Zool— Vol. I.] EISEN—PLASMOCYTES. 47 



this theory further we might conclude that before this sym- 

 biosis took place the " contracting parties" were independ- 

 ent elements, perhaps similar to the nucleus and the plasmo- 

 cyte. As far as we know at the present time these plasmo- 

 cytes are the only independent organisms which do not 

 possess a nucleus, and the suggestion that the non-nucleated 

 ancestors of our nucleated cells resembled them is, I think, 

 entirely reasonable. 



From the observations of Heidenhain on the budding of 

 the centrosome, it seems probable that this body follows the 

 same law which teaches that the nucleus comes only from a 

 nucleus and the cytoplasm only from a previous cytoplasm. 

 The centrosome would thus always originate from a previ- 

 ous centrosome and not from any microsome that might 

 happen to be present. If such be the case, and in my opin- 

 ion all indications point that way, then the theory of sym- 

 biosis must be extended. Instead of a duality in the sym- 

 biosis we would have at least a trinity — cytosome, archo- 

 some, and caryosome — each contributing towards the forma- 

 tion of the perfect cell; they may not, perhaps, be of equal 

 importance, but each is of sufficient consequence to make 

 cell life impossible without it. We now know that the arch- 

 osome is a complete body of three distinct parts. To what 

 extent are these independent of each other? What is their 

 phylogenetic development? Is the centrosphere of the same 

 nature as the centrosome and somosphere, or have we to 

 deal with a further symbiosis of the integral parts of the 

 archosome? 



I have already suggested that the somosphere is the as- 

 similative organ of the archosome and that it probably stands 

 in the same relation to this body as does the granosphere to 

 the cytosome or to the cell at large ; but is it not probable 

 that, regardless of the somosphere, the archosome may be 

 composed of two distinct symbiotic entities, the centrosphere 

 and the centrosome, each of which is apparently, to some 

 extent at least, independent of the other. I will here call 

 attention to the observation made by Rawitz on the tempo- 

 rary connection of the archoplasms in adjoining cells, while 



