Dixon — On the Chromosomes of Lilium longiflorum. 711 



there are one to three large nucleoli in the nucleus, and in each is 

 usually to he recognised a large vacuole. In one preparation, at this 

 stage, I saw a chromosome hent into the form of a ring surrounding a 

 nucleolus. Sometimes the chromosomes are hent into the form of a V, 

 but this appears to be the exception. During this period they are not 

 rery susceptible of the usual nuclear stains.^ 



The slender parallel portions of each of the chromosomes gradually 

 grow shorter and thicker (fig. 4), so that they appear roughly quadrate, 

 triangular, or ring-shaped (fig. 5). Tliis latter form arises probably from 

 the more complete approximation of the two portions at the ends than 

 in the middle. "^ In each case the distinction into two portions is more 

 or less completely preserved. When the chromosomes, having these 

 forms, re-arrange themselves after the solution of the nuclear membrane 

 to form the nuclear plate, they dispose themselves in such a manner that 

 when viewed from the poles the division between the two portions is 

 apparent ; at least I believe that the division one sees under these 

 circumstances corresponds to the division in the slender chromosomes, 

 between the approximated portions of the thread, as this division alone 

 is visible at the first formation of the nuclear plate (fig. 7). But as each 

 of the chromosomes forming the nuclear plate appears, in most cases, 

 divided longitudinally into four portions by two planes at right angles 

 to one another, it is difiicult to decide which of these two planes of 

 division corresponds to the space separating the two portions of each 

 chromosome which have come together by the approximation of 

 different parts of the thread. If the horizontal plane of division 

 corresponds to this space, it is evident that the division might be a 

 " reduction division." Often the shape of the chromosome, seen in 

 profile, seems to support this view ; but I think on the whole that it 

 is more probable, for the reasons already given, that the vertical 

 plane of division corresponds to the space separating the two portions of 

 the nuclear thread. In this latter case there could be no " reduction 

 division." At a slightly later period, inspection from the equatorial 

 plane shows these bodies (the short thick chromosomes) beginning to 

 undergo horizontal longitudinal fission (fig. 6) which commences at the 



^ The nucleus of the pollen mother-cells in this stage present a remarkahle simi- 

 larity to I lie corresponding stages in the development of the nucleus of the ovum 

 figured by V. Ilaeker (Arch. f. Mikr. Anat., 1895). 



'^ Cf. Farmer, Ueber Kemtheilung in Lilium-Antheren; Flora, 1895. 



Eelajett'also notices other peculiar forms assumed by the chromosomes in the 

 division of the nucleus of the pollen mother-cells of Larix europcea ; Flora, 1894. 



