抄録
Animals are thought to be suitable for experiment, when following criteria are fulfilled: 1) The animals should either respond to a given stimulus markedly or remain without reactions. 2) The indications before treatment never show the great individual difference. 3) The indications after treatment never show the great individuality as well.
Firstly, responses to a treatment may be conjectured by Tukey's or Scheffé's method. Secondly, the stability of indications before treatment are given by coefficients of variance. The smaller the coefficient, the more uniform the indication.Finally, the uniformity of the response to a given stimulus is examined with the following formula :
√VB (P) /X
where
VB (P) =unbiased variance among individual animals
X=arithmetic mean
The smaller the value, the uniform the response.
The method of selection of animals for experiment, as well as the determination of its critical parameters and the administered dose, are given as an example of benzene poisoning. And mature rats of Sprague-Dawly strain were found the most suitable animals for the experimental work in benzene poisoning.
