HERPYLLOBIUS ARCTICUS. 
399 
Jensen is right in stating that a membrane formed by the 
worm is present. In cases where the parasite and host liave 
been cut together the epidermis of the worm folds in over the 
hold-fast (PL 22, fig. 7, ep.), and although the actual connection 
is not evident, the membrane [mb.) surrounding tlie roots seems 
to be a part of this. In sections of a whole worm from which 
the parasite had become detaclied the epidermis of the host 
(Pi. 22, fig’. 8, ep.), although broken at the critical spot, is 
probably continuous with the membrane [mb.). This mem- 
bx’ane may be traced down the worm as far as the roots 
of the parasite extend, although all the proximal part of 
the root system was withdrawn when the copepod was re- 
moved. A section lower down the worm (PI. 22, fig. 9) shows 
the distal end of the i-oots of Herpyllobius surrounded at 
some distance by a well-marked membrane [mb.), and sections 
of another worm infested with three parasites show the root 
system of each surrounded by a membrane, Avhich is always 
separated by some distance from the tissue of the root, and 
therefore is not likely to belong to it, as the root itself does 
not appear to be at all shrunken. Giard and Bonnier^ state 
that in the case of crabs the parasitic Entoniscan is always 
entirely surrounded by a thin transparent membrane, except 
at one spot which is the opening to the invagination 
caused by the penetration of the young Entoniscan into its 
host; this membrane presents the same histological structure 
as the hypodermis of the crab. As a like process is prob- 
ably effected by Herpyllobius it may then be described as 
an ectoparasite. 
As to the cuticle of the roots themselves, I am inclined to 
think that if this is present it is extremely thin, and only 
evident in the region of the hold-fast (PI. 22, fig. 7, ; in 
sections of the worm with the parasite roots, these are not 
surrounded by any well-mai-ked cuticle (PI. 22, fig. 9). 
The tissue of the root .system is fibrous and does not show 
any definite cell boundaries; these fibres are strongly 
developed round the hold-fast, to the serrated lower edges of 
' Loc. cit. 
