328 ROBERT H. BOWEN. 



the mitochondria the early steps in this process are much clearer 

 in the large cells, but in most respects the small cells are the same, 

 and important differences will be pointed out by way of com- 

 parison. In the large cells the individual Golgi bodies spread 

 out more or less as do the mitochondria and at the same time 

 begin to increase in size (Fig. 4). Growth is rapid and is con- 

 tinued up to the end of the growth-period proper. The Golgi 

 bodies soon begin to migrate out into the general cytoplasm, 

 their movement seeming to be approximately synchronous with 

 that of the mitochondria. Thus they soon come to be scattered 

 in .hap-hazard fashion throughout the cytoplasm (Fig. 6), a con- 

 dition which is maintained throughout the growth-period. Their 

 number seems to be about twenty-five or thirty but it is not pos- 

 sible to make a very accurate count because each cell extends 

 over several sections and the bodies may vary as to impregnation. 

 The point particularly to be emphasized is that the Golgi bodies 

 never at any time form a condensed aggregate in any way com- 

 parable to the so-called " Nebenkern " of pulmonates or the idio- 

 some or " sphere " of many other forms particularly vertebrates. 

 In the small cells a similar growth and migration of the Golgi 

 bodies occurs, but it is characteristic of these generations that 

 the general distribution is preceded by a stage in which the Golgi 

 bodies become more or less completely condensed into one or 

 several masses. Fig. 2 shows a cell in the early growth period in 

 which most of the Golgi elements have condensed into a single 

 body reminiscent of the idiosome of the opossum as figured by 

 Duesberg ('20). It should be noticed that even in the case fig- 

 tired, a few Golgi bodies have failed of incorporation in the 

 " idiosome " and usually the fusion is much less complete, result- 

 ing in several masses with numerous outlying, separate bodies. 

 The condition in the large cells (where this phenomenon never 

 occurs), together with the appearances often found in material 

 fixed with Flemming and the customary failure to form a close 

 aggregate even in Golgi preparations, leads me to believe that the 

 fusion may possibly be in part a result of poor fixation. How- 

 ever, the constancy of this phenomenon argues something in 

 favor of its reality, and the similarity to the idiosome and " Ne- 

 tenkern " is very suggestive. It is this body which Montgomery 



