Desirable Agricultural Experiments. 261 
of trials should be made simultaneously by different associations 
— the more the better — under identical conditions. 
The first trial would be arranged to determine which of two 
breeds would prove the more profitable in beef-making, and the 
two I would propose are Herefords and Dexters. The plan 
would be to have ten good representatives of each breed to feed 
from weaning time until they were three years old. But as it . 
is not easy to tell how calves will turn out, and as some might 
die early, it would be advisable to start with fifteen calves of 
each breed, the best that good judges of each can select. At 
the end of the first six months any balance over ten in each lot 
could be taken out of the trial, the least promising being with- 
drawn. If one or two died afterwards, the loss would not 
materially affect the trial, as average results would supply the 
tests upon which a decision would rest. All the animals should 
have the same kinds of food, and as much as they could 
consume with advantage, all food, of course, being weighed, and 
the weights recorded. Unless health demanded a run on pasture 
in the summer, it would be best to keep the animals in yards 
when young, tying them up or putting them in boxes for the 
last few months of the trial. Both breeds being ordinarily 
grazing ones, a trial under such conditions would be no more 
unnatural to one than to the other, and it would be amenable to 
more exact conditions than if a run on pasture were allowed. 
But if it should be decided that a run would be desirable, the 
two lots should graze on different parts of the same pasture, care 
being taken to secure equality of grazing advantages, and the 
allowance to each of only a comparatively small division at a 
time, so that the difference of space fed off to the same degree of 
bareness could be ascertained. In this case the acreage of grass 
consumed, valued at the same rate for both breeds, would be 
charged against each. Roots and other food grown on the 
farm would, of course, be charged at fixed rates, and purchased 
food at cost price. The live weight of each animal should be 
taken and recorded at the end of every month, and the dead 
weight with all desirable details at the finish. The value of the 
carcass, loose fat, and offal would also be ascertained. It would 
then be possible to state the average expenses and returns for 
each lot, and to see which had proved the more profitable. 
Valuable information, too, would be obtained as to the difference 
in the consumption of food by good beef-making representatives 
of the large and small breeds of cattle. 
One such trial would be of comparatively little value, as it 
could not be regarded as decisive. But if six, or still better, 
ten such trials were carried on simultaneously in different 
