212 



Althougb the results are somewliat limited, it is stated that this is only 

 the beginning of a work to be more extensively followed up. 



If we compare Nos. 2, 3, 22, au.l 8 [Ihe time since last calf being in each case two 

 hundred and forty days] we find that No. 2 produced twice as much butter fat as No. 

 3, and nearly five and one half times as much butter fat as No. 8, and that No. 22 

 produced seven and one half times as much butter fat as No. 8. 



Comparing No. 13 with No. 14 shows that nearly twice as much milk must be 

 handled by the owner to get the sama weight of butter fat from No.. 14 as from No. 

 13. Besides these extreme cases mentioned, cows can be found all along the line 

 from very profitable to very unprofitable. 



(2) Variations in themilJc supplied to creameries hi/ (liferent patrons.— 

 Tests were made of the milk brought by one hundred and thirteen 

 patrons to two creameries on the same day. The results, stated in a table, 

 show the percentages of fat in the milk, the pounds of milk per pound 

 of fat, pounds of fat in the milk, and pounds of milk. The milk supplied 

 by two patrons of creamery A contained 3 and 3.1 per cent of fat each, 

 and that from seven patrons 4 to 4.4 per cent, "making a difference 

 of 1.4 per cent between the highest and the lowest," that is to say "the 

 richest milk was 47 per cent richer than the poorest." The pounds of 

 milk per pound of fat, in the milk from fifty-five patrons, ranged from 

 22.7 to 33.3 pounds and averaged 27.3 pounds. 



Equally striking illustrations could be drawn from the record of creamery^, B. 

 Eighty per cent of the patrons supplied milk ranging from 3.2 to 3.8 per cent of fat, 

 but the milk brought by one patron ran as low as 2.3 per cent, and that brought by 

 another as high as 4.6 per cent of fat; that is, one contained just twice as much but- 

 ter fat in 100 pounds of milk. If the richer milk is received at $1 jier 100 pounds, 

 for the poorer but 50 cents per 100 pounds should be paid. 



Trials of milk tests. — The methods of Short, Cochran, Patrick, Fail- 

 yer and Willard, and Parsons are briefly described, and the results of 

 analyses by the first three as compared with those by the gravimetric 

 (on sand and on paper) are tabulated. On the whole the comparison 

 was very favorable for the quick methods, the results rarely being more 

 than 0.2 above or below the gravimetric, and the average difference 

 very small. The determinations of fat in buttermilk were somewhat 

 less satisfactory than those in whole and skim-milk. Observations are 

 given on the kinds and costof reagents used, approximate time required 

 for analyses and precautious necessary to accuracy. It is stated that 

 with the Cochran method a single determination could be made iu a 

 half hour, or twenty-four analyses in one and one fourth hours, and at 

 a cost of about one half cent for re-agents per determination ; and with 

 the Patrick method a single determination in about twenty minutes, 

 or six determinations iu a half hour, the cost of chemicals at the fig- 

 ures given being about 1 cent per determination. 



Outside of the correct graduation of the measuring parts of the apparatus, the ac- 

 curacy of the results obtained by using such methods as these, depends, to a control, 

 ling degree, on constant attention to such small details as measuring the fat in clean 

 glass, and carefully and uniformly reading and recording the length of the fat col- 



