KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 39. N:o |. itf 
to the depth of this cleft, and also as to the site where the retractors are attached 
to the body-wall. By maintaining the earlier view, that we have to do with a single 
retractor, one would run the risk of referring one and the same animal either to se- 
veral distinct species, or to a group of distantly related species or even to a wrong 
genus. That is also what has happened. The incorrect statement given by KOREN 
and DANIELSSEN, that Phascolosoma Lilljeborgii is provided with a single retractor, 
later on induced SELENKA to place it in his synopsis close to Phascolosoma squamatum, 
a form which is in all respects quite different. The Norwegian authors themselves, 
evidently for the same reason, have erroneously referred two true Phascolosoma-spe- 
cies, Ph. Sarsi and Ph. olaciale, to the genus Onchnesoma etc. 
From this it is evident that, in order to avoid misapprehension and exclude 
arbitrariness, the question as to the number of the retractors, especially as to that 
of the ventral ones, must be treated with the greatest caution. 
In the recent, true Sipunculids the largest known number of the retractors is 
four, namely two dorsal and two ventral. However, there seems to be but little 
doubt that the number was originally larger, and that, moreover, the ancestors, like 
Sipunculus and Physcosoma, had the inner musele-layer of the body-wall broken up 
into longitudinal bundles or bands. According to my opinion, the fact that in the 
two genera mentioned the roots of the retractors run out from two or more longitu- 
dinal bundles, gives us a hint that the retractors were originally more numerous, 
though in the course of time they have united and become fewer. The genus Pria- 
pulus, belonging to a family which is considered to bear some relation to the Sipun- 
culids, still retains its original character in so far as to possess numerous retractor- 
muscles. 
Afterwards, a differentiation has bhegun to take place. Except in Sipuncul!us, 
Physcosoma and some forms of Aspidosiphon, where they remain, the longitudinal 
muscele-bundles of the body-wall have gradually disappeared and been replaced by a 
continuous layer, as is the case in Phascolosoma, Phascolion etc. Simultaneously the 
retractor-muscles, originating in some way from the longitudinal bundles of the body- 
wall, become reduced in number, so that finally only four, three, two or even one 
remain. The reduction always begins with the dorsal retractor muscles, these beco- 
ming first obliterated. It seems also likely that such a reduction has been accom- 
panied by a gradual removal of the retractors towards the posterior portion of the 
body-cavity. 
As a rule, forms with four retractors, two dorsal and two ventral, have these 
attached somewhat more to the front, the dorsal being more so than the ventral. 
In other species which are provided with only two ventral retractors, there not in- 
frequently occurs a displacement of their attachment backwards, though that is by 
no means always the case. Lastly, in such species, as are in possession of three re- 
tractors or of only a single one, the attaching points of these muscles are removed 
far behind, in the latter case as far as to the posterior extremity of the body. In 
the case of a single retractor being present, this must certainly: belong to the ven- 
tral system, 
