California, Peru and Chile, and Argentina, which, as I shall presently show, 

 constitute a distinct genus ; the even more strikingly pantemperate 

 jackmackerel {Trachums) and chub mackerel [Pneumatophorus) groups ; 

 the saury genus Scomberesox (of North Atlantic and South Temperate 

 waters) ; the species of Seriola related to 5. dorsalis ; Centrolophus ; 

 several tunas (Thunnidse) and tuna-like fishes ; also Alepisaunts and 

 various other rather pelagic types. 



As emphasized by Berg (1933), not only these fishes, but also various 

 other organisms almost certainly had a surface-water connection between 

 the northern and southern hemispheres at a geologically very recent time, 

 very probably, in large part, during the last period of continental 

 glaciation. Notable examples are sealions (Otariidae), right whales 

 (E^ibalaena) , and the giant kelp Macrocystis, which .is said to be repre- 

 sented in both hemispheres by the same two ecotypic varieties of a single 

 species. 



Zoogeographical, paleontological, and oceanographical data keep 

 accumulating to force the conclusion that the last period of glaciation 

 was synchronous in the two hemispheres and that the isotherms of sea 

 and land were then markedly displaced equatorward. The increased 

 winds from polar toward equatorial regions must have accentuated the 

 convergence of the cool currents, along the eastern shores of the two 

 great oceans. The tropical belt must have been narrowed and con- 

 siderably cooled everywhere. Paleontological data from Europe and 

 Africa, the occurrence of a land-locked salmon {Oncorhvnchiis) in Formosa, 

 and the ichthyological data in the following paragraphs are samples of 

 the evidence on which these statements rest. 



The paucity of the Panamanian marine fish fauna, particularly among 

 the reef fishes, suggests widespread extermination b}/ a marked cooling 

 of the tropical eastern Pacific. This fauna is outstandingly rich only in 

 certain famihes that characteristically inhabit mud and sand bottom, 

 notabl}' the drums (Sciaenidae) , marine catfishes (Ariidae), and anchovies 

 (Engrauhdidse). A decimated reef fauna would nicely explain the 

 obviously very recent establishment on the tropical American Pacific 

 Coast, particularly on the outlying islands, of Indo-Pacific fishes. The 

 broad intervening expanse of open water, stressed by Ekman (1935) as 

 the most notable barrier in the circumtropical distribution of life, has no 

 doubt allowed only rare strays to reach x\merica from Oceania. Had 

 they reached shores as saturated with life as are those of the Indo-Pacific 

 area it . is hardly conceivable that these strays could have established 

 populations. 



The former cooling of now essentially tropical waters, about the 

 Cape region of Baja California, is indicated by the entrapment, in the 

 upper part of the Gulf of California, of a rather large faunal element 

 related to that of north-western Baja California and southern California. 

 Most of the constituent types are no longer able to round the tip of the 

 peninsula. The varying differentiation of the Gulf representatives, 

 generally very slight and still unmeasured for some species, suggests the 

 origin of these upper-Gulf types during several periods of Pleistocene 

 global cooling, but chiefly during the last ice period. Somewhat similarly, 

 the occurence in the Galapagos Islands of many Temperate Peruvian 

 types, in part more or less differentiated, suggests transfer and 

 establishment during Pleistocene periods of general oceanic cooling. ' 



327 



