The Analysis of Covariance

THE WHOLE OF this discussion is based on the data of a single experiment, the details of which have been published under the title Effect of Atropine and Quinidine Sulphate on Atrophy and Fibrillation in Deniervated Skeletal Muscle. 1 For the present, we may adopt the view that the experiment was conducted to compare the effects of four different drugs in delaying the atrophy of denervated muscles. The structure of the experiment was, briefly, as follows. A number of rats were put randomly into four groups; a certain muscle in the hind leg of each rat was deprived of its nerve supply by severing the appropriate nerves, the leg to be denervated (right or left) being chosen at random. Each of the four groups was -assigned to treatment by one of the drugs and the treatments were continued throughout the course of the experiment. Four days after treatment was begun, four rats were chosen randomly from each of the four groups and measures of atrophy were obtained from them. This procedure was repeated after 8 and 12 days. Atrophy is usually measured by the loss of weight of the muscle after denervation. The weight of the muscle can be determined only by killing the animal and removing and weighing the muscle; hence the weight of the muscle at-the beginning of the experiment is not known and a direct measure of weight loss is not obtainable. The device ordinarily used to circumvent this difficulty is to obtain the weight of the same muscle from the other leg (which was not denervated), at the same time as the denervated muscle is taken, and to assume that the weight of the intact muscle, at the end of the experiment, is the same as that of the denervated muscle at the beginning of the experiment.