64 ROBERT H. BOWEN. 



tained in the stage of Fig. i, in which the chromophobic material 

 has run together to form many small, clear masses or vacuoles, 

 the walls of which are formed, as it were, of the chromophilic 

 substance. The process of condensation which has produced 

 this initial differentiation now continues rapidly (Fig. 2), larger 

 and larger aggregates of chromophobic material being produced 

 by the further condensation of the chromophilic substance. This 

 condensation has seemed, from the beginning, to progress from 

 without inward, so that the primary vacuolization is in the form 

 of concentric shells of vacuoles, each equal in thickness to the 

 diameter of a single vacuole. Thus by the absorption of the par- 

 titions of chromophilic material which at first separate the 

 vacuoles of any one layer, there is produced a very typical figure. 

 The chromophilic material has accumulated in a thin lamella be- 

 tween each adjacent layer of vacuoles, while the vacuoles them- 

 selves have run together to form a more or less homogeneous 

 ground substance which surrounds, and fills in all the interspaces 

 between the chromophilic lamella?. These lamellae seem to be 

 more or less inter-connected and wrinkled, and are possibly per- 

 forated in numerous places, producing a sort of skeletal plate- 

 work, the main outlines of which are molded on the original vac- 

 uolization of the nebenkern. The "spireme" of Gatenby ('17^ 

 and the " plate-work " which I have described are thus directly 

 comparable in every way. Their apparent difference depends., it 

 would seem, on the original form and method of fusion of the 

 spermatidi chondriosomes, which are in one case large, vesicular 

 granules and in the other long, delicate threads. 



I believe, therefore, that, depending on the primary morphology 

 of the auxocyte chondriosom.es and possibly other unknown 

 factors, the nebenkern may develop a " spireme " structure or a 

 system of plates (or possibly other structural types not yet de- 

 scribed). Gatenby ('18) on the other hand seems inclined to the 

 view that all the nebenkern structures are essentially a thread- 

 work comparable to the " spirerrfe " in Lepidoptera. He states, 

 in fact, that a "very finely-coiled spireme" does occur in Tene- 

 brio, and presumably other Coleoptera. It is always possible, as 

 Gatenby has shown in the Lepidoptera, that faulty technique will 



