590 
a Gi he RR irae 2,000 1,700 ‘ See te ae ee se 74,520 176,620 
: Dre Bae 6) okt « PA 0) 4,190 * iO RSE Use ant ends © 30 , 240 Ja PSD) 
a? 1 Recerca Caen 1,100 3,860 a i AY (eee MP Ue ae Pati 1,940 2,700 
is on Ss ae orice 4,140 6,200 4 UO Wer eis glare they: 14,200 e250 
- Sete tats ee es 9 , 200 5,040 ETE OTR eA Oe ait 34,350 7,500 
Fs evransiwate feicte The re) 9,040 
PARVIE DUC CN cist nudy osc eiielladens 3 3,650 OOO | PAVeTa Se’ falaen sofas cuseers 3 26,110 6,260 
General average...... 4,405 General average..... 16,185 



These results indicate that the use of cotton in the suction 
filter alone will markedly improve the quality of the product. 
This improvement is shown especially in the reduced range of 
variation inthe counts. 
The average of the results where no cotton was used has 
little value on account of the wide variation in the results but 
this variation is characteristic of the absence of filters, and 
the average expresses the marked tendency toward higher re- 
sults. In computing this average the high count from Bess 
on March 9 was omitted. Her milk entered the side of the 
machine opposite the suction filter where the absence of the 
cotton ordinarily had less effect and the high count was un- 
