4 Papers from the Department of Marine Biology. 
complete. But the retention of the mantle and the mantle cavity is 
due to the exigencies of the reproductive method in these forms, which 
compels an elaborate mechanism of ducts and incubatory spaces. 
The most remarkable feature of the group, however, is the endo- 
parasitic life-history, which has its only parallel in the Crustacea in 
the copepod family Monstrillide. The Nauplius and Cypris larve 
resemble those of other Cirripedes except in the absence of a gut. 
The latter fix at the base of hairs on the carapace of crabs and an 
internal mass of cells leaves the larval cuticle and passes through the 
gap in the carapace, which the articulation affords, into the body-cavity 
of the host. After a period of wandering it becomes attached to the 
intestine in the region of the abdomen and here it grows, absorbing 
nourishment from the blood and differentiating into a root system and 
a body destined to become the external sac. The internal stage of 
development is terminated by a moult in Sacculina, the sac-like body 
eating its way through the muscle and epithelium of the abdominal 
wall so that it emerges when the chitinous exoskeleton is cast. In 
adult life connection with the internal root system is maintained 
through the narrow peduncle occupying the aperture which served for 
escape. 
Though it has been pointed out above that other parasitic Crustacea 
possess a root system, this is always developed, so far as is known, by 
a secondary inpushing of tissue from the externally situated body. 
The complete description by Delage (2) of the development of Saccu- 
lina was published in 1884, but so extraordinary was its nature that 
for a long time doubt was cast upon the correctness of the account. 
In particular the late Alfred Giard maintained that the Cypris larva 
underwent metamorphosis into the adult at the precise position of 
fixation, as proved by his personal observation. It was not in fact 
until Geoffrey Smith, in 1906, published a striking confirmation of 
Delage’s story that the matter was put beyond doubt. That an 
endoparasitic stage occurs in the life-history of Sacculina and Pelto- 
gaster is now certain; but it has not been actually proved to occur in 
any other genera of the group. For the most part they are similar 
in structure and habit and it may be safely assumed that their develop- 
ment runs through a course not unlike that of Sacculina. But, in one 
genus at least, Thompsonia, which forms the subject of this paper, the 
facts that an enormous number of external sacs are associated on the 
same host and that these sacs have an exceedingly simple structure 
seem to suggest that a different method of development may occur 
here. The endoparasitic stage must have been interpolated in the 
life-history and it is not perhaps unreasonable to suppose that amongst 
the many forms of the Rhizocephala one should be found to exhibit a 
simpler and, in its omission of the endoparisitic stage, more primitive 
life-history than that of Sacculina. Coutiére (1), who has described the 
