66 Nils Hj. Odhner. 
characters on account of which they have been included in the Mol- 
lusca, it necessarily follows that, as the grounds for their reference 
are not maintainable, the Selon beanies must be excluded from the 
Molluscan phylum. | 
One might, of course, extend the limits of the Moltitaeit phylum 
so as to comprise also the Solenogastres in spite of their lack of the 
characteristics mentioned, by laying stress exclusively on the pre- 
sence of a pericardium. But in this case it might be objected that 
a pericardium is not an absolute molluscan characteristic, since it 
exists also in the Tunicata. I leave it open whether such a change 
in a well-grounded definition would be justified from the view of 
nemenclature. As to the actual facts one would certainly be justified 
in making it in the case that no close relations to any other group of 
animals could be established. 
With regard to the Solenogastres, however, there can be shown 
many points of affinities in other directions, as we shall find in 
the sequel. 
We have assumed that the ventral sole of Gymnomenia is a 
primitive creeping disc and that its entire absence as in Chaetoderma 
signifies the most ancient condition, which has its closest correspon- 
dence among the worms. If we start from the assumption, supported 
by WIREN’S investigations, that it is in the Annelids, that the closest 
vermian affinities of the Solenogastres are to be sought, we find it 
rather difficult to establish the most intimately related type among 
these worms, which as we know, have developed in several directions 
with various specializations. There exist, however, one group of 
Annelids with a more simple organization than the main stock, a 
group which in reality shows a rather striking conformity in many 
respects with the Solenogastres: I refer to the Archiannelids. 
Apart from the cephalic tentacles, the above-mentioned forms 
show similarity in many details with the Solenogastres, which seems 
to be more than an accidental convergency. This is especially the 
case with the genus Protodrilus, which has been thoroughly in- 
vestigated and admirably monographed by PIERANTONI (1908). 
Here a ventral ciliated furrow is present, which does not exist in 
all Archiannelids (e. g. not in the genus Polygordius). Thus in their 
external shape the Archiannelids unite the same characters as are 
represented by the Chaetodermatidae on the one hand, and the 
Neomeniina on the other. In Protodrilus, moreover, the pharynx I | 
is muscular and contains a yellow cuticular substance, which may — 
