196 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 30 Ser. 



These are fastened to the chromosome in two groups, one 

 at each end. All the fibers of the " Centralkorper " finally 

 become directed toward one or another of the poles, whose 

 number is finally reduced to two, probably by fusion. 



It will be seen that the process in Equisetum has but little 

 in common with that in Passiflora, and that the process in 

 Larix, on the other hand, agrees with that in Passijlora in 

 some very essential points. In both Larix and Passijlora 

 the cytoplasm and linin take part in the formation of a 

 central network which is pulled out to form cones, which 

 later become composed of free fibers. The cones then 

 fuse to form the bipolar spindle. Various details, how- 

 ever, are different in the two. The peculiar arrange- 

 ment of the cytoplasm showing a zone around the nucleus 

 (fig. 2) in which the meshes are drawn out parallel to the 

 nuclear wall recalls Belajeff's description, but, as in 

 Equisetum, is of short duration and it is difficult to say 

 what part, if any, it plays in spindle-formation. The same 

 arrangement has been described by Strasburger (1895) for 

 Larix. He finds it to have about the same connection with 

 the formation of the spindle as does Belajeff. Other 

 observers have described a similar arrangement, but do 

 not explain it. 



The radial arrangement of the cytoplasmic network (fig. 

 3) corresponds very closely with Belajeff's description of 

 the state of the cytoplasm of Larix before the meshes begin 

 to be drawn out parallel to the nuclear wall. In Passijlora, 

 however, the radial arrangement occurs after this stage. 

 The radial arrangement described by Guignard (1891), 

 Farmer (1893 and 1895), Osterhout (1897), and Mottier 

 (1897a) seems to be due to the presence of free fibers, not 

 to a network. 



Belajeff speaks of fibers which extend out from the cen- 

 tral network to the cell-wall; these run radially or tangen- 

 tially. The appearance of fig. 4 certainly seems to indicate 

 that there is a similar condition here. Some of the fibers 

 can be traced to the cell-wall, but there is no evidence 

 that they pull the central network out into cones as they 



