304 CG. CLIFFORD DOBELL. 
Now let us consider all these facts about chromidia, regard- 
less of any hypotheses which have already been introduced to 
“ explain” them. 
First, it seems to me that the evidence at present is 
strongly in favour of the view that in the Metazoa most of 
the so-called chromidia are really scattered remnants of 
centroplasmic fibrils or their derivatives—properly speaking, 
not nuclear chromatin at all. Consequently, I believe that 
any hypothesis which is based upon the assumption of their 
nuclear nature! has a very insecure foundation. But before 
we have more facts to go upon it seems to me premature to 
argue the matter further. 
Secondly, I believe that certain facts regarding the 
Protista are sufficiently well established to permit of generali- 
sations being made. 
It is perfectly evident that under the name chromidia 
at least four quite distinctly different things are 
comprised, whose morphological resemblance alone allows 
of their sharing a common title. Physiologically they are 
quite different. First, chromidia may represent the normal 
condition of the chromatin in a vegetative cell which 
has no formed nucleus (e. g. in Bacteria, Siedleckia, etc.). 
Secondly, chromidia may be the products of cell meta- 
bolism—either of the nucleus (e.g. Actinospherium)? 
or of the cytoplasm (e.g. Stenophora).®? Thirdly, chromidia 
may be decomposition products of the nucleus, due to 
degeneration or death of the cell (e.g. degenerating 
Opalina).* And fourthly, chromidia may represent one 
stage ina process of multiple nuclear division (e.g. 
Mastigella).’? This process of nuclear division occurs fre- 
quently—though not exclusively (cf. isospores of Radiolaria, 
p. 289)—in the formation of gametes. 
' IT do not mean to imply that the centrosome and centroplasm were 
not originally themselves derived from the nucleus. On the contrary, I 
regard this as highly probable. 
2 See p. 282. 3 See p. 297. 
4 See p. 293. 5 See p. 288. 
