STUDIES ON THE DETERMINATION OF DIGESTIBILITY
                                COEFFICIENTS :

    II. THE ESTIMATION AND COMPUTATION OF DIGES-
        TIBILITY COEFFICIENTS FROM INDIVIDUAL TESTS
        AND THEIR ORDER OF PRECISION

                                           BY

                M. CARBERY, D.S.O., M.C., M.A, B.Sc,

          Agricultural Chemist to the Government of Bengal,

                                         AND

               INDU BHUSHAN CHATTERJEE, L.AG.,

                         Physiological Chemist, Bengal.*

               (Received for publication on 30th July 1935).

                                    INTRODUCTION

In a recent paper by Carbery, Chatterjee and Hye [1934]. a method of experi-
mentation and computation has been described, which is likely to ensure a greater
degree of accuracy in arriving at the digestibility values of individual feeds and
their components in a mixed ration. The chief point in favour of this method
is that it enables direct estimation of the digestibilities without having recourse
either to the use of assumed values (from standard books) for one or other items
of the mixed feed, or to the alternative of conducting the trial with a single feed
in the first instance, followed by the combined feed.

The method developed by the above authors is the result of an elaborate
test involving eighteen individual estimations conducted in a cyclic order on
six animals in a restricted randomised distribution. It need hardly be stated
that the larger the number of tests and animals, the greater is the accuracy and
representative character of the mean values. At the same time it cannot be
overlooked that an increasing number of tests implies increasing outfits in staff,
material and expenses. Thus, in many cases, it is not possible to have a large
number of tests owing to shortage of men, money and materials.

On such an occasion a shorter method can be the only possible way to meet
such special requirements. The development of a method giving a reasonable
order of accuracy in such a contingency will be of great practical value. Naturally
since the initiation of original tests this aspect engaged the close attention of the
authors, and in the present paper a partial solution of this complicated problem

* Under the grant of the Imperial Council of Agricultural Research, India.

                                                (87)