16 THEEL, NORTHERN AND ARCTIC INVERTEBRATES. I. SIPUNCULIDS. 
Among all the Sipunculids provided with tentacles, the genera Petalostoma, 
Aspidosiphon and, above all, Physcosoma, form exceptions to the rule, though here 
too the general plan is discernible. In the two genera first mentioned the transfor- 
mation is less remarkable, a simple reduction having taken place, so that either only ” 
two dorsal, rather large and solid, lamelliform, tentacles remain (Petalostoma), or else 
there are several small ones, forming a dorsal semicirele (some species of Aspidosiphon). 
Somewhat more complex is the transformation in the genus Physcosoma. Here the 
rim of the so-called disk forms above the mouth a dorsal horseshoe-shaped duplicature, 
from which alone numerous simple tentacles protrude. Thus, it appears to be a rule 
that, when a reduction of tentacles takes place, it always begins on the ventral part 
of the rim and proceeds towards the dorsal part, where the tentacles remain longest. 
In all other Sipunculids the tentacles form a erown round the mouth; but among 
them too a certain difference exists as to their shape. While the majority have them 
simple, more or less elongate and finger-like, others, e. g. a section of the genus 
Sipunculus, have them in the shape of a lobate membrane; and, finally, it is to be 
remembered that the genus Dendrostoma is marked by possessing only a few (4—6) 
pinnate tentacles. 
The retractor-muscles. The retractors play an important part in the classification 
of the forms. But here too a confusion seems to prevail with regard to the true 
number of them. As an instance I may adduce the following illustration: It is said 
that several species of the genus Phascolosoma, devoid of dorsal retractors, are pro- 
vided with two ventral retractors, others with only a single one. In the latter case 
the retractor has the base more or less distinetly bifureate, and the nerve-cord runs 
between the two roots or to the vicinity of the interspace between them. Now it is 
a fact that in the same species the extent of the bifurcation is very variable, and it 
is quite impossible to draw a line of demarcation between two quite separate retrac- 
tors and one retractor with two so-called roots. All possible transition-stages are to 
be identified. Moreover, it ought to be remembered that the relation between the 
nerve-cord and the two so-called roots of the retractor indisputably manmifests that 
there originally existed two separate ventral retractor-muscles, which have gradually 
become more or less intimately united, and finally remain free only at the bases. 
In 1877 KOREN and DANIELSSEN Wrote with regard to their Phascolosoma pallidum: 
«The nervous cord runs as usual between both retractors''. Their Fig. 23, Pl. XIV, 
also confirms this statement. Nevertheless, both their statement and drawing are 
wrong. The nerve-cord runs in reality to the so-called roots of the ventral retractor. 
In view of the above-mentioned facts it is evident that the greatest arbitra- 
riness and confusion must prevail as to the number of the ventral retractors. One 
investigator tells us that an animal has two retractors, another that the same spe- 
cimen is in possession of a single one with a more or less deep incision at the 
base, all depending upon the particular idea held of the significance of the cleft. 
According to the former the cleft proves the presence of two retractors, which still 
remain separate at the bases. Now my own investigations of Phascolosoma Sabel- 
Jarie and improvisum reveal that there exists a considerable, individual variation as 
