u 



tation into chromosomes takes place. Gates' investigations 

 on CEnothera^° incline him to the latter view rather than 

 that of the presence and lateral approximation of two 

 separate spirems. Probably single or double threads may 

 characterise different plants (Fig. ii). Contraction and 

 consequent thickening bring the thread or threads into a 

 tangled skein or ball, which takes up a position at one side 

 of the wall surrounding the nuclear cavity (Fig. 12). The 

 nucleolus is usually caught and held within the thread, a 

 little later it becomes free and finally disappears. In this 

 first contraction of the spirem, known as the synapsis stage, 

 one is always sure of dealing with the cell as a spore 

 mother-cell, and not with its earlier form of a sporogenous 

 cell. 



Coming out of synapsis the convoluted thread loosens 

 into a shorter and thicker hollow spirem (Fig. 13), and 

 about this time the first or second longitudinal splitting, as 

 the case may be, takes place. Preceding this division of 

 the thread a bipartation and pairing of the chromatin 

 granules is described, resulting in a double row of chromo- 

 meres running along the linin strands (Fig. 14). The 

 thread now breaks up into segments representing the 

 reduced number of chromosomes characteristic of 

 the gametophyte forms of the plant. At first the 

 chromosomes may be variously twisted and looped, 

 and each such segment seems to consist of two 

 somatic chromosomes of the sporogenous cell that were 

 arranged end to end along the spirem (Fig. 1 5). These are 

 cut off in pairs, which bend round, or otherwise come to lie 

 side by side and form double or " bivalent " chromosomes, 

 thus by their union reducing the number to one-half of that 

 contained in the sporogenous cell. They gradually pass 

 into typically V-shaped forms, move towards the equator of 

 the cell, and there constitute the nuclear plate (Fig. 16). 



