viii 



from may not always be actually taken cognizance of in the diagnoses of the 

 groups, they more or less influence the adoption of groups characterized by 

 modifications of such parts. But it is only within certain limits that these 

 modifications are indicative of affinity ; often, for example, only recalling 

 ordinal relations determined by the number of the bones and their devel 

 opment. If, in many other cases, the nearer relations of forms have been 

 correctly inferred, it is rather from the tact which practice confers on the 

 student and the suggestions furnished by modifications which may be of 

 slight moment apparently, but which, on account of eccentricity or other 

 cause, strike the observer and often yield true clews to affinities. It is 

 logically, although the premises might be strenuously disavowed, the result 

 of a quasi-adoption of the doctrine of evolution, and the assumption that 

 certain characteristics peculiar to and common (but perhaps only in part) 

 to certain forms, especially when non-adaptive, are indicative of community 

 of origin, and therefore of immediate affinity. Such combinations are 

 often indefinable at first, but are frequently justified finally on a complete 

 study of the anatomy. But those combinations, when not definable, cannot 

 be considered as established, and are deservedly open to suspicion. The 

 author for many years has been collecting the skeletons and especially the 

 skulls of fishes, and their study has assured him of the affinities of many 

 forms whose relations would otherwise have been very doubtful. He has 

 meanwhile been anticipated in the announcement of certain of the results 

 of his studies by Prof. Cope, who has been fortunate in being able to 

 avail himself of the largest collection of skeletons of fishes known to exist. 



CLASSIFICATION. 



At a future time the views of the author respecting the principles of 

 classification and their application to the fishes will be published in detail. 



At present, it need only be stated that he entirely concurs with Prof. 

 Cope in the view that under the general term &quot; Fishes,&quot; three perfectly 

 distinct classes (PISCES, MARSIPOBRANCHII, and LEPTOCARDII) are con 

 founded, and he is inclined to agree with Prof. Hackel in the recognition 

 of even wider and certainly more obvious gaps between the typical fishes 

 and the two inferior classes than between any other contiguous classes of 

 vertebrates, but he cannot, with the latter naturalist, admit the title of the 

 Dipnoi to classical rank. As he urged in 18G1, 1 the Dipnoi and Polyp- 

 terids (Crossopterygia, Huxley) exhibit so many characters in common 

 that they cannot be very widely separated, and are not even entitled to 

 subclassical distinction. 



1 GILL (Theodore Nicholas). Catalogue of the Fishes of the Eastern Coast of North 

 America, .... [Philadelphia, The Academy of Natural Sciences, 13C1,] pp. 12-20. 



