BOT.— Vol. II.] HUS— CASSIA TOMENTOSA L. 337 



time after the nuclear wall has disappeared. Only when 

 the multipolar spindle has been fully formed, and in fact 

 not until a rearrangement of the poles (with a view to the 

 formation of a bipolar spindle) has begun, do they begin 

 their migration toward the spot where the equatorial plate is 

 to be formed. Throughout the formation of the cones and 

 the subsequent changes from multipolar to bipolar spindle, 

 no indication of a centrosome could be observed. 



As soon as the fibers composing the cones have become 

 smooth, the nuclear wall breaks down and the blue-staining 

 fibers penetrate the nuclear cavity, where they lie in contact 

 with the bright red chromosomes. The multipolar spindle 

 thus formed has a variable number of cones; the largest 

 number observed in a single section was eight (fig. 10) ; 

 but as the cell was always cut into a number of sections, 

 the number of cones must be greater. 



The granular zone now presents a uniform appearance, 

 and stains from yellow-brown to brown-violet; the granular 

 threads which were formerly so prominent have disappeared ; 

 nothing remains but small meshes which are fairly uniform 

 in size. Between and upon the threads composing the 

 meshes a large number of granules can be observed. The 

 poles protrude into the granular zone. Surrounding the 

 granular zone we find an irregular line still staining a deep 

 blue. The cytoplasm outside this line retains the same 

 structure as in the last stage. 



In the formation of the multipolar spindle a highly inter- 

 esting phenomenon was observed. It has been stated 

 before that, as a rule, a cone situated at one end of the 

 nucleus had from the first an ascendency over the others ; 

 and while it was not observed that another cone (situated 

 directly opposite this one) was always the second to be 

 formed, yet one could not fail to see that after a little while 

 two cones (one at each end of the nucleus) were more 

 prominent than the others, or were even the only cones 

 present, in which case the nuclear wall disappeared beneath 

 these cones while persisting everywhere else. Such a case 

 is shown in figure 8. " In nearly every instance this promi- 



