390 bulletin: museum of comparative zoology. 



tions 10,202, 10,203, 10,211, 10,212. They are provisionally identi- 

 fied as L. sculigera, because their gonads agree with that fotm (Mc- 

 Crady, 1857, Brooks, 1886). 



Liriope species? 



Four Liriopes in which the gonads were entirely destroyed were 

 taken at Stations 10,203 and 10,209. Their condition is so poor that 

 it is idle to attempt to identify them. 



Geryonia proboscidalis (Forsk^l). 



Medusa proboscidalis Forskal, 1775, p. 108; 1776, pi. 36, fig. 1.. 

 (For synonj^my, see Bigelow, 1909a, p. 116). 



This well-known species occurred twice, at Station 10,206, where 

 fragments of a large specimen were taken in a haul from 100 meters, 

 and at Station 10,200, surface, 3 specimens so young that the gonads 

 have not appeared. 



NARCOMEDUSAE. 



The classification of the Narcomedusae has been the subject of 

 much discussion. And while all recent authors agree that Haeckel's 

 (1879) system is unnatural, the revisions proposed by Maas (1904a, 

 1904b) and adopted, in its essentials by Mayer (1910) and by me 

 (1909a, 1913) difl^ers fundamentally from Vanhoffen's scheme (1908a, 

 1912a). The characters which have been used to separate, first, 

 families, then genera in this group, are, the presence or absence of 

 gastric pockets, the location of the gastric pockets, radial or inter- 

 radial, and the presence or absence of a peripheral canal-system. To 

 Maas, and to me, the gastric pockets seem the more important, while 

 Vanhoffen chooses the canal-system, these two systems resulting in 

 widely divergent alignments of genera, and even species. I have 

 already (1909a, p. 48), ghen reasons for considering the state of the 

 gastric pockets the more significant character, phylogenetically, of the 

 two: they are, in brief that canals may, or may not, be present, in 

 genera obviously closely allied, if not directly related in phylogeny 

 (e. g., Aegina, Solmundella, and Aeginopsis), and even in different 

 species of a genus (c. g., in Cunina) : that different genera may exhibit 

 gradations in the degree of development of the canals, and that at 

 least two Narcomedusae {Cunina prohoscidea, and Pegantha clara) have 



