88     THE INDIAN JOURNAL OF VETERINARY SCIENCE AND ANIMAL HUSBANDRY [VI, I

has been attempted by directing attention to two main lines of approach, viz.,
(1) the method of computation and (2) the minimum number of tests necessary
for a certain specified standard of accuracy.

                    METHOD OF COMPUTATION

With respect to the method of computation a mathematical formula can
be satisfactorily worked out, in fact, has been done. It is based, of course, on
the assumption that the behaviour of the animals is uniform. This assumption
is implicit in all digestion tests and has been made by all investigators. If the
animals were perfectly uniform, the data obtained from a test performed on one
single animal would be quite sufficient to work on this formula.

In such a case the equation can be set up on the following basis. If R and C
represent the amount of roughage and concentrate actually consumed and if D
be the amount digested from the combined feed, then in a course of two digestion
trials we get the following simultaneous equations :—

where x and y are the digestibility values (unitary basis) of R and C, and are
assumed to be constant.

By solving these two equations we get:—

x = D₁ C₂— D₂C₁ / R₁ C₂— R₂C₁ . . (3)

y = D₂ R₁— D₁ R₂ / R₁ C₂— R₂ C₁ . . (4)

Similarly, if there are three feeds, say R, C and F and if x, y and z be their digesti-
bility coefficients (unitary basis) then in the course of three digestion trials, we
obtain the following three equations :—

By solving these we obtain the following values which for convenience are shown
in terms of determinants :—