36 AUGUSTA ÄRNBÄCK-CHRISTIE-L1NDE, NORTHERN AND ARCTIC INVERTEBRATES. 



pare those species it is evident that they are closely allied to each other and not 

 sharply distinct. It seems possible that further investigations based on a more ample 

 material raight prove some of them to be local species only. 



According to Van Name and Hartmeyer C. mollis Stimpson is no doubt 

 identical with Styela vestita, re-described by Ålder a. Hancock and occurring off the 

 British coast. Judging from the descriptions of the species, C. mollis and C. mor- 

 tenseni are both distinguished by the presence of intermediate longitudinal vessels; 

 in the former there are, on each side, three well-developed branchial folds, in the 

 latter there is only one. In the other species the branchial folds are all more or 

 less rudimentary and the intermediate vessels are reduced. In C. cirrata rudiments 

 of intermediate vessels have been observed. The number of the gonads is reduced 

 in both C. cirrata and C. mortenseni; but the alimentary canal of the former agrees 

 with that of G. mollis and C. mollispina, the stomach being short and straight and 

 the intestine of moderate length. In C. mortenseni the alimentary canal, character- 

 ized by an elongated stomach and intestine of great length, shows resemblance to 

 that of G. rhizopus. It should be noticed that the last-mentioned species has not 

 hitherto been reported from West Greenland where C. cirrata seems to be abundant. 



As to the externa! aspect C. mollisfina resembles certain Molguloid Ascidians 

 being of almost globular form and covered with sand (ef. Pl. 2, fig. 23). The test 

 of that single specimen of the collection shows a peculiar strueture, being provided 

 with soft spine-like processes which are clearly shown from fig. 23. As regards the 

 internal anatomy the reduetion of the number of the endocarps is noticeable. In 

 all other members of the group here examined there are a great number of endo- 

 carps scattered över the inner surface of the body-wall. In G. mollispina there are 

 only two, one on each side. (Pl. 2, fig. 24.) Even the dorsal tuberele of the species 

 in question differs from that of the other Cnemidocarpae; in the former it is elon- 

 gated and narrow, the opening is somewhat curved, in the latter it is large, of a 

 rounded form and the opening is more or less circular. (Pl. 2, fig. 25.) From a 

 comparison between figs. 22 & 24 it appears that the alimentary canal of C. mollispina 

 is of quite another strueture than that of G. rhizopus; its resemblance to that of C. 

 cirrata is evident. 



The gonads of the three last-mentioned species show different details in their 

 strueture, though they agree in principal. 



The polycarp-like aspect of the gonads of G. cirrata should be noticed. (Pl. 

 2, fig. 29 & 32.) 



Gynthia villosa, described by Kupffer (1874) and re-described by Hartmeyer 

 1903 (1. c.) under the name of Styela villosa, belongs to the rhizopus group. Whether 

 it is synonymous with C. rhizopus — which seems most probabie — cannot be decided 

 without a re-examination of the specimen in question. If this would prove to be 

 the case, the name of C. rhizopus must perhaps be changed. 



Styela sabulifera Ritter (1913) from the Bering Sea should be included in the 

 genus Cnemidocarpa ; it is closely related to C. rhizopus. Judging from the descrip- 

 tion, there are no principal differences between the two species. The branchial folds 



