42 U.S. NIATIONAL MUSEUM BULLETUST 271 



The cold-water orientation of marine Gammaridea and the diversity 

 of brackish water and marine species (the Anisogammarus complex) in 

 the north Pacific Basin, with affinities to Baikalian genera, also sup- 

 port the thesis that Gammaridea may have had a fresh-water origin 

 in the vicinity of Baikal and then invaded the sea. The thesis becomes 

 stronger if one confines the theory only to the living members of the 

 family Gammaridae, for that family is generically impoverished not 

 only in tropical shallows but in all deep-seas of the world. Numerous 

 species of Maera, Elasmopus and Ceradocus are prominent in the 

 tropical sublittoral but generic radiation is comparatively low. 

 Isaeids and hyalids also dominate tropical shallows. 



The importance of limnetic environments to the development of 

 Amphipoda is also attested to by the species-flock of Hyalellidae in 

 Lake Titicaca, a lake now of alpine character in the tropics but a lake 

 perhaps of warmer climes in earlier times. It seems more than hap- 

 penstance that most other talitroideans (like the Hyalellidae) are now 

 strongly oriented to the tropics. They include the fully (and only) 

 terrestrial Amphipoda that occur mainly on the islands of the Indo- 

 Pacific region, plus the beachhoppers that also have primarily tropical 

 affinities. Only the aquatic genus Hyalella has successfully invaded the 

 Nearctic realm dominated by Gammaridae (which also dominate 

 Palearctic and occur in African (Ethiopian) realms. Talitroideans are 

 more specialized or more advanced morphologically than members of 

 Gammaridae in uropods, mandibles, and antennae and no one has 

 been able to trace any direct relationships between the two groups, 

 even though one may see in the Beaudettiidae a species presumably 

 derived from Elasmopus (Gammaridae) that has undergone a revolu- 

 tionary change in the direction toward Hyale (Talitroidea) . 



One may consider that by the time fresh-water Gammaridae had 

 become dispersed to a cordilleran pathway contiguous to South 

 America that the latter continent had become disjunct or that the cir- 

 cumferential Sonoran belt afforded a major barrier. There may be a 

 strong competitional stress between Gammaridae and Hyalel- 

 lidae for the northward migration of one hyalellid into the Nearctic 

 realm does suggest the presence of a migratory pathway now in exist- 

 ence, but only one species has managed to make the crossing. 



Dispersal of marine Gammaridae to the Neotropical realm also 

 must have come after the successful invasion by hyalellids, because it 

 would appear that even in tropical waters marine gammarids have been 

 able to invade limnetic environments. 



All through the ancient Tethyan Sea marine gammarids invaded 

 underground waters, forming the nipharigid and other troglobitic 

 groups. The South African genus Paramelita may have its origin in 

 marine melitids and have undergone convergence towards the Baika- 



