36 



HYDROIDA 



so great a range of variability that the number stated in Laiiipra atlantica must only be considered 

 as an extreme variation. The variation of the individuals found in the Troudhjem fjord is even larger 

 than stated by Swenander (1903); individuals have been found with only 15 tentacles in the proximal 

 whorl. The difference in the shape of the gonophores, being either subglobular or oval with attempts 

 at tapering, is of little importance. Swenander points out that the female gonophores are more 

 globose, the male more oval. In a specimen picked up north east of the Seydisfjord (Iceland) the gono- 

 phores are partly globular, partly oval, with the same attempt at tapering towards the distal end as 

 is shown in the figure of Bounevie (1899, Tab. Ill, Fig. 1 a). However, one criterion is still left, 

 the distal tentacles of La?npra purpurea being arranged in two separate circles, while in other spe- 

 cies they are placed in several irregular and close-set circles. From the comments of Bonnevieit 

 appears that the description has mainly been based on the drawings of G. O. Sars, which were exe- 

 cuted on board ship. In these drawings the double whorl is not peculiarly clearly rendered ; nor can, 

 from the remnants of the original specimen kept, the character be ascertained any longer. From the 

 table set up of the specimens from the Danish collections, however, it appears that, from individuals 

 with a distal whorl consisting of two rows to individuals with a distal tentacle whorl of several rows, 

 every transition may be pointed out. Thus neither this table nor the criteria stated in the other table 

 may be turned to account for the purpose of attaching any peculiar systematical importance to this 

 character. Lampra atlantica and Lampra purpurea, therefore, cannot be recognized as two separate 

 species, and are moreover identical with Corymorpha groenlandica. 



For a thorough examination of the species we are indebted to .Swenander (1903), who considered 

 the Lampra of the Trondhjem fjord as a peculiar species, Lampra socia. From what I have stated 

 above, the shape of the gonophores, applied by Swenander as a main character distinguishing the 

 species from Lampra purpurea Bonnevie, cannot be maintained as a criterion. The number of the 

 blastostyles then remains; Bonnevie, for both the species mentioned, states 10 blastostyles; Swen- 

 ander, for Lampra socia, 18 — 32. Still Swenander has not found the minimum for the individuals 

 of the Trondhjem fjord; as a matter of fact, specimens with only 15 blastostyles are now in hand. 

 In the Danish material collected from the northern Atlantic the number is throughout lower, varying 

 from 10 to 28, thus bridging the difference between Lampra socia and the species stated by Bonne- 

 vie. Also the other distinguishing characters, the emphatic demarkation of the hydrocaulus from the 

 polyp, or its gradual transition into the latter, and the colour of the animals, are varying from one 

 individual to the other, forming no strongly defined limits. The demarkation of the stem from the 

 polyp more or less emphatic, is, in living individuals, varying according to the state of contraction, as 

 in other species of Corymorpha. Therefore, also Lampra socia must be considered as a synonym of 

 Corymorpha groenlandica. 



Jaderholm (1909) and Broch (1909) describe two new species, respectively Lampra arctica 

 and Corymorpha spitzbergensis. The table at once shows that they come within the range of variation 

 of Corymorpha groenlandica. As to Lampra arctica the short and thick blastostyles, maintained by 

 Jaderholm (1909) as a good criterion, are in fact to be looked upon as a mere phenomenon of con- 

 traction which may be partly observed even in preserved material, where in the same individual some 

 blastostyles may be short and thick, while others are thin and strongly extended. Illustrative in this 



