1871.] 329 [Wilder. 



the entire habit of fishes, and with their mode of life, they best indi- 

 cate their natural affinities, and indeed prove to be the most constant 

 and general characters." 



As to generic criteria, Miiller and Henle enumerate 1 the charac- 

 ters found by them most useful among selachians; and Parker is 

 explicit respecting the unimportance of certain characters, for the 

 determination of groups more comprehensive than genera. 2 



Specific characters of the Pycnogonidee are enumerated by H. D. 

 S. Goodsir, 3 and those of the tortoises by Owen (62, 1, 162.) 



Finally, a great part of Agassiz's later works (200 and 201), is de- 

 voted to the effort to show not only that groups really exist in na- 

 ture, but that they are based upon distinct " categories of structure." 

 I quote the following also from my notes of his lectures on Selach- 

 ians. 4 " Zoologists take very different criteria or different parts as 

 foundation for the same kind of group, or the same criteria for dif- 

 ferent kinds of groups, so that their results are very diverse. We 

 must have some means of determining the value of characters." 



Accepting provisionally Agassiz's abstract enunciation of these cri- 

 teria and then* subordination as to value, as summed up on page 261 of 

 201, and likewise considering the only direct application of these prin- 

 ciples to a single group, the Testudinata and its subdivisions (200, 1> 

 Part n), I have endeavored to translate the zoological criteria into 

 anatomical language, and in this way to at least indicate the means 

 by which we may sometime be able to determine the exact morphi- 

 cal value of any anatomical character. The conclusions which I 

 reached are given in the diagram (page 179), and afterward briefly 

 explained ; but I must here admit that I feel sure of being right upon 

 only the following points : 



1. That both plan of structure and form are displayed upon a 

 vertico-lateral section of an animal. 5 



1 Ann. and Mag. of Nat. Hist., 1844, pp. 1 and 4. 



2Proc. Zool. Soc, 1863, p. 572. 



s Ann. of Nat. Hist., July, 1844, p. 1. 



4 Given at the Museum of Comp. Zool., 1867 - 1868. 



5 As between Vertebrata and Eadiata, or between either of these and the Mol- 

 lusca, and Articulata this is clear enough ; but since the relative positions of di- 

 gestive, nervous and circulatory systems seem nearly identical in the two latter 

 branches, the respiratory and perhaps some other systems must be included in our 

 representation of a vertico-lateral section. See Huxley's diagrams, 151, fig. 30. 

 As to the view that Vertebrata and Mollusca may find connecting links in Amphi- 

 oxus and the ascidians (references to which are given in 336). I have not yet seen 

 any comparison of the vertico-lateral sections of these animals, or any statement 

 that they are identical. 



