Effect of Fibrin 171 
coefficient of expansion of butter fat is more than three times as 
great as that of water, so that in order to maintain the same 
relative difference in their specific gravities, when the tempera- 
ture is falling, the milk serum must cool more than three times 
as rapidly as the fat. In other words, when the milk serum has 
cooled from 90° F. to 40° F., or through 50°, the fat globules 
should have lost less than 17°, and should still have a tempera- 
ture of over 73° F., a difference between the temperature of the 
fat and serum of more than 33°. Such a condition is mani- 
festly impossible, but any less difference than this would cause 
the fat to become relatively heavier than at first, and would 
operate against the creaming. 
On the whole, the most satisfactory explanation of 
the good effect of the low temperature in the deep 
setting system is that advanced hy Babeock,* that 
the presence of fibrin in milk, especially when it has 
coagulated in the form of threadlike masses, permeat- 
ing the fluid in all directions, offers a considerable 
obstacle to the rising of the fat globules. The sudden 
reduction of the temperature quickly after the milk 
is drawn, by preventing the formation of these 
fibrin clots or threads, aids in the separation of the 
fat. This would be entirely satisfactory were it not 
for the fact that it has been shown that while it 
is usually of advantage to cool the milk imme. 
diately after it is drawn, in some cases, at least, 
the setting and cooling may have been delayed for 
a time long enough to permit the formation of 
fibrin clots without appreciable effects upon the sep- 
aration, as the following tables} very clearly show, 
‘the efficiency of creaming being measured by the 
percentage of fat in the skimmed milk: 
* Loe. cit. ie 
_ ¢ Cornell. University Agricultural Experiment Station, Bull. No. 29, p. 78. 
