Andrews, Karyokinesis in Magnolia and Liriodendron. 141 



daiighter niiclei divided simultaneously and connecting fibres often 

 extend tbrough tlie isthmus of the undivided motlier cell as figiired 

 by Guig-nard^) (1 c. fig. 28). 



In form the spindle of this division is rather short and 

 naiTOw especially in the latter stages It is also very blunt at 

 the poles (Fig. 7) but assumes more pointed terminations at matu- 

 rity. The heavy band of protoplasm, mentioned above, for the 

 first division is again in evidence, bat likewise rapidly dis- 

 appears. 



In the tapetum cells it was noticed that one, two, three or 

 even more nuclei were often present. The number of these seemed 

 generally greater in those cells nearest the pollen-mother cells. 

 These nuclei multiplied by indirect division since mitotic figures 

 were frequently seen. Again other of these nuclei appeared to 

 be in an apparant State of fusion and of this fusion process all 

 stages could be observed. The same behavior of tapetum nuclei 

 has been noticed by M o 1 1 i e r in the tapetum cells of Podojjhyllum. 



The facts observed in Magnolia and in Liriodendron indicate 

 that: — 



I. In the first mitosis. 



1. The chromosomes arise from the resting nucleus as ir- 

 regulär masses without a previous formation of the usual 

 and uniform spirem. 



2. That the resulting chromosomes are mostly U-shaped 

 though many are in the form of open or closed rings or 

 ellipses. 



3. That they divide here again longitudinally. No longi- 

 tudinal division of the chromosomes was observed during 

 the meta or anaphase of the first division It probably 

 does not occur since the daughter chromosomes of the 

 first division lose their identity during the reconstruction 

 of the daughter nucleus, and as these daughter nuclei 

 pass into the resting condition, it is extremely difficult 

 to conceive of the purpose of a second longitudinal Split- 

 ting of the chromatin during the first mitosis. 



II. In the second mitosis: 



1. The identity of the chromosomes, therefore from the first 

 to the second mitosis is not maintained. 



2. That the chromosomes arise by the segmentation of an 

 irregulär spirem and are at first lumpy bodies. 



3. That the chromosomes arise mostly in the form of shallow 

 U's but in a few cases rings were formed. 



This work was performed in the botanical department of 

 Indiana University U. S. A. and it is my pleasure to express my 

 thanks to Professor Mottier for bis many kind suggestions. 



*) Les centres cin^tiques chez les vcg^taux. (Annales des Sciences 

 Naturelles. Botaniques. Serie VIII. Tome V.) 



