108 THE MEDUSAE. 



but in its habits. While the latter have usually been recorded from land- 

 locked ponds or harbors, where they often cling to Algae, G. suvaeiisis is 

 more truly pelagic, occurring in open sounds and lagoons, or on the surface 

 of the open sea (Bigelow, : 04). Although it has been observed in life at 

 Mangareva, among the Maldive Islands (Bigelow, : 04), and among the 

 Fiji Islands (Agassiz and Mayer, '99), it has never been seen to exhibit 

 the peculiar " turning " habit of swimming so characteristic of G. vertens, 

 G. vmrbachii, and G. depressum (Goto, : 03). 



This species is widely distributed over the Tropical Pacific and Indian 

 Oceans, but appears to be restricted to shoal-water regions in the close 

 neighborhood of land. 



Olindias F. MtUler, 1861. 



Halicalyx Fewkes, '82s 

 Halicalyx Mayer, : 00''. 



Olindiinae with four radial canals, with two kinds of tentacles, the pri- 

 mary tentacles bearing sucking discs ; with blind centripetal canals. 



Maas (: 05) has pointed out that the Atlantic forms of this genus, 0. phos- 

 phoricaDeWa Chiaje (miiUen Jiaeckel), 0. sambaqiie7isisF. Mliller, and 0. tenuis 

 Mayer (Fewkes ?), can hardly be distinguished, since the various characters 

 which have been supposed to separate them, i. e. conformation and lobing of 

 the gonads, and number of marginal organs, not only change with advanc- 

 ing development, but are subject to great individual variation (see Table, 

 p. 109). As a variety of this one Atlantic species, he identifies specimens iu 

 the " Siboga " collection, from the Malay Archipelago, under the name O. 

 phosphorica, var. malayensis. 



Browne (: 04) has recently described under the name O. singularis another 

 species distinguishable from 0. mulleri and its varieties by the presence of 

 distinct bulbs at the bases of the secondary (velar) tentacles, and of only one 

 otocyst, instead of a pair, at the base of each primary tentacle. The present 

 specimens, as well as an excellently preserved specimen from Dutch Papua, 

 appear to belong to 0. singularis Browne, although this large series shows 

 that Browne's original description, taken from but a single specimen, must 

 be somewhat modified. 



Maas (: 05, p. 48) has compiled a table of the numerical condition of 

 canals and tentacular structures ; but since his figures, though illustrating 

 very well their variability, do not altogether agree with my observations 



