272     Effect of Certain Drugs etc. on Helminthiasis of Ruminants

same. These two conditions were not however observed in the 1934-35 experi-
ments. In the next two years the animals were so selected that the differences
fn initial weights were not very much and the grouping of animals for giving
several feeds was done in such a way that the total weights of all the groups were
almost equal. In the case of calves only, the experiments continued success-
fully as there were ery few deaths. In the case of sheep, however, there were
a number of deaths during 1935-36 and 1936-37, so that the final weights of
only a iew animals could be recorded.

Calves.—In 1934-35, the percentages of increase in live-weight varied
from 94.8 in the case of balanced ration to such a low figure as 0 • 7 in the case
of grass Adjoining Deg Nullah. The increase due to balanced ration is signifi-
cantly greater than the rest, while the differences between increases amongst
other feeds are not significant. Again, in the year 1935-36 the same results
were obtained. In the final year, however, there was a decrease in weight on
the average in the Deg Nullah and Adjoining Deg Nullah groups. The increase
due to the balanced ration was significantly greater than the rest. The
difference between normal area grass and Adjoining Deg Nullah was also
significant. This is not, however, of much importance as the Adjoining Deg
Nullah grass actually reduced the weight.

Sheep.—As mentioned already the data of 1934-35 only could be analysed
in the case of sheep. The Deg Nullah grass has decreased the live-weight.
The only significant difference is that between balanced ration and Deg Nullah
grass.

On the whole the balanced ration only appears to increase the live-weight
significantly. The normal area grass stands next, with the grass from Deg
Nullah and Adjoining Deg Nullah having a tendency to reduce the live-weight
in certain cases.

                                    CORRELATIONS

An attempt was made to find out whether there is any correlation between
the different factors, viz., protein consumption, mineral intake, increase in live-
weight and parasitic counts in spite of the fact that the figures of protein con-
sumption and the mineral intake are available only for the group as a whole
and not for individual animals. The correlations were, therefore, worked out
with these figures and the average figures of each group in the case of live-
weight and parasitic counts.

It may be mentioned that huge variations exist in the live-weight and
parasitic counts within animals in the same group, though the number in each
group is only four or five and sometimes only two. Finer methods of correla-
tional analysis were, therefore, not adopted. In working out the correlations
the three years' data for each set of two characters was treated as a whole.
The variation due to years and feeds were eliminated and the correlation was
calculated from the residual variation only.