498 EDWJN S. GOODRICH. 



at my disposal. Some appenrances have led me to believe 

 tliat tliey are fused togethei* in parallel rows along the top of 

 the loiig proximal segment of the dermal rays. At all events 

 they are set very close together (fig. 49 a). It was, perhaps, 

 such an arrangement as I have just described in Cheirolepis 

 which led Agassiz to believe that the fins are provided with 

 dermal rays lying below more superficial scales. It must be 

 remembered, however, that the elongated proximal segment 

 of the lepidotrichia is situated, not in the fin, bnt in the 

 body itself, and probably overlapped the cndo-skeleton. 

 Pander (28) fully adopts the view of Agassiz when describing 

 Cheirolepis. From his figures and statements there can be 

 no doubt that he fell into the mistake mentioned above. 

 Traquair, on the contrary, considers that the rays consist of 

 one layer of small segments only (35). The interpretations 

 of these two authors can, however, be reconciled if we assume 

 that Pander examined the proximal end of the rays, and 

 Traquair the more distal i-egion in the fin itself. 



On comparing Cheirolepis with Pal^oniscus the simplest 

 conclusion wonld be that the proximal elongated joint buried 

 beneath the scales in the former, is homologous with the 

 proximal but only partially covered joint of the ray in the 

 latter. If this basal joint, which in Palieoniscus is still 

 covered with ganoine at its outer exposed end, were to 

 become longer and to sink deeper beneath the body-scales, 

 we should obtain the structure found in Cheirolepis. The 

 proximal joint of the Cheirolepis ray is rounded in section, 

 and exhibits unmistakable signs ot" concentric layeis of 

 growth [Gg. 47), a fact which raises some doubt as to its 

 original derivation from a superficial scale ; but there is 

 a general tendency, so to speak, for the deeper ends of the 

 lepidotrichia to acquire this concentric structure, which is 

 visible to some extent in the more distal segments them- 

 selves and in the scales (figs. 4G and 47). 



Another interpretation is, indeed, possible — namely, that 

 these dermal rays are compound structures formed by the 

 fusion in the fin region of deep rays, homologous with the 



