475 



on the periphery. In fig. 1 rf is shown the anaphase condilion, 

 the partners of the gemini reaching the poles and the univalents 

 dividlng. Apparently there is, subsequent lo the start of the gemini 

 partners toward the poles, no congrcgation of the single chromo- 

 somes in the equatorial region preparatory to their division. On 

 the other hand, they seem lo retain throughout the same relative 

 position which they occupy when first they appear on the spindle. 



It would appear from fig. 1 rf that ralher widely varying nnm- 

 bers of chromosomes would be represented in the daughler nuclei. 

 Thns onc might suppose that some of tlie undivided univalents 

 near the equatorial region would divide too late to permit their 

 daughler elements to reach the poles before the rounding uj) of 

 the daughler nuclei or that both halves of univalent chromosomes 

 dividing near one pole w'ould be included in one daughler nucleus. 

 Allhough, wilh the material at hand, the fåle of the single chro- 

 mosomes or their products is not entirely clear, it seems probable 

 that the latter situation occurs. On the other hand, it is certainly 

 Iruc that no conspicuous number of chromosomes is left behind 

 in the plasma and indecd at interkinesis many P. M. C. show no 

 such chromosomes and olh.crs only one or two. The number of 

 chromosomes appearing in a homotypic metaphase is rarely less 

 than ninetecn or twenty. 



Fig. 2 a shows a characterislic condilion following interkinesis, 

 Ihe significance of which is not entirely clear. It should be noted 

 that, allhough widely distribuled, the chromosomes all lie in ap- 

 proximately the same plane within the group and are ralher 

 strikingly large. Il is provisionally assumed that such stages, and 

 they are very numerous, rcsull from abnormal growlh Of daughter 

 nuclei during interkinesis, the area occupied by the widely distri- 

 buled chromosome groups (fig. 2 a) being the same as that occupied 

 by abnormal ly large daughter nuclei of ten to be observed. Since 

 diads are formed among the teträds and since intermediate stages 

 have been observed, it is suggested that such P. M. C. nndergo 

 no furlher divisions and represent early stages of diad formation. 



In contrast to the "abnormal" condilion described above, "normal" 

 interkinesis stages give rise to typical homotypic metaphases and 

 to the anaphase condilion shown in fig. 2 b. The condilion 

 illustraled is characterislic of the more "normal" spindles observed. 

 In many cases the homotypic spindles are highly irregular and in 

 some cases the two nnite to produce a monstrous but regularly 



