75 
that forms interpreted as parasites must show definite specia- 
lization to their new modus of living. The term parasitism must 
necessarily remain somewhat elastic, as, in respect to habits, the 
categories are created in our mind, in Nature they are fluent. 
Descending from free-living ancestors, the parasites have by degrees 
evolved their parasitism. Under such circumstances there are 
always forms which present intermediate stages. In using the 
words true parasites, I shall take as starting-point that in ordinary 
life the animals, firstly, are wholly dependent upon the food they 
receive from the host, secondly, that they clearly show adaptations 
for their stay on the host. 
These introductory remarks serve only the purpose of fixing 
my standpoint for the present case, and are therefore cut brief. 
True parasitism is in this sense unknown in the Polyclads. 
Several forms are, however, on the way to evolve it, but only the initial 
steps have as yet been taken. I might here refer to the enume- 
ration of such Polyclads in the first chapter of this paper. In other 
Turbellarian groups there are nice series of examples of intimate 
relation to the host (v. Gralf 1903, Wahl 1906-1910, Beklemicheff 
1915, Caullery et Mesnil 1903, etc.). The most advanced adap¬ 
tation occurs in the Rhabdocoeles (f. i. Fecampia) and Acoeles. 
Far more bound to the host than any Polyclad are, for instance, 
Bdelloura species among the Triclads, not to speak of the group 
Temnocephaloidea. In the Polyclads the dependence upon the 
host is evidently very slight, and they would be able to carry on 
their life even if compelled to leave their host. There would scarcely 
be any other difference than that they would have lost the eventual 
protective value of the stay and the struggle for food would be 
somewhat augmented during a completely free-living State. Recog- 
nized clearly, this demands that the question of true parasitism 
must be left out of consideration in the present case. If I were to 
make a supposition as to the relation of the Polyclads to their 
hosts, it would be that Hoploplana inquilina (Wheeler) on the 
Gastropod Sycotypus canaliculatus Gili and the Pagurid Polyclads 
described here are more intimately associated with their hosts than 
other known Polyclads. 
A case parallel to the two latter ones is mentioned in the 
literature, inasmuch as it refers to a Polyclad living with Pagurids. 
