On the Rhizocephalan Genus Thompsonia, etc. 25 
well prove that the slitting up of a single central tumour into the rudiments of 
several individuals takes place at a still earlier stage, before any differentiation 
of the tumour has occurred, possibly soon after the entrance of the embryonic 
cells of the Cypris larva. The final test of this hypothesis must, however, be 
left to the future, for someone who has the opportunity of studying either 
this species or some such form as Thylacoplethus, where material is abundant.’’* 
I have given this quotation from Smith in full because I feel certain 
of the general correctness of his suggestion that where ‘‘numerous 
individuals on a single host’’} occur in the Rhizocephala they are really 
produced from a single Cypris larva by a process of budding. It is 
amply proved in Thompsonia, I claim, by the evidence here put for- 
ward, even though I have not been able to describe the endoparasitic 
stages; and in Peltogaster socialis, with which the passage I quote is 
specially concerned, I feel confident that a similar phenomenon takes 
place. 
It is perfectly certain, however, that the details of the budding pro- 
cess differ in the two genera. Smith has shown in P. socialis, there are, 
in the endoparasitic stage, separate central tumours (the anlagen of 
the visceral mass) and separate root systems belonging to them. 
He suggests that division of the original embryo may even take place 
before there has been any differentiation. Thus early division and 
consequent development of separate embryos, each with a separate 
central tumour and root system, are probably characteristic of the 
budding process in P. socialis. In Thompsonia the parasitic organism 
is at all stages continuous within the body of the host and consists of 
a single root system which is established first, and a large number of 
reproductive sacs which are budded off and become external at a later 
phase. Their number and late and recurrent appearance are character- 
istic of Thompsonia and correlated with a degenerate structure. 
It will be of interest in connection with this discussion to give a small 
series of drawings of a social Rhizocephalan very similar to Peltogaster 
socialis and, I think, identical with the form from Japan described by 
Kriiger as a new genus and species (Peltogasterella socialis). In 1911, 
while enjoying the delightful hospitality of Professor Trevor Kincaid 
at the Puget Sound Marine Biological Laboratory at Friday Harbour, 
I collected a number of individuals of the hermit crab, Pagurus alas- 
kensis heavily infected by this form. Plate 2 represents the para- 
site at three widely different periods of growth, but in each case all the 
individuals are of uniform size, and, it can hardly be doubted, of the 
same age too. The only alternative to a theory of budding is that of 
simultaneous fixation of the Cypris larve. This theory indeed sup- 
*Smith, l. ¢., pp. 57-58. 
+This does not apply to all the numerous instances where a crab bears two or three Sacculine 
or a hermit crab more than one Peltogaster. Far more commonly these must be cases of suc- 
cessive infections. 
