714 Proceedings of the Royal Irish Academy. 



As the V-shaped chromosomes move away from the equator, the 

 distance between each leg of the V increases, and at the same time 

 they begin to exhibit a distinctly beaded appearance, each leg being 

 formed of two to four bead-like masses of chromatin (fig. 9). At the 

 angle of the V the space between the bead-like masses becomes greater 

 than along the legs, and. finally, at the poles, or before the V-shaped 

 chromosomes have reached the poles, they each break into two portions 

 lying parallel to one another, and representing the legs of the V 

 (fig. 10); each of these straight chromatia bodies is composed of two 

 to four chromatin granules. In the daughter nucleus, even before it 

 is provided with its membrane, each of these chromatin bodies breaks 

 up into its constituent granules, so that these nuclei are composed of a 

 large number of spherical granules apparently unconnected with one 

 another (figs. 11 and 12). 



Judging from the appearances just described, I think it probable 

 that each of the chromosomes into which the nuclear thread of the 

 pollen-mother-cells breaks, prior to karyokinesis, corresponds to two 

 chromosomes of previous divisions. These two chromosomes in the 

 early stages of division may be united end to end, and at first bend 

 loosely on one another or lie across one another in various ways. 

 Later on they come to lie side by side and present the appearance of 

 being formed by the longitudinal fusion of one chi'omosome. If this 

 view is correct, the doubling of the chromosomes previous to entering 

 into the nuclear plate does not correspond to the division of the 

 chromosomes which takes place in the equatorial plate of other cells ; 

 this latter cleavage, however, takes place in the pollen-mother-cells 

 when by means of the horizontal fission the two V-shaped daughter 

 chromosomes separate one from another. 



The double nature of the chromosome in this karyokinesis is not 

 only thus shown by their mode of origin, but also their appearance in 

 the nuclear plate points to the same fact ; and, finally, when the 

 daughter chromosomes approach the poles they sever themselves more 

 or less completely into two parts, again suggesting, as they do so, 

 their compound nature. In this way these observations seem to bear 

 out the view expressed by Strasburger,^ that the reduction in number 

 of the chi'omosomes is effected by the union of the chromosomes of 

 previous divisions in pairs end to end. 



Unfortunately none of the material of Lilium longiflorum was fixed 

 at the right time to catch the division of the daughter nuclei to form 



Annals of Botany, September, 1894. 



