NARCOMEDUSAE — SOLMARIDAE. 431 



adopt, excepting that he divides my iEginidae into two families, Cunanthidae with undivided 

 gastric pouches equal in number to, and in the radii of, the tentacles; iEginidae with gas- 

 tric pockets commonly adradial in the adult, usually 2 to 4 times as numerous as tentacles. 

 In ALgina alternans, a new form found by Bigelow, however, the gastric pockets are niter- 

 radial and alternate with the tentacles. This may have been brought about through the 

 obliteration of the interradial cleft between the originally adradial pockets, but the devel- 

 opment is as yet unknown. 



Family SOLMARIDAE Haeckel, 1879, sensu Maas. 



Pe°anthiiU+ Solmaridc, Hafxkf.l, 1879, Syst. der Mcdusen, pp. 323, 346. 



Solmaridx, Maas, 1904, Result. Camp. Sci. Prince de Monaco, fasc. 28, p. 41; 1905, Craspedoten Medusen der Siboga Exped., 



Monog. 10, p. 80; 1906, Fauna Arctica, Bd. 14, Lfg. 3, p. 498.— Bigelow, H. B., 1909, Mem. Mus. Comp. Zool. at 



Harvard College, vol. 37, pp. 50, Si. 



FAMILY CHARACTERS. 



Narcomedusae in which the periphery of the stomach is simple and circular without 

 peripheral stomach-pouches. With or without saccules in the floor of the subumbrella wall of 

 the stomach. With a variable number of tentacles, the solid entodermal cores of which are 

 attached to the periphery of the central stomach, while the tentacles themselves project stiffly 

 from the sides of the bell. 



The bell-margin is cleft into lappets, which alternate with the tentacles, and the margins 

 of these lappets bear sensory-clubs with concretions of entodermal origin. There may or may 

 not be ectodermal sensory bristle-bearing tracts (otoporpae) over the exumbrella above these 

 sensory-clubs. The velum is a wide diaphragm and also extends upward as a web, bridging 

 over the clefts between the lappets. The peripheral, marginal, ring-canal system is either 

 degenerate or wanting altogether, or well developed. 



The development appears to be direct, the free-floating, actinula larva being gradually 

 transformed into the form of a medusa. Thus the ontogeny of these medusae is widely different 

 from that of the Anthomedusae and Leptomedusae, and the medusa-shape appears to have been 

 attained in an independent manner. The Narcomedusae may, I think, be regarded as medusi- 

 form actinulae and the bell as a secondarily acquired organ, which does not begin to grow out 

 until after some, at least, of the tentacles have developed upon the sides of the oral zone of the 

 actinula. In the Anthomedusae and Leptomedusae, on the other hand, the bell develops before 

 the tentacles and lithocysts, whereas the reverse is the case in Narcomedusae. Thus the bells 

 in the two groups are not homologous structures, but have been acquired independently in 

 response to similar environmental conditions. The entodermal lithocysts of Narcomedusae are 

 evidently not homologous with the ectodermal lithocysts of the Anthomedusae and Lepto- 

 medusae, but are, phylogenetically speaking, merely modified tentacles. 



The tentacles have not migrated upward from the bell-margin as is maintained by Haeckel, 

 etc., for the development shows that the bell grows downward and outward, beyond the bases 

 of the tentacles which simply retain their primitive position. 



The Solmaridae are very widely distributed over warm and tropical seas, but are poorly 

 represented in the Polar Regions. The Mediterranean appears to contain the greatest number 

 of species, whereas these forms are very rare off the tropical coasts of America. 



Haeckel, 1879, states that the Solmaridae have no ring-canal, no peronial canals, and no 

 otoporpae at the bases of the sensory-clubs; whereas his Peganthidae possess these structures. 

 Haeckel, however, cut no sections. As Maas has shown, the ring-canal system in other 

 families of Narcomedusae is very degenerate and subject to great variability in development, 

 even in different individuals of one and the same species. It is, therefore, not a suitable 

 criterion for the separation of families, and is, indeed, hardly of specific value. 



The presence or absence of otoporpae appears to me to be of too slight an importance 

 to serve as a sole distinction between families, and Haeckel's Peganthidae should be merged 

 with the Solmaridae. I would define the Solmaridae as follows: Narcomedusae without 

 marginal stomach-pouches. With or without saccules in the subumbrella -floor of the stomach. 



