660 RADIATION BIOLOGY 



produce a so-called " isochromatid break" (Sax, 1941; Swanson, 1943; 

 Catcheside, Lea, and Thoday, 1946b; Catcheside, 1948; Bishop, 1950). 

 In the grasshopper, Chortophaga, isochromatid breaks were observed by 

 Creighton (1941) in cells that presumably were irradiated in late prophase 

 stages. She suggested that in these cases breakage was probably attrib- 

 utable to an ion cluster rather than to a single ionization. The present 

 author (unpublished data) has confirmed by direct observation on living 

 neuroblasts of Chortophaga the fact that isochromatid breaks can be 

 produced by irradiation at late prophase stages when the chromosomes are 

 visibly separated into chromatids (Fig. 9-3, column 2). Such evidence 

 provides no support for the contention of Darlington and Koller (1947) 

 that isochromatid breaks are "fictitious." Accepting the validity of 

 their occurrence — and there are many other reasons for doing so, as 

 Catcheside (1948) has indicated — it follows, as Lea (1946) has empha- 

 sized, that evidence obtained from X-ray studies that a chromosome is 

 unsplit is far from conclusive. 



Some of the rearrangements detected in Drosophila by salivary-gland-, 

 chromosome analysis have involved both chromosome and chromatid 

 breaks in the same chromosome. A typical rearrangement (Fig. 9-1 lc) 

 included two chromosome breaks, one proximal and the other distal, and 

 in the region between them two chromatid breaks, one in each of the 

 two sister strands. On the assumption that chromatid and chromosome 

 breaks reveal whether a chromosome has or has not divided, the chromo- 

 somes of the spermatozoon at the time of irradiation in this case would 

 appear to have been in the process of division. The condensed, inactive 

 condition of the chromosomes of the sperm head, however, makes division 

 at this stage highly improbable. On the assumption that the chromo- 

 some was single at the time of treatment, potential breaks were induced 

 at the four loci indicated, and were duplicated subsequently in each sister 

 chromatid. Restitution must then have occurred in each of the two 

 strands at that locus at which a break is detectable only in the sister 

 strand. Under such conditions a higher proportion of duplications might 

 be expected than the analysis of a large group of induced alterations has 

 indicated, for the reason that broken ends in both sister strands would be 

 available for recombination. On the assumption that the chromosome 

 was longitudinally double in the mature spermatozoon, either chromatid 

 or isochromatid breaks might be produced, as revealed in the described 

 rearrangement. An unequivocal decision in favor of one or another of 

 these alternatives is complicated by the fact that the chromosomes have 

 not been identified in the spermatozoon, so that no cytological evidence 

 is available concerning the time of division into chromatids. 



Despite this lack of direct evidence, individual rearrangements have 

 been observed that can best be interpreted on the assumption that the 

 chromosomes of the spermatozoon are longitudinally double (Kaufmann, 



