20 BULLETIN 501, U. S. DEPARTMENT OE AGRICULTURE. 
The data from Massachusetts and Connecticut are results from the 
experiment station. herds, while the data from New Hampshire are a 
compilation from all the herds of the Lyndeboro Cow Testing Asso- 
ciation. These data differ in some respects from that procured from 
complete cost records. However, considering the difference of con- 
ditions and methods under which the herds were handled, the results 
from all these sources closely coincide. 
In each case the feed cost has been determined from actual records. 
Although many of the items other than feed were determined from 
estimates for the New England herds, they compare closely with the 
same items from actual records. By comparing Tables I and XIII, 
it will be seen that the total cost of keeping a cow, both on the two 
experiment station farms and in the cow-testing association, is some- 
what higher than on the four farms discussed above. However, in 
terms of the proportion which feed, labor, and other costs bear to 
total cost, they check closely. In the feeding of a grain ration the 
entire year the Michigan herd is handled similarly to the Connecticut 
and Massachusetts experiment station herds, and trie cost of feed in 
these three herds is 57.2, 56.7, and 55 per cent, respectively, of the 
total cost of keeping a cow. 
The New Hampshire figures were compiled from farm herds where 
pasture was influential in reducing the feed cost. In this respect 
these herds are not unlike the Wisconsin, Pennsylvania, and North 
Carolina herds, and the feed is about one-half of the total cost. 
The cost for labor, as shown by the New England records, is a little 
higher than on the four farms; however, the per cent of the total 
cost is a trifle lower. The cow-testing association record shows the 
total labor to be slightly more than one-fourth of the total cost, 
whereas at the experiment stations, where no marketing is included, 
the labor is less than one-fourth. 
The individual items other than feed and labor show greater diver- 
gencies, but considering them as a whole they compare fairly well. 
Some of the charges included in the individual items are not the same. 
For instance, "Overhead" is entirely omitted from the New England 
records, and a part of equipment-cost charges are either omitted or 
were too closely linked with other items to be separated. 
The item "Miscellaneous," as shown in Table XIII, takes care of 
a share of what might otherwise be called overhead. It also includes 
in each case a charge of about $5 per cow for bedding. This item of 
expense may show up on many herds, but was so small on the four 
farms that it was included in the feed cost. Its consideration here 
tends to offset the items not given separately on the New England 
records. 
