1905.] NATURAL SCIENCES OF PHILADELPHIA. 179 



the case when its univalent parts are placed in one line. Sometimes, 

 as with the extreme right and left ones of fig. 57, the long axis of the 

 chromosome appears to lie in the equatorial plane ; this results also in 

 a reduction division, however, because here there is a bent instead of a 

 straight bivalent chromosome, with consequent convergent disposition 

 of the two univalent chromosomes. Lycosa is particularly demon- 

 strative of this first mitosis being a reduction mitosis, on account of the 

 simple form of the chromosomes and of the occasional perceptible per- 

 sistence of the longitudinal split at this stage. No chromosomal rings 

 occur at this stage; the nearest approach to them are oval forms like 

 the two largest in fig. 55, with very exceptional width of the longitudinal 

 split; such forms are individual variations, not found in every cell, 

 as one sees on comparison with pole views (figs. 53, 54) showing all the 

 chromosomes, yet even in them the original long axis of the chromo- 

 some is recognizable. 



In metakinesis (fig. 59) all the bivalent chromosomes undergo a 

 reductional halving. Figs. 60-63 show pole views of the chromosomal 

 plates of the daughter cells, second spermatocytes. Disregarding the 

 two minute bodies (S.) of 62 and 63, we find in 60 fifteen chromosomes, 

 in 63 fourteen, in 61 thirteen, in 62 twelve. There would then seem 

 to be a range in number from twelve to fifteen. This I believe is due 

 rather to some unexplained individual variation than to the possibliity 

 of a normal unequal distribution of the chromosomes. For in the nine 

 cases where they could be easily counted the numbers fifteen, fourteen 

 and twelve were each represented by only one case, whereas thirteen 

 appeared in six cases; and in the only two cases where the chromo- 

 somes of the second spermatocyte could be counted on lateral view 

 (one of these shown in fig. 65) there were in both cases thirteen chromo- 

 somes. There were thirteen bivalent chromosomes in the first sper- 

 matocyte, and the counts show that in the majority of cases, so prob- 

 ably as the normal phenomenon, there are thirteen univalent ones in 

 the second spermatocyte. As the chromosomes of the first spermato- 

 cyte separate in the anaphase (fig. 59) each daughter chromosome 

 shows a divergent split widest at the equatorial end ; this can be noth- 

 ing else than the reopening of the original longitudinal split, if one com- 

 pares the appearances in fig. 59 with the chromosome most to the left 

 in fig. 55. So each bent chromosome of the second spermatocytes 

 (figs. 60-65) is a univalent chromosome so split longitudinally that the 

 cleft is narrow at one end and widens out toward the other. There 

 is no proof of any kind that this is either a transverse break or a line of 

 separation between whole univalent chromosomes. 



