28 J. 0. Merriam — Triassic Ichthyosauria. 



Longipinnati, such as Ichthyosaurus tenuirostris. - The ante- 

 rior digit was considered as representing digit II of the primi- 

 tive limb, digit I having disappeared. Latipinnate forms, such 

 as I intermedins, were supposed to be more specialized, the 

 additional digit in the middle of the hand having been pro- 

 duced by the splitting of digit III. The longipinnate group 

 would then be the more primitive and the latipinnate forms be 

 derived from it by intercalation or splitting of digits.* 



Strongly suggestive of the latipinnate and longipinnate pad- 

 dles we find also in the Triassic genera a broad and a narrow 

 type, the broad form occurring in Mixosaurus, the narrow 

 form in all of the Californian genera. In the narrow type the 

 limb is even more reduced than in the Longipinnati and is 

 really tridactyle. So far as can be determined, Toretocnemus 

 seems to be the most primitive of these forms. The rudimen- 

 tary fourth digit is larger than in the others and the third 

 digit is as large as the first. In Delphinosaurus the third digit 

 is much smaller than the others and in Shastasaurus osmonti 

 it was probably smaller than in Delphinosaurus. This series 

 showing progressive reduction of the posterior side of the limb 

 indicates that the type is probably not a primitive one, but is 

 derived from an earlier form with five digits. 



While we can understand the origin of the narrow type of 

 paddle in the Trias, the broad form is not so easily explained 

 if we hold that digit number one has disappeared. Mixo- 

 saurus had five digits of nearly equal size and made up largely 

 of the shafted or primitive type of phalanges. The extremi- 

 ties of this form were, however, already specialized paddles, 

 and, if the interpretation of the structure of the paddle of 

 Ichthyosaurus given above is correct, we shall have to suppose 

 that in this form digit I was lost and another digit added. 



If finger I in Mixosaurus corresponds to primitive digit IT, 

 the added digit is either below the intermedium or on the pos- 

 terior border of the limb. The presence of a supernumerary 

 ossicle behind the carpus seems to give support to the idea 

 that the last digit is not primitive, as elements of this charac- 

 ter are known to develop secondarily, particularly on the pos- 

 terior borders of the limbs of aquatic forms. Their presence 

 does not^ prove the case, however. The supernumerary ossicle 

 may be secondary and the digit primitive, or the presence of 

 the ossicle may be due to upward movement of the last digit 

 along the posterior side of the carpus. Such movements have 

 occurred frequently in Ichthyosaurian paddles, regardless of 

 the theory by which we account for them. 



To suppose that one of the digits below the intermedium is 



*Mr. Lydekker has recently expressed himself as in accord with the views 

 concerning the primitive character of the Mixosaurian paddle which are 

 presented in this paper. See p. 29. 





