34 
Journal of Agricultural Research 
Vol. VIII, No. 2 
by breeders. Some indication of this is given in Table III. It is obvi¬ 
ously difficult to determine the absolute yield in animals that are not 
milked regularly by hand. The classification shown here is therefore 
based on a relative comparison rather than on absolute yield. This classi¬ 
fication was based on close observation and verified by the actual quan¬ 
tities drawn for the determination of fat. Four grades were established 
to indicate relative yield: (i) High, (2) good, (3) fair, and (4) poor. The 
results shown in the last column of Table III and also by the curves in 
figure 1 bring out clearly the economic significance of high yield. The 
number of ewes represented is fairly large and the results should there- 
PlG. i-—Curves showing rate of increase in weight of lambs (in pounds) by 4-week periods from different 
quantities of ewes’ milk. 
fore possess a fairly representative value as to facts. The difference in 
weight increase between lambs from high-milking ewes and good-milking 
ewes is 16 percent; between high- and fair- milking ewes, 38 percent; and 
between high- and poor- milking ewes the difference is 79 per cent. The 
same results from different yields of milk are also represented graphically 
in figure 1. The curves here show a rise almost directly proportional to 
yield. As shown in 4-week periods, the greatest rate of increase is found 
in the second period for all groups. A slight decline in the rate of growth 
is found in the third period, except in the low yield, and is followed by a 
still further decline in the fourth period. Since all lambs had free access 
to a liberal amount of grain and hay, which they ate greedily as they 
