THE TWO MAIN LINES 467 



There are, however, other longipinnatids which are still rare iu the 

 Triassic (I refer especially to Toretocnemus, with ribs much divided 

 proximally), but which branch out in the Liassic and later form Stenop- 

 ierygius, including quadriscissus and many others. These animals are 

 very perfectly adapted to a swimming mode of life, no less than are the 

 contemporaneous latipinnatids. Throughout the Upper Jurassic forms 

 are found with very small paddles and exceptionally strong tail (the 

 functions of which compensate each other) of the group Ichthyo- 

 saurus euthecodon. I have proposed for them the new generic name 

 N anno pterygium. There is a form in the Middle Cretaceous described 

 by Broili as I. platydactylus, which has specialized in another direction. 

 This form has very broad paddles, with sesamoid bones at the anterior 

 and posterior borders. The tip of the tail was short and the animal was 

 probably therefore a very poor propeller. This genus I call Flatyptery- 

 gius. It is the latest latipinnatid known. 



Land tetrapods when first adopting aquatic life row with both pairs 

 of extremities, a condition seen in the Mesosauria. The better adapted 

 they become to that mode of locomotion, the more the function of the 

 hind limbs is emphasized (as, for instance, in the crocodiles). This is 

 still seen to some extent in one division of the Longipinnatidse, namely, 

 Cymhospondylus, the shastasaurids and Leptopterygius. Originally the 

 slender, long-tailed bodies of the unknown earliest ichthyosaurs prob- 

 ably swam with a natural serpentine motion. By and by this was found 

 useful for propulsion and became more localized in the flexible tail. In 

 the same degree the neck was shortened and the head enlarged. This 

 must have been true of all Ichthyosauria, but to a greater extent in the 

 branch which later became the Latipinnatidse. The function of the tail 

 must very early have become one of propulsion, and at its strongest point 

 a fin border ("Flossensaum") was acquired, later developing into a more 

 or less symmetrical diplocercal caudal fin for a propeller. For a long 

 time the paddles remained as organs of propulsion, and were used for 

 that purpose in most of the earlier longipinnatids, but in the Stenop- 

 terygius branch and in all later longipinnatids, as well as in all latipin- 

 natids, the hind paddles became reduced because of the perfect function 

 of the tail-propeller and the fore paddles became more and more en- 

 larged for vertical steering and for balancing. In the fore paddle of the 

 latipinnatids the number of digits is not reduced, as it is in all longi- 

 pinnatids. This may be explained by a differentiation in the very earliest 

 mode of adaptation, when the two main divisions of the ichthyosaurs were 

 just forming. In swimming, the upper wing of the diplocercal caudal 



