344 REPORT OF NATIONAL MU&EUM, 1892. 



waters that geologists usually regard the discovery of remains of simi- 

 lar fishes in any given geological formation as evidence of its marine 

 origin. Still, members of certain of these families present remarkable 

 exceptions to the general rule thus indicated; such, for example, as the 

 presence of Selachians in Lake Nicaragua and other fresh waters. 

 Therefore, in case the fossils associated with such lish remains should 

 distinctly indicate the fresh-water origin of the strata containing them, 

 their presence may be held as not necessarily constituting conflicting 

 evidence because of the known exceptions to the rule that their living 

 congeners have a marine habitat. 



Of the remaining eleven families mentioned in the foregoing notes 

 three are more abundantly represented in fresh than in saline waters. 

 To these the converse of the foregoing remarks will apply. Because of 

 the varying range of habitat of the remaining eight families mentioned 

 in the foregoing notes tlw discovery of remains referable to any of 

 them in a given formation would be of little value as evidence in deter- 

 mining the character of the water in which it was deposited unless 

 supported by other and more definite evidence. 



The Marsipobranchii and Leptocardii are by some authors included 

 in the class of fishes proper, but Dr. Gill and others regard them as 

 separate classes coordinate with tishes, reptiles, etc. 



Two of the three families belonging to the Marsipobranchii are known 

 only in marine waters. Most of the members of the other family, 

 namely, the Petromyzontidae, range from marine into brackish and 

 fresh waters, as is well known in the case of the lampreys. 



No representation of the Leptocardii are known in other than marine 

 waters. 



This review of the tishes is confined to those families which have 

 living representatives, and the criteria relating to the different kinds 

 of aquatic habitat of fossil fishes which may be based upon this review 

 apply directly only to the families here named. 



Fossil remains of a large number of kinds of fishes have been dis- 

 covered, especially in the paleozoic formations, which differ so much 

 from all living kinds that they can be referred to no family, and some- 

 times to no order, which has living representatives. 



The character of the water in which such tishes lived might be con- 

 jectured by reference to their most nearly related forms, but the most 

 reliable indication is furnished by such other fossil remains as may be 

 found associated with them. 



MOLLUSCA.* 



The Mollusca are of peculiar importance in connection with the sub- 

 ject of this review, because the greater part of the members of this sub- 

 kingdom have aqueous respiration, and because in the matter of 



* The classification of the Mollusca used in this review is that of Dr. Theodore 

 (Jill's "Arrangement of the Families of Mollusks." See Smithsonian Miscellaneous 

 Publications, No. 227, 1871. 



