18 THOS. H. MONTGOMERY jr., 



The raetaphase (Figs. 38 — 43) shows the two plates of daughter 

 chromosomes gradually separating from one another. As they become 

 separated one notes between corresponding daughter chromosomes 

 thick achromatic fibres (Figs. 41 — 45), which are true connective fibres, 

 derivatives of the chromatin contained in the chromosomes. Each of 

 these fibres appears doubly-contoured on optical section, as if each 

 were made up of two separate fibres. But this appearance is illusory, 

 careful focussing of the microscope showing that each such fibre is 

 in reality a hollow cylinder of linin, so that when a chromosome 

 divides its linin becomes drawn out in the form of a hollow tube 

 between the daughter chromosomes. The undivided chromosome is 

 so dense in structure that one cannot determine in what manner the 

 linin is arranged in it; but the preceding observations show that 

 very probably a great part of it at least forms a sheath around the 

 chromosome. 



From the stage of the monaster up to the dyaster the distance 

 between the centrosomes increases (compare Figs. 30, 31 with Figs. 41 

 — 44), and this is due in great probability, in accord with many recent 

 observations, to the extension of the central spindle fibres. This could 

 not be directly observed, however, since these fibres are obscured by 

 the thick connective fibres. This increasing separation of the centro- 

 somes is not clearly shown in all my figures, for there are drawn 

 here cells which possibly belong to two different generations, since in 

 them the spindles at the same stages may be of different lengths. 

 But the preceding conclusion is proved by the facts: 1) that in one 

 and the same spermatocyst, where spermatogonia of the same generation 

 occur, the spindles of the dyaster are longer than those of the monaster ; 

 and 2) that the longest spindles found in the dyaster stage are of 

 greater length than the longest of the monaster. The separation of 

 the centrosomes continues until they have nearly or quite reached 

 the cell membrane. I never noted any indentation of this membrane 

 above a centrosome. In all stages the pole rays are very feebly 

 developed, though in the spermatocytic divisions they are seen with 

 great distinctness. The metakinesis is of very short duration. 



At this stage of the metaphase each daughter chromosomal plate, 

 when seen from the pole, shows clearly 14 chromosomes (Figs. 46, 

 47), so grouped around the centrosome as to completely hide the 

 latter. The chromosomes next become more widely separated from 

 one another, with the result that the connective fibres converge from 

 them towards the equator of the cell (Fig. 45). The constriction of 



