Dixon — On the Chromosomes of Lilium longiflorum. 715 



the pollen-tetrads. I supplemented this deficiency by observing this 

 division in Lilium chalcedonicum and L. Martagon. In the former, as 

 Farmer has observed for L. Martagon, the second division of the pollen- 

 mother-cell conforms to the normal type of plant karyokinesis, i.e., the 

 chromosomes do not appear double before the nuclear plate is formed, 

 but in the nuclear plate they undergo longitudinal fission (figs. 13 and 

 14) ; the daughter chromosomes become V-shaped, by passing through 

 J -shaped stages and the chromosomes are not so short and thick as in 

 the preceding division. A point that may be noticed in this division 

 is the fact that very often two of the daughter chromosomes remain in 

 contact with one another by one end, after their other end and the 

 other chromosomes have been gathered into the daughter nuclei. 

 Again, during even the later stages of division, fragments of the 

 nucleolus remain in proximity to the achromatic spindle and daughter 

 nuclei. 



Development of the Embryo-sac. 



In nuclear divisions in the very young ovule nuclei are to be found 

 with sixteen and with twenty-four chromosomes. In the earliest 

 stages I examined the embryo-sac was only distinguishable from the 

 surrounding cells of the ovule by its very slightly larger size, and by 

 its more granular protoplasm, but it is not uncommon to find two to 

 four cells in the axis of each ovule, all of which, at this stage, are 

 larger than their neighbours, and their protoplasm is distinguished 

 by its more granular appearance. In a division of one of the large 

 cells lying beside that which eventually becomes the embryo-sac, I 

 have counted sixteen chromosomes. 



The primary nucleus of the embryo -sac, like that of the pollen- 

 mother-cells, passes through the "synaptic" stage, prior to entering 

 into the earlier stages of division, and its behaviour is similar in all 

 the stages of karyokinesis and the stages leading up to it, which I 

 have examined, to that of the nucleus of the pollen-mother-cells. 

 Naturally, so many preparations of this nucleus could not be examined 

 as of the nucleus of the pollen-mother- cells and, consequently, the 

 changes were not followed in such detail. 



Before the nucleus becomes " synaptic," the embryo-sac has 

 enlarged considerably and the nucleus usually lies towards its basal 

 end. Above the nucleus in the protoplasm, fibres of great distinctness 

 become apparent, forming a spindle in the finely granular protoplasm, 

 the axis of which is at right angles to the axis of the embryo-sac (fig. 

 15), The appearance and position reminds one strongly of F. Herman's 



