MALE CELLS OF RHOMALEUM MICROPTERUM BEAUV. 5 



chondria are usually in two separate masses with a clear space 

 close to the nuclear wall between. An extreme case of this 

 sort is shown in a cell in the bouquet stage (Fig. 6) which will 

 be referred to later. Here it is very clear that a spherical mass 

 must separate the mitochondria. This characteristic clumping 

 of the mitochondria in two groups is mentioned by Lewis and 

 Robertson, but no explanation is attempted. The behavior of 

 this sphere shows that it is in the nature of an attraction sphere 

 or idiozome (Meves). 



Finally there are in the cytoplasm of these cells of the early 

 growth period from three to six or more of the "neutral red 

 granules" mentioned above. They are larger and more promi- 

 nent than in the spermatogonia, and are shown stained with 

 haematoxylin in Figs. 2 and 3. They may lie anywhere in the 

 cytoplasm, but are more often among the mitochondria and 

 close to the idiozome. In the living material they do not appear 

 in the unstained cells, but when cells are stained with neutral red, 

 they take the stain faintly some time after treatment. There 

 are then four sorts of inclusions in the cytoplasm of the cells in 

 the early stages of the growth period: (i) the mitosome, (2) the 

 idiozome, (3) the fragments called "neutral red granules," (4) 

 the masses of granular mitochondria. 



From the stage of the massive bodies the chromosomes uncoil 

 into the leptotene condition. The threads are fine and at first 

 appear as a tangled mass. The monosome is of course an excep- 

 tion to this rule for it remains as a heavy densely staining mass. 

 Shortly after the leptotenes begin to suggest a polarization, 

 the threads become doubled and pass rapidly into the diplotene 

 condition. Here they remain until the growth period is com- 

 pleted and diakinesis begins. In this process of polarization 

 the spherical idiozome in the cytoplasm apparently plays an 

 important part. It is not always visible in the stained material 

 even when the preparations are carefully extracted with a view 

 to making it clear, but in every case in which it can be seen in the 

 late leptotene or diplotene stages it is found that the chromatin 

 threads are polarized towards it. Two clear cases of polarized 

 diplotenes are shown in Figs. 5 and 6. Both are approximately 

 the same stage, the first from a fixed and mounted preparation, 



