Receive (d fot 1 publicatiol August 12, 196.5 IN a previous paper (Fare, 1964) it was found that there was a significant difference between boxes of mice given identical carcinogenic treatment as regards the times at which the animals first developed tumours. It was presumed that since the tumour induction time for a particular mouse was influenced by which box it happened to live in, then different tumour incidences in a given population would result in a fixed time if the number of mice to a box was varied. The experiments described here were carried out to investigate this suggestion.
MIATERIALS AND METHODSMice.-The albino mice used were from our closed stock colonv which has been used for carcinogenicity tests on a wide range of compounds including polycyclic lhydrocarbons (Woodhouse, 1959), 4-nitroquinoline-N-oxide (Searle andWoodhouse, 1964). various anminoazo dyes (Fare, unpublished work), beta-propiolactone (Searle. 1961), tobacco tar (Hamer and Woodhouse, 1956) and mineral oils (Hieger and Woodhouse, 1952). The results have generally been reproducible. anid in agreement with the published results of other workers where applicable.On rare occasions. it has been suspected in the past that the incideicees of tumours have been non-random, but attempts to investigate this statistically by an analysis of the records have been unsatisfactory, largely because of limitations imposed by large numbers of early deaths and enforced killings during the experiments. The treatment used to produce tumours in the present experiments is un-ique in our experience in that a one hundred per cent tumour vield results in our mice in a relatively short time, before those mice whiclh develop tumours at an early stage have to be killed.Thlree groups of 30 mice wrere actually used in the tests, but to alloM for possible early deaths during the " equilibration " period before treatment was started, 1l() were taken initially. They were all males. to obviate difficulties due to breast tumours, both spontaneous and treatment-induced, which occur in a high proportion of female mice of this stock in long-term experiments. Even during the shortterm experiments described here, some breast tumours would be anticipated.The mice, aged 6-8 weeks, were selected by hand, 5 into each of 22 numbered boxes. The mice in any one box were identified by colour markings whichl was considered to be a more uniform procedure than ear-clipping, when it is inevitable that the degree of trauma varies from mouse to mouse.Twenty-two bottle caps were numbered from 1 to 22 and placed in a box. Five glass marbles, coloured similarly to the dyes used to mark the mice, were placed in another.