Sex-Deter m* nation in the Gall-Fly, N. lenticularis. 195 



perspective. Further, although there are considerable differences in detail, 

 in the rather earlier stages especially there is a uniformity about the figures 

 which makes it difficult to believe that the appearances are due to artifact. 



I am inclined, therefore, to believe that the figures are at least to a 

 considerable extent genuine representations of what exists in the living egg, 

 and that the maturation divisions of the Cynipidse exhibit a type of mitosis 

 not hitherto observed in other forms. Its essence appears to be the drawing 

 out of threads from a reticulum ; the threads, at least at their bases, are 

 double, and the ends of the longitudinal halves diverge and anastomose with 

 the threads of the network. Gradually the network disappears, apparently 

 becoming absorbed into the drawn out threads, forming an inner and outer 

 group. These two groups then each divide, but, as was said above, there are 

 no clear indications of how this second division is effected ; it is probable, 

 however, that the division consists in a separation of the longitudinal halves 

 of the originally double threads. To make this description more intelligible, 

 I give in a text-figure a purely diagrammatic representation of the kind of 



Diagrams 1-4. — Eepresenting tlie kind of mitotic division suggested by the figures 

 observed in the eggs of Newrotveus. 



1. Threads becoming drawn out on opposite sides of a reticular nucleus. 



2. Separation and sinking-in of the inner threads, each of which is longitudinally split. 



3. Division (lougitudiual !) of the inner threads, and formation of split chromosomes from 



the outer group. 



4. Completion of the maturation divisions. The innermost group forms the egg-nucleus, 



the other three the groups of polar chromosomes. 



(In diagrams 2, 3, 4, only half of the actual number of chromosomes is represented). 



system which the figures suggest, but it will be understood that this 

 representation is imagiuaiy, and is not intended to show at all accurately the 

 structures as they actually exist. 



