On the Rhizocephalan Genus Thompsonia, etc. 23 
sacs, then, evidently originate from the internal club-shaped processes 
of the root system; the rootlets which supplied the old external sacs do 
not regenerate new ones at once. 
Clearly, then, successive crops of sacs containing the reproductive 
organs are produced by the root system. It would seem likely that 
the period of development of the parasite has been adjusted to corre- 
spond to the time elapsing between moults of the host. This theory 
is supported by the fact that in the case cited above the external sacs 
contained advanced larve at the time of moulting. I imagine that the 
disturbance connected with capture and change of conditions hastened 
the moult, which otherwise might not have occurred until the larve 
were liberated. If the correspondence does not exist, or is only partial, 
an immense waste must go on; for surely the moult of the host and 
consequent disconnection from the root system will suspend the devel- 
opment of the parasite and interfere with the mechanism for securing 
the liberation of the larve. 
DEVELOPMENT OF GERM CELLS IN THE ROOT SYSTEM. 
I have pointed out that the peripheral root system contains in its 
lacunar space multitudes of minute cells, which, in the formation of 
internal buds, migrate into the interior and form the ovary of the 
future external sac. At first they all appear exactly similar, but it is 
clear that only a small proportion actually become ova, the remainder 
being interstitial cells. We do not know how the germ cells are stimu- 
lated to simultaneous development, but whatever stimulus it is that so 
acts is not confined to the external sacs, but is also felt throughout the 
peripheral root system; for when this is examined in a specimen with 
well-developed external sacs there are seen, among the cellules of the 
lacunar space, much larger bodies corresponding in size and structure 
with the segmenting eggs found in the external sacs. One of these is 
figured here (text-figure 8, ov.). These eggs, for such they undoubt- 
edly are, are not so numerous as those concentrated in the external 
sacs, but they are widely diffused throughout the peripheral system. 
They develop until gastrulation begins, but how much further than 
that I do not know. They certainly do not produce larve; yet the 
early segmentation seems to be normal and the egg is surrounded by a 
chitinous shell. In one of my preparations the egg appeared to 
break up at a later stage and the blastomeres to separate, but this 
was indicated only indistinctly. The existence of these bodies in the 
root system is exceedingly interesting and the problem of their fate 
is important. 
I have not observed whether the stages of development of the eggs 
in the external sacs and in the root system correspond exactly. 
