Dixon — On the Chromosomes of Lilium longiflormn. 713 



this view, and as I have also observed the chromosomes of L. chalce- 

 donicum and L. Martagon in the nuclear plate of the succeeding division 

 undergo longitudinal fission, there is additional reason to believe that 

 the process observed in the nucleus of the mother-cell is not a double 

 longitudinal fission. (Figs. 13 «, 13^, 14.) 



Eefore leaving the description of this stage I would like to direct 

 attention to an appearance which frequently presents itself when the 

 nuclear plate is regarded from the poles : the strands of the achromatic 

 spindle (at all events the larger ones) seem to alternate with the 

 chromosomes and do not appear to be attached to them (fig. 7). 



The halving of the chromosomes, which takes place in the equatorial 

 plate, divides each chromosome in a horizonal sense, and it begins at 

 that part of the chromosome which lies towards the axis, while the free 

 ends lying towards the periphery remain in contact with one another for 

 a. short time longer. The horizontal parts of the mother-chromosome, 

 which originally lay towards the centre, become bent abruptly towards 

 the poles, while the peripheral portions of these parts, which are still 

 in contact with one another, remain in. the equatorial plane ; the 

 result is that when the whole chromosome is seen in jjrofile from 

 the equatorial plane it is T-shaped, the vertical bar of the T lying in 

 the equatorial plane with the lower portion towards the peiiphery, 

 while the horizontal bar extends towards the poles. Just before this 

 horizontal cleavage has reached the peripheral end of the chromosome, 

 this end undergoes cleavage along the line of separation between the 

 two rods proceeding from the periphery in a vertical plane, so that the 

 peripheral ends of the rods part asunder from one another horizontally. 

 This cleavage, however, is never complete while the chromosome remains 

 in the equatorial plate, for while at the equator the inner ends of the 

 rods remain united. This double cleavage results in the formation of 

 two V-shaped chromosomes touching one another by their ends while 

 the apices are parted asunder and point towards the poles. The legs 

 of each V include at first a very small angle, and lie almost parallel 

 and in contact with one another (c/. fig. 8, and Belajeff, Plora, 1894). 



In this formation of V-shaped chromosomes for the daughter 

 nuclei, it is to be noticed (as Eelajeff, loc. cit., has already pointed out for 

 Larix, Lilium, and Fritillaiia), that there is no bending over of 

 originally rod-shaped bodies through the intermediate S- and J -shaped 

 bodies, but they are formed by the direct fission of V-shaped mother 

 chromosomes.' 



' BeliijeJi {loc. cit.) considers that this is the case in all cases of iiaryokinesis 

 in plants, but this, I think, is not home out by observation. 



