levine: somatic and reduction divisions in drosera 133 



By slow changes the spireme begins to contract and move toward 

 the periphery of the nucleus and the synaptic stage is formed. 

 All stages may be found in this contraction of the spireme. An- 

 thers in the same flower are found in different stages other than 

 synapsis. I cannot agree with Lawson ('11) that the synaptic 

 knot is not an active contraction of the chromatin thread. The 

 coils of the spireme can be readily traced in the early stages but as 

 the knot becomes smaller and smaller the difficulty is increased 

 but at no time is it impossible to trace the individual coils (Figs. 

 17-18). The nucleole is generally enclosed in the meshes of the 

 synaptic knot; that also helps to make the winding of the band 

 more difficult to follow. 



As the chromatin mass emerges from the synaptic knot there 

 is a simultaneous shortening of the spireme and the pachyneme 

 spireme is formed. Drosera filiformis is particularly favorable 

 material for the study of this stage (see Figs. 29a, 2gb). When 

 the pachyneme is fully developed it may be seen to consist of a 

 number of loops which are evenly distributed through the nuclear 

 cavity. The band is homogeneous and while it occasionally shows 

 signs of a beaded structure it is as a rule very smooth. The 

 chromatin material of the spireme next seems to accumulate in 

 certain areas (FiG. 30) leaving clear spaces in the band between 

 them. This stage is followed by further contraction and thicken- 

 ing and finally the breaking of the spireme into a number of seg- 

 ments as shown in Fig. 31. In Figs. 19, 32a, b, c, we have the 

 representation of nuclei in diakinesis. The ten bivalent chromo- 

 somes are peripherally distributed. They are somewhat angular 

 and approximately all the same size. Occasionally the chromo- 

 somes appear quadripartite (FiG. 9). In the embryo sac mother- 

 cell of Drosera longifolia I have been able to observe the breaking 

 up of the pachyneme (FiGS. 40a, h) spireme. The cells are 

 large and the nuclei in sections 5 /x thick may be cut twice. 

 I have been unable at this stage also to determine the number of 

 chromosomes but I hope that in the coming summer Avith more 

 material this question can be definitely settled. 



The nucleole in this stage stains a vivid ruby red and a number 

 of vacuolar areas are visible. The nucleole has in no way changed 

 except that it has increased in size. I must agree with Martins 

 Mano ('05) that there is no evidence here for the contention of 



