﻿2 6o • ' BOTANICAL GAZETTE [a?ril 



i 



somes are arranged both radially and tangentially. The two 

 elements of each chromosome may lie parallel with each other, 

 or they may remain almost entirely separate, being in contact 

 only at the ends that are attached to the spindle fibers i^fig. 14). 

 This seems to be more frequently the case in Lilhim candidiim 

 and Fritillaria. In the latter arrangement both elements may 

 lie upon the same side or upon opposite sides of the equatorial 

 plane. The chromosomes are fastened to the spindle fibers at 

 or near the ends, rarely near the middle. When they are fas- 

 tened near the ends, these are generally bent somewhat so that 

 the retreating segments appear in the form of rods which are 

 slightly hooked at the ends turned toward the poles [^fig- ^7)- 



Strasburger (1900), Guignard, ('99), and Gregoire ('99) 

 claim that the two segments of each chromosome in the second 

 division correspond exactly to the pairs of granddaughter seg- 

 ments which form the V's or U's of the first mitosis, and that 

 they are fastened to the spindle at the places of bending or at 

 the angles of the V's as in the first mitosis. Theoretically there 

 may be little objection to this statement, as the matter in ques- 

 tion may be a detail of secondary importance, but it seems that 

 the observed facts will not admit of its general application in 

 Lilium. Figs. 13,14, and 75 show clearly that the free ends of 

 the loops or V's are as likely to be attached to the spindle. 

 Furthermore, we cannot say with certainty that the chromosomes 

 of the second mitosis are identical with the pairs of granddaugh- 

 ter elements of the first division, for it does not seem impossi- 

 ble that the spirem may segment in such a manner that the 

 segments of each chromosome of the second division may be 

 derived from different pairs of granddaughter elements. 



This may be brought about by the spirem segmenting at the 

 apex of the V's rather than at the points corresponding to the 

 free ends of the elements of each V, assuming that each long 

 loop of the spirem represents a V. Under such circumstances, 

 if all the granddaughter elements composing the spirem were so 

 oriented in the preceding anaphase as to form V's, then all of 

 the chromosomes of the second division would be made up of 

 elements from different pairs of sister segments. But on the 



