GERM CELLS IN THE ARTHROPODA 135 
in the autumn but return to the mitotic method in 
the spring, later giving rise to functional spermato- 
gonia. Vom Rath finds amitosis but contends that 
the cells that divide in this way do not become sper- 
matozoa but are degenerating, being used as nutritive 
material by the other spermatogonia. The amitotic 
divisions described by McGregor (1899) in Am- 
phiuma differ in certain respects from those of 
both Meves and vom Rath. In this species the 
primary spermatogonia divide by amitosis; their 
products later divide by mitosis and produce func- 
tional spermatozoa. Our knowledge concerning ami- 
tosis in the spermatogonia of AMPHIBIA is therefore 
in an unsatisfactory state, although the observations 
of Meves and McGregor argue strongly in favor of 
this method. 
Ca@LENTERATA. While no direct nuclear divi- 
sions were recorded by Hargitt (1906) in the germ 
cells of Clava leptostyla the absence of mitotic figures 
in the early cleavage stages of the egg led him to the 
conclusion that the ‘nuclear activity differs greatly 
from the oridinary forms of mitosis, and appears 
to involve direct or amitotic division” (p. 229). 
If this were true, the germ cells which are derived 
from these cleavage cells must be descended from 
cells which once divided amitotically. This case 
of supposed amitosis has been cleared up by the sub- 
sequent studies of Beckwith (1909), who collected 
material of Clava very early in the morning and found 
typical mitotic divisions during the maturation and 
early cleavage of the egg and no evidence of amitosis. 
