722 SCIENTIFIC RECORD FOR 1883 



rosouiids, many Cyprinodontids, and most Gasterosteids ; and (3) one 

 (the eel) perhaps should be considered as a salt rather than a fresh water 

 species, inasmuch as it is catadromous, and appears to breed only in the 

 sea. Conversely, those fishes which resort to fresh water to spawn and 

 speud their early days therein may be considered to be fresh- water forms. 

 If all species which to some extent run up into fresh water were in- 

 cluded, the list might be very greatly increased. 



Number of North American fishes. — A much needed work was com. 

 pleted and published during 1883 (although dated 1882), under the title 

 " Synopsis of the Fishes of North America," by David S. Jordan and 

 Charles H. Gilbert : it is the sixteenth " Bulletin of the United States 

 National Museum." Bearing the same title as a work published in 3846 

 by Dr. D. H.Storer, it not only contrasts with the latter in fullness of 

 details and as an epitome of all tbat has been done in North American 

 ichthyology, but is greatly superior in the mode of treatment of its sub- 

 ject, and is truly a work of eminent scientific merit. It embraces within 

 its scope all the species which visit any part of our extended coast, as well 

 as those ranging to the farthest north, and the inhabitants of our inland 

 rivers and lakes ; it does not, however, include any of the West Indian 

 fishes, except those which touch on our coast, and thus has a more re- 

 stricted aim than Dr. Storer's work. The classification adopted by the 

 authors "is essentially based on the views of Professors Gill and Cope, 

 who have," it seemed to our authors, "been more fortunate in reflecting 

 nature in their groupings of the fishes than have any of the European 

 systematists." They commence with the lowest or most generalized 

 forms, and successively take up the more specialized. The four classes of 

 Leptocardians, Marsipobranchs, Elasmobranchs, and Fishes proper are 

 adopted. The true fishes, so far as the North American species are con- 

 cerned at least, are subdivided into the "series Ganoidei," with the 

 "subclasses" Chondrostei and Holostei, and the "series Teleostei" with 

 the "subclasses" Physostomi and Physoclisti. The last two "sub- 

 classes" seem to be unnecessary, or, rather, not entitled to such rank, 

 for they not only intergrade, but the presence or absence of the duct 

 may be of minor importance. For example, the duct is not obliterated, 

 it has been urged, in such forms as Holoeenlrum, Priacanthus, Gcesio, &c, 

 and inasmuch as the bladder is developed as a diverticulum from the 

 intestinal canal, it is ever liable to resume the evidence of such origin 

 in the persistence of the duct. Twenty-three orders are recognized for 

 all the North American fish-like types, from the " Cirrostomi" upwards. 



Reduced to their several elements, there are along our east coast 

 about 436 species ; from the Gulf coast about 307 have been obtained ; 

 but it must be remembered that the latter region is comparatively but 

 little known ichthyologically. From the west coast 310 species have 

 already been secured, which contrast remarkably with the few that were 

 alone known to Dr. Storer in 1846. As many as 617 species have 

 been attributed to fresh-water, and of exclusively fresh-water types 



