﻿266 BOTANICAL GAZETTE [april 



some results. At the time of or even before the formation of 

 the spindle, the daughter segments contract to such an extent 

 that each may be in the form of a short, thick crescent. Each 

 ring-like chromosome will therefore consist of two crescents 

 whose ends adhere {^fig- Jo) . It not infrequently happens that 

 the daughter segments contract in such a manner that closed 

 rings are formed. If the daughter segments remain adhering at 

 one end, each chromosome in the nuclear plate is seen to con- 

 sist of two thick pieces in the form of either a short crescent, a 

 broad U, or only slightly bent at one or both ends (Jig Ji, on 

 the left). Under these circumstances the whole chromosome is 

 relatively long, reaching almost from pole to pole. 



As a rule the chromosomes are oriented tangentially upon 

 the spindle, the fibers being fastened at the middle of each seg- 

 ment or near one end. It is interesting to note that the place 

 to which the fibers are attached Is almost always in the form of a 

 small protuberance, giving the impression that a pull is exerted 

 by the spindle fibers. In fig. jo is shown the singular phenome- 

 non of two chromosomes fastened together, a fact which is of 

 course without any special significance. 



No indication of a second longitudinal splitting is seen in 

 Tradescantia when the chromosomes are in the nuclear plate, 

 and, so far as could be determined with certainty, not even dur- 

 ing metakinesis, although it may occur here as early as in Lilium 

 ifg^' 3^y 3^) ; but by the time the daughter chromosomes have 

 traversed one-half the distance toward the poles, it can be usually 

 seen that each consists of a pair of granddaughter segments, 

 arising from the second longitudinal splitting {fig- 33 



Having arrived at the poles, the pairs of granddaughter seg- 

 ments become somewhat crowded together, and shortly the 

 nuclear membrane is laid down [fig. Ji). The segments, or 

 some of them at least, now separate slightly ; but as they begin 

 to reticulate, becoming lumpy and very irregular, their identity 

 is soon lost. The process of reticulation continues, and at a lit- 

 tle later stage the lumpy chromatin masses which are connected 

 by numerous fine threads present the appearance of fig- 3^- 

 As in many other plants, there is not an even and uniform 



