SEMAEOSTOME^E — PELAGIA. 



571 



seems probable that tbe foldings observed liy Vaiihiiffen in tlie netrlc-warts are laigel)' due 

 to shrinkage in alcohol. 



The medusae oi Prlngia, being pelagic in all stages, are creatines of the high seas; and 

 one would expect the species to be of world-wide distribution and at the same time to have 

 developed many local varieties which are not very clearly differentiated from their parent 

 stocks. I believe that all of the forms may be grouped into 4 cohorts as follows; 



(1) P. nocliluca, neglecia, ami crassa, of the Mcditrrranean and Atlantic, with large, clongati' ncltlc-warts over the 



exumbrclla. 



(2) Allied to (l) in the Pacific, we find P. f areola, denticulala, tahiliana, and fapillata with large oval, erect nettlc-warts. 



(3) P. cfanella, peria, discoldea, phosphora and minuta of the Atlantic with small, rounded nettle-warts. 



(4) P. panopyra and placenta of the Pacific with small, flatly dome-like ncttle-warts. 



Cohorts (i) and (2) are closely related; and (3) and (4) form another group. 

 As in Cyauea and AtirrUia so in Pilagia we find that the Linnean system is inadequate 

 to e.xpress the relationship of the numerous, closely related forms. 



♦Development without alternation of generations. 



fDevelopment through a pelagic larva without alternation of generations. 



JToo imperfectly known to be retained. 



The remarkable development, which is direct, without a sessile larval stage, is described 

 in detail under P. noctiluca and P. cyanella. The young medusa passes through a 4-tentacled 

 stage, before acquiring its 8 adradial tentacles. 



"Zonepkvra corona," Agassiz and Mayer, 1902 (Mem. Museum Comp. Zool. at Harvard 

 College, vol. 26, p. 157, pi. 4, figs. 19, 20), is probably a young Pclagia. 



