Sex -Determination in the Gall-Fly, N. lenticularis. 193 



nature of the original strands from which the chromosomes were formed. 

 Without deciding definitely, therefore, it may probably be concluded that 

 the division is longitudinal, and that each of the rods arranged lengthwise on 

 the spindle splits longitudinally, and one half slips over the other and 

 passes to the inner pole, leaving the other nearly stationary. The form of 

 some of the chromosomes represented in Fig. X, 2, might correspond with a 

 division of this sort. That such a division is very unlike what is found in 

 other organisms is obvious, but it appears to be most nearly what the 

 figures observed suggest. 



The division of the outer group which follows is perhaps of the same kind, 

 as is indicated by such figures as X, 5, in which some of the separating 

 chromosomes appear to be attached to one another in such a way that the 

 outer end of the inner chromosome is lying parallel and in contact with the 

 inner end of the outer. In other cases the figures give the appearance of 

 a. sorting out of the chromosomes of the inner group, rather than of a 

 division (I, 4 ; X, 4), but the fact that after the two groups have separated 

 some chromosomes may remain connected by long chromatin strands* rather 

 indicates a genuine division. The final result of the division of the outer 

 group of chromosomes in eggs of this class is usually a rather regular 

 group of about 10 chromosomes arranged in a sheaf near the edge of the 

 egg, and a less regular group at a somewhat deeper level (Figs. IV, 5, 6 ; 

 HI, 7). 



The above description applies to eggs of females whose eggs in their 

 earlier stages have " top-shaped " nuclei, and my few good figures of eggs of 

 the " spherical nucleus " class make it doubtful whether the divisions are 

 quite similar. The only really satisfactory figure of the stage immediately 

 preceding the division of the inner group of chromosomes is shown in 

 Fig. VIII, 2, and it will be seen that in this case the inner group, instead of 

 consisting of a regular sheaf of rod-like chromosomes (cf. Figs. II, 4 ; III, 5) 

 is made up of a rather irregular assemblage of rods and strands. It is 

 possible that this appearance is due to the chromosomes being actually in 

 division, and in process of separation from one another, for the number is 

 clearly more than 10, though the irregular arrangement makes an exact 

 count impossible. A rather later stage is shown in VII, 7, in which eight 

 or nine chromosomes are clearly seen sinking in to form the egg-nucleus, 

 while those of the outer half of the inner group have loop-like connections 

 with one another. In both figures the chromosomes of the outer group 

 show no clear sign of division at this stage, though traces of spindle fibres 

 may be seen running outwards towards the edge of the egg. Fig. VII, 8, a, b, 

 * Cf. Part I, Plate 2, fig. 32, ' Roy. Soc, Proc.,' B, vol. 82. 



