STUDIES ON INSECT SPERMATOGENESIS. 57 



it has not yet been possible to distinguish any internal structural 

 features. However, in one case, Murgantia histrionica, I have 

 found that the mitochondria at an early spermatocyte period pass 

 through a granular stage in which it is possible clearly to make 

 out the same differentiation into two substances which others 

 have noted in the like granular mitochondria of Lepidoptera. Sub- 

 sequently the granules produce the long, thread-like mitochondria 

 typical of Eiischistus and other pentatomids. It is, T think, rea- 

 sonable to presume that the structure of the granules is carried 

 over into the threads, which would thus correspond in structure 

 with those of Tcnebrio. Without going further into these purely 

 structural aspects, it seems to me fair to conclude that in the male 

 germ cells at least the mitochondria are generally composed of 

 two distinct substances, related to each other as I have described 

 above. Whether there is any relation between this morphologi- 

 cal duplicity and the presumed chemical composition (lipoid plus 

 albumenoid) of the mitochondria is entirely problematical. In- 

 deed, it is not unlikely that our ideas of the chemical nature of 

 these bodies will have to be revised in the light of recent cytologi- 

 cal investigations. 



Assuming, then, a duplex morphological structure for the 

 mitochondria, we are now in position to take up the structure of 

 the nebenkern at the time when it has just rounded out into its 

 typical shape. Here, again, our problem is simplified by a pre- 

 liminary study of the formation of a nebenkern from the granular 

 type of mitochondria. This was first observed in detail by Meves 

 ('oo) in P \gccra, and more recently the same ground has been 

 worked over by Gatenby ('i/) in a most illuminating manner. 

 Meves showed that at the close of the second maturation division, 

 the vesicular mitochondria begin to fuse together into an irregu- 

 lar, darkly staining mass containing many large, non-staining 

 vacuoles. Subsequently, the stained substance is aggregated into 

 a roughly spherical mass enclosed in a non-staining envelope 

 which is very sharply delimited externally and is traversed by 

 numerous strands passing out from the central, chromophilic 

 core. These latter disappear eventually, and the nebenkern is 

 smoothed out into the two zones observed many years before by 



