19 
In these tables we cannot but notice the individuality of the cows. 
Juno had a less capricious appetite than Ceres, drank less water, and 
required less food in proportion to her live weight ; while Ceres was 
the better milk yielder. These facts find comparison in the follow- 
ing table : 
Average daily record for each period. 
Wt. of cows. Wt. of food. Perct. relation. Wt. water drank. Milk yield 
Ceres. Juno. Ceres. Juno. Food to weight. Ceres. Juno. Ceres. Juno 
Lbs. Lbs. Lbs. Lbs. Ceres. Juno. Lbs. Lbs. Lbs. Lbs. 
Period I. 759 «967 22. 22. 2.9 2.3 772) 70.9. 20.6.:°19.8 
IL. 744 919 82.2 37.7 438 41 58.2 41.4 24.4 28.9 
1¢: 746\. 905. 22:4 19.9 . 3.0. 2.2... 76:3 . 66.1, 19.9 18.8 
‘ea 745 893 244 215 3.38 24 74.7 82.9 245 20.7 
; ae ec eseo) eh og | 28 IG. Oy) wOsek nore 
Ng TOL 00a “Boo 204 Sel a 2.06~ 69,L) | OD) 24.6 ) 104 
VII. 726 ' 891 ‘834 361 °°46 41° 49.9 843 ‘21.6 17.5 
VIII. Che STC 920250 17.65.2.9. 20. 6B 04ST NRL TS 14.5 
The hay fed was of excellent quality, and when mixed foods were 
fed, a different proportion was consumed by the two cows. We 
hence supplement this table by another, in which the amount of 
albuminoid consumed daily is calculated, and the average fat in the 
morning and evening milk is given, but there were fewer analyses 
of the evening than of the morning’s milk, which destroys the com- 
parison except as between the two cows. 
Per ct.of butter 
Albuminoid con- Per cent. offatin Percent.offatin churned from 
sumedin ration. morning’s milk. evening’s milk. morning’s milk 
rae: jung, Ceres. Juno. Ceres. Juno. Ceres. Juno. 
Ss. Ss. 
Period I. 2.28 2.28 4.23 4.73 4.06 4.29 
II. 2.64 2.84 3.85 3.89 4.68 5.27 3.70 3.04 
Jit. 2.32 2.06 3.92 4.09 4.12 5.65 3.49 3.04 
IV. 2.97 2.68 3.65 Spats 4.62 5.88 3.52 3.58 
V. 2.99 2.64 3.50 3.75 3.52 3.38 
MI. 2.86 2.88 3.37 8.75 3.41 3.24 
VII. 2.08 2.08 3 95 3.96 3.61 3.93 
VILL. 2.12 1.82 3,40 8.46 3.54 2.74 
Calculating the various rations into their equivalent in hay, we 
find that the albuminoid which represents 24.4 lbs. of hay, or 3.3 
per cent. of her average live weight sufficed to maintain Ceres un- 
der the conditions of the trial, with an average daily milk yield of 
23.3 lbs ; while the albuminoid equivalent to 23.2 lbs of hay, or 2.6 
per cent. of her live weight sufficed to maintain Juno under the 
conditions of the trial, with an average daily milk yield of 20.2 lbs. 
When we consider the teaching of our tabular data, that even 
when fed ad libitum, neither is the same amount of food eaten by 
the two cows, nor the same amount by the same cow during succes- 
Sive intervals, it is seen that we have not figures which can justly 
be incorporated into averages for the two cows, and hence we must 
examine, as to the influence of the food, the records of each cow sep- 
arately. If we do otherwise, and average our results, the figures in 
many cases cancel each other, instead of becoming incorporated. 
Thus if one of the two cows has a strong appetite, and feeds largely, 
while the other cow has a capricious appetite and refuses her food, 
the union of the data of the two into averages tends to destroy the 
value of co-relations which might appear between the food and the 
yield, whether expressed or live weight or milk. 
Hence to rightly study our data, we must examine the co-relations 
