76 ARVID R. MOLANDER, ALCYONACEA. 
Eunephthya racemosa. 
As early as 1909 THOMPSON expresses a doubt whether we can consider Eunephthya 
racemosa, STUDER, as a distinct species, among other reasons because he supposes the 
dumb-bells observed by STUDER in the bark of the trunk to be nothing but the broken 
ends of the club spicules which are seen there. This well-founded remark is, however, 
based in this case on a wrong supposition. The dumb-bells in the trunk bark exist in 
PE. racemosa, just as in E. glomerata and KE. flavescens, and, consequently, E. racemosa 
does not prove any exception in this respect. These dumb-bells exist most numerously 
in Z. flavescens, a species that £. racemosa resembles very much as regards the shape and 
size of the anthocodiae, the mode of growth of the colony, ete. From THOMPSON's descrip- 
tion it appears that the species mentioned by him is £. flavescens. Here, too, we find the nu- 
merous, long and small spindles and the more irregular clubs. Z. racemosa is marked by 
rather low colonies, 4—35 cm., and by its few branches. On every branch there is an in- 
significant number of rather large anthocodiae. The trunk tapers greatly towards the 
point. The anthocodiae are club-shaped, 4 mm. in length. The spicules are similar in 
shape and size to those of £. flavescens. 'The walls between the coelenterons are filled 
with spicules, and there is no sign of any endodermal, indirect canal system. The con- 
nection between the coelenterons passes through wide, endodermal direct canals. This 
deviates a great deal from the canal system in general in this genus, and also from that 
in E. flavescens. The great development of spicules in the walls between the coelenterons 
possibly may produce direct connection, as we see in the case of Hunephthya groenlandica. 
The insignificant ramification of the colony, however, and the short branches produce 
great crowding among the coelenterons. In other specimens this crowding is prevented 
by the profuse ramification, from which follows a suitable distribution of the coelenterons 
in the trunk and branches. But in this case ramification does not exist, and direct con- 
nection occurs. This construction of the canal system cannot be relied upon to differen- 
tiate the species, and I consequently consider Z. racemosa, STUDER, as homologous with 
H. flavescens. 
According to THOMPSON, the specimens investigated by him should be classed as vivi- 
para, something already remarked by DANIELSEN (1887) in connection with E. flavescens. 
The indirect canal system in ZE. flavescens is well developed, except the specimens, 
which are described as E. racemosa. The walls between the coelenterons are filled with 
wide canals, which communicate actively with each other. I have not found direct 
connection between the coelenterons in the cases investigated, and there exists no crow- 
ding in general between the coelenterons, this seeming to be the cause of the want of 
direct connection. 
Habitat: 
Newfoundland, Lat. 52? 5' N., Long. 52” 19' W., 290 met., sand mixed with clay 
(Ingegärd and Gladan, !/, 1871), 1 sp. 
Davis Sound, Lat. 63” 47! N., Long. 52” 25' W., 63 met., shells (Ingegärd and Gla- 
dan, -s/- 187158 spe: 
Baffin Bay, Lat. 68” 8 N., Long. 58” 47! W., 304 met., stone, clay (Ingegärd and 
