Manchester Memoirs, Vol. Lx. (1916), No. %. 17 



tively uncrushed, serve excellently for the study of the 

 trunk and of the paired fins, but none of those at our 

 disposal show the median fins. The pectoral fins were well 

 illustrated by J. Ward, and all the specimens which we 

 have seen agree with that which he has figured in having 

 pectoral fins closely applied to the ventral surface of the 

 body and inserted at the junction of the two rami of the 

 clavicle. They are short lobate fins, the scaled area 

 being rounded and about half the length of the fin. The 

 pelvic fins stand free from the body vertically in the 

 matrix. They also, are short Crossopterygian fins, pre- 

 senting in themselves no special features, but each accom- 

 panied by a special free scale or spine arising from the 

 fin root between the fin and the body and passing back- 

 ward. This spine is not part of the general squamation, 

 and may perhaps be an intromittent organ analogous to 

 the Elasmobranch clasper, but it is present in all specimens 

 which might be expected to show it, which have come 

 under our examination. The anus lies between the 

 pelvic fins, its position being indicated by a slight modi- 

 fication of the squamation. {PI. /., Fig. 2.) 



Our account of the median fins is founded mainly on 

 Williamson's small original specimen. The relative size 

 of this individual has been determined from its well 

 exposed operculum, but the known variation in proportion 

 of this bone suggests that the median fins shown in our 

 restoration are not necessarily of appropriate size, although 

 they cannot be greatly in error. Their structure will be 

 best understood from the figures, but attention may be 

 called to the occurrence of an elongated ridge scale 

 behind each dorsal. 



Individuals of Rhizodopsis sauroides differ very con- 

 siderably in size, our smallest fish having been about 

 fifteen centimetres long, and our largest nearly a metre. 



