TUE DERMAL FIN-RAYS OE EISHES. 497 



This buried portion of the ray may be compared to the 

 bevelled part of a scale, which, is covered over by the next 

 scale in front, and both are devoid of ganoine (fig. 53). 



We must now study the interesting and aberrant genus 

 Cheirolepis, which, it will be remembered, is the oldest 

 known representative of the group. Cheirolepis is remark- 

 able for the possession of a covering of very minute scales, 

 so small that the skin resembles rather the shagreen of a 

 shark than the ordinary scaling of a fossil '' Ganoid." Yet, 

 judging by their microscopic structure, these dermal ossifica- 

 tions are not denticles, but merely ganoid scales of unusually 

 small size (figs. 47 and 49). The fins are provided with the 

 familiar fulcral scales at their front edge, and are supported 

 by lepidotrichia, the segments of which are very numerous 

 and minute. The segments of these va,ys are, in fact, of just 

 the same size and appearance as the neighbouring body- 

 scales, and, like them, are covered with a thick layer of 

 ganoine, and ornamented with the same superficial sculptur- 

 ing (figs. 47 and 49). Sections through the body-scales show 

 that they are provided with a rather shallow base of bony 

 substance containing bone-cells. Similar sections through 

 the lepidotrichia, taken near the body, show that the seg- 

 ments have a much more extensive bony base of identical 

 histological structure, but of much greater bulk and of more 

 rounded shape (figs. 46 and 48). Towards the periphery of 

 the fin the rays become flatter. At their pi'oximal ends the 

 lepidotrichia are continued beneath the body-scales in the 

 form of unjoiuted bony rods, some six or seven times the 

 leugth of a single outer segment. These unjoiuted pieces are 

 seen to be of oval section, without ganoine, and of exactly 

 the same nature as the inner part of the more distal region 

 of the dermal ray, of which they form, so to speak, the con- 

 tinuation. The overlying ganoid scales ai-e arranged in 

 regular rows, closely fitting the rays, and perhaps really 

 alternating with them, as seems to be suggested by the sec- 

 tion figured (fig. 47). But the exact relative position of these 

 superficial scales is very diflScult to determine in the fossils 



VOL. 47, PART 4. NEW SEEIES. I I 



