GREGORY, PRESENT STATUS OF ORIGIN OF TETRAPODA 357 



of all fins were short and scale-like ; the teeth were conical, but of small 

 size. In general appearance, except for the shorter body, these primitive 

 Ehipidistia may have been similar to Osteolepis. 



Tarrasius proMematicus Traquair (1881, 1890) from the Lower Car- 

 boniferous of Scotland is usually referred to the Crossopterygii and placed 

 before the Ehipidistia in a separate order or suborder Haplistia. It is 

 a very small fish (about 3 inches long) with a continuous dorso-caudo- 

 anal fin and obtusely lobate pectorals; at least the posterior part of the 

 body is covered with small granular rhombic scales; the anterior region 

 of the trunk may have been naked ; the notochord is persistent ; there are 

 neural and haemal arches ; the median fins are supported by rods, which 

 are more numerous than the vertebral arches. Tarrasius was provision- 

 ally referred by Zittel to the Dipnoi, but Traquair (1890, p. 494) said 

 that ''^the obtusely lobate character of the pectoral fin seems to point 

 toward the Crossopterygii." Goodrich (1909, p. 284) states that the 

 "dermal bones of the skull and operculum appear on the whole to re- . 

 semble those of Osteolepids." 



The interest of this fish in the present connection lies in its continuous 

 median fin. By the upholders of the "fin-fold" theory it is assumed that 

 ancestral fishes once had both the paired and the median fins continuous. 

 But in the Actinopterygii continuous median fins are invariably a sign 

 of aberrant specialization; and the frequently copied reconstruction of a 

 long-bodied "primitive" fish with a continuous fin-fold bears a suspicious 

 resemblance to such highly specialized types as Fierasfer, Murcena or 

 Gymnotus. Tarrassius may or may not be a crossopterygian ; but I find 

 no evidence for believing that its continuous dorsal became subdivided 

 into the two dorsals of Ehipidistia. 



The point is that while paired fins and median fins probably arose in 

 the same way, they may well have been purely local outgrowths like the 

 dorsal fin of ostracoderms. It is not necessary to conclude that the two 

 dorsal fins of Ehipidistia and the single dorsal fin of primitive Actino- 

 pterygii arose by subdivision or abbreviation of an originally continuous 

 dorso-anal fin. The Actinopterygii, Ehipidistia and Dipnoi may well 

 represent parallel offshoots from primitive short-bodied gnathostomes 

 that had no continuous dorso-anal fin at all, but only low outgrowths of 

 the skin or of the body-wall, placed at nodal points of mechanical ad- 

 vantage in securing a "purchase" against the water. 



Tarrasius prohJematicus is also of special interest because, if it is a 

 primitive crossopterygian, it may also be related to the stem of the 

 Tetrapoda, retaining perhaps the lobate fins that gave rise to the cheirop- 

 tergyia, retaining in part a primitive granular scalation, and developing 



