MATURATION 153 



making up the chromosome are usually very obscure, particu- 

 larly when synizesis is pronounced. 



This whole subject is in a rather more hypothetical state 

 than one might wish, considering the importance of the con 

 elusions to be drawn. Upon the assumption that the qualities 

 of a chromosome differ from end to end, a longitudinal fission 

 of the chromosome would divide it into exactly similar halves, 

 while a transverse fission would of course divide it dissimilarly. 

 In many forms it is clear that the first or heterotype division 

 is longitudinal and the second or homotype division is trans- 

 verse so that the four resulting nuclei are of two categories. In 

 other cases it appears superficially that both divisions are 

 longitudinal, while in still others it is really impossible to say 

 definitely whether a given division is longitudinal or transverse. 

 A transverse division would be reducing, however, only upon 

 the assumption of an end to end differentiation of the chromo- 

 some, and upon the further assumption, which should be clearly 

 apprehended, that no rearrangement of the chromatin granules 

 occurs during the maturation divisions. And the additional 

 fact must be taken into consideration that the chromosomes 

 of the primary spermatocyte or oocyte are bivalent, that they 

 represent two chromosomes, which as wholes have fused in 

 either parasynapsis or telosynapsis, and the actual chromatin 

 granules composing them may have an arrangement in fusion 

 which is not indicated by the behavior of the whole chromosome. 

 The question whether a given longitudinal or transverse 

 division of a chromosome is equational or reducing then can be 

 determined only by taking all of these preliminary arrange- 

 ments into account, and in most cases this is extremely difficult 

 or even at present impossible. The small size of the chromo- 

 somes themselves and the minuteness, often invisibility, of the 

 chromatin granules, often put the facts of their arrangement 

 and behavior beyond the possibility of observation, and we 

 can only infer their arrangement and history from subsequent 

 events. In most cases we know only that reduction occurs. 

 And from the few cases in which the course of events seems 

 clear, we infer that in all maturation divisions, one is equational, 



