25 — 



each level, down to half divisions ; subtract the smaller number from 

 the larger, and in the following table, against the number corre- 

 sponding to this remainder, find the per cent, of fat in the milk. 



Mr. Cochran stated in his original article on this subject, that 

 this method of determining fat in milk was then (1889) in use in 

 upwards of fifty creameries, mostly in southeastern Pennsylvania, 

 and that with the largest and most improved form of apparatus 

 for the heating of the samples, and after becoming perfectly 

 familiar with the manipulation, sixty tests can be made in from 

 two to three hours. 



If scrupulous attention is paid to even- detail of the manipula- 

 tion, as above described, I think that any one who is not too 

 clumsy-fingered and is accustomed to nice careful work, can get 

 reliable results with the method after some practice, such as 

 any method would require. But in any case a little training un- 

 der the instruction of someone who is perfectly familiar with it 

 would undoubtedly be profitable. 



With any such who can spend a week at Ithaca, arrangements 

 will be made, by special correspondence, for such instruction in 

 the laboratory of the Experiment Station. Only two pupils at 

 the most can be taken at once, and therefore this instruction will 

 necessarily be distributed over a large part of the year, in case 

 many should apply for it. All correspondence in regard to the 

 matter should be addressed to Chemical Department, Cornell 

 University. G. C. CALDWELL. 



