4^ 



98 DUGGAR : DEVELOrMENT OF THE POELEX GrAIN 



axis of the spindle is finally determined ; so that it becomes multi- 

 polar in one general plane, and then by a gradual contraction pro- 

 cess it becomes bipolar. 



The chromosomes seem to increase in size just before the dis- 

 appearance of the nuclear membrane. On this entrance of the 

 kinoplasmic fibers there is no indication of the nucleolus, and I 

 have no evidence of the manner of its disappearance. 



During the later stages of spindle formation the chromosomes 

 seem to change their form, possibly the splitting having already 

 begun. When they are finally brought into an equatorial region, 

 the alignment is not perfect (Figs. 25, 26), and the chromosomes 

 arc much scattered. This is characteristic of all first divisions in 

 the production of the reduced number of chromosomes. The 

 dense cytoplasm is apparently repelled from the spindle region, 

 and only the poles of the completed spindle reach into the denser 

 zone (see Fig. 25), thus somewhat resembling the condition found 

 in Hcnierocallis fulva,^ In Bignonia^ however, the spindle fibers 

 are very numerous, and they terminate in a more distinct apex. 

 The spindle is very large for the total chromatin mass concerned. 

 In this plant the nuclei arc much larger than in others under study 

 at the same time, but the chromosomes are much smaller. 



Details of reduction phenomena were not followed, since the 

 plants were not suitable for this purpose, but some especially in- 

 teresting nucleolar phenomena were observed. 



After the separation into the daughter segments the chromo- 

 somes show a very slight indication of a V-form, by short projec- 

 tions pointing towards the equator, as in Fig. 27. They pass to 

 the poles in scattered and unaligned array, so that at this stage it 

 is relatively an easy matter to estimate the number of chromo- 

 somes. From several counts I have concluded that there arc 

 twenty five, although twenty four and twenty six have been 

 counted. In the very thin sections necessary for Ihe study of 

 these divisions the knife usually passes through some of the chro- 

 mosomes, and from this results the only difficulty in counting 

 them accurately. 



^■^ Jucl, II. O. Dir Kenulu'ilunj^r in den rollenmntterzt'llen von Ilctncrocallis fuh'a 

 iind die l)ci denscll^en aiiftretendcn I'nregelmfissigkeiten. Jahrb. wiss. l]ot. 30 : 205- 

 226. 1897. 



