INDUCED CHROMOSOMAL ABERRATIONS IN ANIMALS 1181 



attachments are sometimes transferred from the middle to one of the ends 

 of the genetic chromosome. Individuals carrying these inversions were 

 found cytologically to possess very long rod-shaped, or J-shaped, second 

 chromosomes instead of the normal V-shaped ones (Van Atta, 136, 137; 

 Schultz and Dobzhansky, 116). Since the shape of a chromosome at the 

 equatorial plate stage is determined by the location of the attachment 

 constriction (the latter being directed toward the center of the plate, and 

 the free end of the chromosome toward the periphery), these results 

 suggest that the spindle attachment is a permanent feature of the 

 organization of chromosomes. 



Still more conclusive evidence in favor of the permanence of the 

 spindle attachments is afforded by comparative studies on various 

 translocations. Chromosomes are fragmented by X-rays; some of the 

 fragments may remain free and others may reattach to different chromo- 

 somes. The question is, then, whether a fragment not including the 

 spindle attachment of the old chromosome can remain free and develop 

 a spindle attachment de novo, or whether two fragments both possessing 

 a spindle attachment can unite and form a new chromosome with two 

 attachments. This question may be answered in the negative, since no 

 such cases have been encountered, in spite of the large number of trans- 

 locations analyzed in detail. The fragments remaining free are invariably 

 those which include the spindle attachment of the old chromosomes, and 

 the fragments that become attached to other chromosomes are fiberless. 

 The only possible conclusion is, therefore, that fiberless fragments and 

 compounds having more than one spindle attachment are eliminated 

 because of their inability to behave normally at mitosis. 



This conclusion seems to be contradictory to a series of facts dealing 

 with compounding and fragmentation of chromosomes. Can two or 

 more separate chromosomes unite permanently into one, and can a 

 chromosome be broken into several separate chromosomes? It was 

 assumed for a long time that such changes take place. Even disregarding 

 the temporary unions and fragmentations of chromosomes observed in 

 Ascaris and in the gametogenesis of certain insects (which may be an 

 entirely different phenomenon), changes in chromosome numbers, and 

 consequently also in the number of spindle attachments, undoubtedly 

 take place in phylogeny. Even within the genus Drosophila the chromo- 

 some number varies from three to six pairs (there being no evidence of 

 reduplications and los.ses of whole chromosomes), and .similar differences 

 are observed between races oi Phragmatohia (Seller), in certain Orthoptera 

 (McClung), and in other species of animals and plants. 



It should be pointed out that the evaluation of evidence derived from 

 comparison of races and species of unknown origin is always a matter of 

 contention. An exact study of the situation can be carried through 

 only in cases when a comparative genetico-cytological analysis of the 



