404 BOWEN. 



process affecting the cytoplasm so vitally, its several components 

 would develop morphological differences (other than those of mere 

 size or amount) of one sort or another. Such, however, is not the case, 

 and with the exception of a few minor differences upon some of which 

 I have already touched (Bowen '20), the general progress of events in 

 the cytoplasm is the same in all cells regardless of size. With respect 

 to staining reactions and the response to impregnation methods in- 

 volving silver nitrate there are, however, some rather definite differ- 

 ences between the large and small cells, but these are perhaps referable 

 to a physical rather than a physiological difference. 



The centrioles were stated by Montgomery ('98) to vary in size 

 directly with the amount of cytoplasm, but his figures were not en- 

 tirely convincing. I have reexamined this point in Arvelius, in which 

 the facts are demonstrated with unusual clearness. The centrioles 

 in this form are rod-like in shape, often arranged in the primary 

 spermatocytes in a "V" formation recalling the condition in Lepidop- 

 tera. The centrioles in the large generation are strikingly larger than 

 in the smaller generations. (Compare Figs. 1 and 2, in both of which 

 the size of the centrioles has been slightly exaggerated). 



The mitochondria were found by Montgomery (he at first called 

 the mitochondrial material an idiozome) to vary in amount with the 

 cytoplasm. This is easily checked by comparing the large and small 

 spermatids during the stage when the mitochondria are aggregated 

 into the compact, rather accurately spherical nebenkern. (Compare 

 Figs. 3 and 4). 



The Golgi apparatus was not considered by Montgomery. In its 

 scattered condition in the spermatocytes (see Bowen '20) comparisons 

 are not very satisfactory, but after the aggregation of the Golgi ele- 

 ments to form the acroblast in the spermatid, the large and small 

 generations can be readily contrasted (Figs. 7 and 8). As with the 

 mitochondria and the centrioles, the large cells have much the larger 

 amount of Golgi material. So too, the acrosome formed in connection 

 with the acroblast is much larger in the large cells, sometimes very 

 strikingly so (Figs. 9 and 10), and the cast-off acroblast (Golgi rem- 

 nant) varies of course in a similar way (Figs. 9 and 10). 



The chromatoid body is another rather constant feature of the penta- 

 tomid germ cell which was not considered by Montgomery. This 

 structure does not lend itself easily to comparison because of its small 

 size, and often subdivided condition. However, in Rhytidolomia 

 senilis, as discovered by Wilson ('13), the chromatoid body is very 

 large and distinct, and makes a fine object for comparison (Figs. 5 and 



