N THE fall of 1936 the first two authors began a series of experiments I to determine the frequency of minute inversions in material treated by X-rays. Before long it was realized that the frequency of such inversions is very low and that a reliable analysis would involve such technical difficulties that it would require a great deal more time than was available to the authors. Therefore, the original plan was revised and a new experiment started to determine the relation between X-ray dosage and the frequency of chromosomal breaks, the distribution of breaks along the chromosomes, and related problems. Subsequently when the experimental part of the problem was well under way the third author joined in the work.The first author is responsible for the cytological analysis of the major portion of the female material, the second author was in charge of breeding and X-raying work and helped with the cytological analysis, and the third author is responsible for the cytological analysis of all male material and of a portion of the female data where Swedish-b females were mated with treated Oregon-R males. Permanent preparations of 1765 pairs of salivary glands were used for this study. Slides were prepared by Mr. HERSCHEL ROMAN, Miss RUTH BATE, and Miss EUNICE WHITE, to whom the authors wish to express their appreciation.When this manuscript was ready for press a paper by CATCHESIDE (1938) appeared describing some of his results obtained in experiments identical with ours. Many conclusions reached by CATCHESIDE are identical with our conclusions.
MATERIAL AND METHODSThe bulk of the data reported here was obtained from experiments in which inbred Oregon-R wild type stock was used. However, before this experiment was completed a female sterility factor appeared in the stock and to avoid difficulties connected with sterility, females of an inbred Swedish-b wild type line were used in later experiments. In the chart (fig. I) showing the relationship between X-ray dosage and effects, the data from Oregon-R females are presented separately because differences may exist in the genetic constitution of the females used. In analyses of other problems such differences would probably be unimportant and,