1 88 ROBERT H. BOWEN. 



Pensa ('13) on living cells indicate that these osmic impregnations 

 are really remarkable demonstrations of the actual condition of 

 the vacuoles. This is borne out by the fact that such networks 

 in the periblem are characteristic only of Vicia (among the 

 forms so far studied by me). 



My results on the vacuoles thus bear out the recent extended 

 researches of the Dangeards and of Guilliermond, and go far 

 toward establishing the vacuome as an independent and perma- 

 nent part of the plant cell system. 



Other Cell Structures. Of the other constituents mentioned in 

 an earlier section as being blackened by osmic acid, none throw 

 any light upon the general problem of cytoplasmic components, 

 and they will not therefore receive further mention here. My 

 results on the blackening of the spindle fibers corroborate, 

 in the main, those of Nassonov ('18), to whose findings I have 

 added the interesting fact that the cell plate can be blackened 

 at all stages of its formation. 



ANTHERIDIA OF POLYTRICHUM. 



The results of my study of root-tips are now fairly complete, 

 and they lead unmistakably to the conclusion that the cytoplasm 

 of plant cells contains four classes of formed elements, which 

 are apparently distinct and independent, and probably self- 

 perpetuating under normal circumstances. In animal cells there 

 are only two such generalized classes of formed bodies chon- 

 driome and Golgi apparatus at least so far as known at present. 

 It is thus impossible to draw any off-hand conclusions as to 

 what components in the plant cell are equivalent to those of 

 the animal cell. Furthermore, the study of meristem cells in 

 root-tips gives no conclusive answer to the problem of homologies. 

 As a means of getting around this unhappy impasse, I suggested 

 in a previous note (Bowen, '266) the possibility of finding 

 processes in the formation of plant sperms which would be 

 comparable to those now known to occur universally in differ- 

 entiating animal sperms. I have already reported briefly on 

 the probability of achieving success in this direction. Further 

 study, particularly of more successful osmic impregnation 

 preparations, has, however, clearly indicated one or two errors 



