314 
Milk. 
aperture, through which a thermometer may be inserted, and the 
temperature of the milk ascertained. 
All that is necessary is to fill these graduated tubes up to zero, 
and after 24 hours to read off the number of degrees occupied by 
the cream. The milk should be kept during the experiment as 
nearly as possible at a temperature of 62' Fahr. In using these 
instruments for comparative trials, it is necessary to observe in 
the first place that the temperature be the same in all trials, for 
direct experiments have shown me that somewhat less cream is 
obtained when the temperature rises above 62' Fahr. ; or rather I 
should say that the cream which is thrown up at a more elevated 
temperature than 62' Fahr., occupies somewhat less space than 
that which rises at a lower temperature. In tlie second place, in 
all trials the milk should be left standing: for the same length of 
time, which may be either 18 or 24 hours. If left for a longer 
time I find that the bulk of cream slightly diminishes. During 
the longer period more cream rises ; but as a more complete 
separation between the liquid portion of milk and the cream 
globules takes place under these circumstances, notwithstanding 
the larger amount of cream, its bulk slightly diminishes, as will 
be seen in the following trials : — 
1. 100 measures of new milk yielded in the creamoraeter 13i 
measures of cream after 18 hours ; the same quantity after 24 
hours, and scarcely 13 measures after 48 hours. 
2. 100 measures of another sample of new milk gave nearly 
the same results. 
3. 100 measures of a third sample showed 13 measures of 
cream after 18 hours; the same quantity after 24 hours, and 12 
measures after 48 hours. 
There are two circumstances w hich seriously interfere with the 
practical use of the creamometer, and make its indications unre- 
liable. The first is that the cream which rises from different 
kinds of milk often varies greatly in composition. Proofs of this 
have been given already in the experiments cited in a former 
part of this communication, by which it was distinctly shown 
that the largest amount or the thickest cream does not always 
give most butter. The indications of the creamometer, there- 
fore, are fallible when samples of milk, produced under very 
different circumstances, have to be tested. 
The second disturbing circumstance in the use of this instru- 
ment lies in the fact that milk which has been agitated, as it 
necessarily will be, when sent by rail, throws up less cream than 
that which has been less disturljed. A direct experiment shows 
this very distinctly : — 
100 measures of new milk, after standing for 24 hours at 
62^ Fahr., gave 12 per cent, of cream by measure, whilst, at the 
