DR. GUNTHER ON THE STUDY OF FISHES. 385 



colony of Queensland, accounts were given of the existence in the 

 rivers of the country of a large fish which the colonists called a 

 " salmon," from the fact of its having salmon-colored flesh, and of 

 sometimes, it was said, rising to a fly. Mr. Krefft, Curator of the 

 Australian Museum, at Sydney, having examined a specimen of this 

 fish in 1869 or 1870, pronounced it to be allied to the mud-fish, but 

 discovered also that it possessed teeth so closely resembling those of 

 certain fossil fishes attributed by Agassiz to the sharks, that no doubt 

 could be entertained of the generic identity of the two forms. Ac- 

 cordingly, the newly found animal was described as a species of 



Ceratodus that being the name which Agassiz had conferred on the 

 creatures whose fossil teeth he had long before made known. Now, 

 all this was in itself sufficiently remarkable, for it proved that Cera- 

 todus, as a genus, had persisted from the Mesozoic era ; but its im- 

 portant bearing was not fully perceived till after some more examples 

 had been obtained and sent to Dr. Giinther. He described the recent 



Ceratodus in great detail in the " Philosophical Transactions " for 

 1871, and was able, furthermore, by its means, to show how the palaeo- 

 zoic as well as the recent ganoids, the sharks and skates both ancient 

 and modern the sturgeons, the mud-fishes, with some other forms 

 that had hitherto been irreconcilable, could all be brought together 

 through some essential characters common to the whole of them, and 

 harmoniously placed in a single class, to which he assigned the name 

 of " Palseichthyes " fishes of ancient type. Professor Huxley had 

 previously pointed out the affinity of the mud-fishes to certain ganoids, 

 but the credit of discovering this comprehensive classification is due to 

 Dr. Giinther. 



Dr. Giinther does not undertake to describe definitely the geograph- 

 ical distribution of fishes in the sense in which the term geographical 

 distribution is used by naturalists of other branches, he holding that 

 " the endeavor to establish, by means of our present fragmentary geo- 

 logical knowledge, the divisions of the fauna of the globe leads us 

 into a maze of conflicting evidence." It is obvious that fishes are not 

 amenable to the laws of geographical distribution which govern land- 

 animals. In treating of their relations in this respect it is, moreover, 

 necessary to separate them into categories, of which Dr. Giinther 

 makes four: 1. Fresh-water fishes ; 2. Brackish-water fishes ; 3. Marine 

 fishes, which are furthermore subdivided into shore-fishes and oceanic 

 or pelagic fishes ; and, 4. Fishes of the deep sea. Even in the case of 

 fresh-water fishes, which of course live under conditions more similar 

 to those of land-animals than do those of the other categories, he dis- 

 allows the six great zoogeographical regions which most geologists 

 have accepted, and would arrange them in three zones Northern, 

 Equatorial, and Southern. These zones, are, however, broken up into 

 regions, which roughly correspond with the six generally received, ex- 

 cept that the Australian region of most zoogeographers is split up into 



TOL. XXI. 25 



