830 DR RAMSAY H. TllAQUAIR ON FOSSIL FISHES COLLECTED BY THE 



smaller specimen from the same horizon and locality, in which both lobes of the caudal 

 fin are seen, but the general form of the body is somewhat obscure. The length of this 

 specimen is 5j inches. 



In fig. 3 we have a small example from the Dovvntonian beds of Seggholm, which 

 gives a very good view of the general shape of the entire creature, the pectoral fin-flaps 

 and the deeply-cleft heterocercal tail being very conspicuous. Fig. 4 is a still smaller 

 specimen from Monk's Burn, in which the tail is also very well exhibited, but the pectoral 

 flaps are rather drawn in. 



The length of the head and body to the posterior origin of the pectoral fin-flaps is 

 contained on an average rather less than three times in the total, while the caudal fin, 

 with its two lobes, occupies also rather less than the terminal third of the fish. 



In some specimens, as in fig. 3, the apices of the pectoral fins seem rounded, but in 

 most, as in fig. 4, they are angulated, and this latter condition I take to represent the 

 normal and unaltered contour. As to the caudal fin, both lobes are pointed, and the 

 upper one is longer than the lower. 



Owing to the state of preservation of the specimens, there are few opportunities of 

 making out the exact form of the scales, an obscurely granulated surface being very 

 often all that is shown by the dermal covering. Careful examination, however, of a 

 considerable number of specimens reveals not only the shape of the scales, but the fact 

 that the scales on the head are of a different form from those further back. I was 

 fortunate enough to find both forms of scales together in an iron-stone nodule from 

 Logan Water, imbedded in a softish material, considered by Mr B. N. Peach to be 

 coprolitic in its nature, and with a little trouble I succeeded in isolating the scales repre- 

 sented in PI. I. figs. 6-9. On comparing these with the scales and impressions of scales 

 which are now and then to be seen in the entire fishes themselves, there can be no doubt 

 as to the identity of the species. 



The anterior or head-scales (figs. 5-7) have the regular Thelodus form, with well- 

 developed base and conspicuous basal opening (fig. 6) ; a smooth, slightly convex crown 

 (fig. 5), separated from the base by a constricted neck, as seen in fig. 7. It will be seen 

 that the edges of the upper surface show eight or nine shallow crenulations or crimp- 

 ings, which are continued down the neck as a sort of "fluting," very much as in the 

 scales of Th. Pagei. 



Fig. 8 represents one of the posterior scales seen from above, and fig. 9 a similar one 

 seen from the side. They can be absolutely identified with those which are seen beauti- 

 fully preserved in situ in some parts of the fish represented in fig. 1, and whose arrange- 

 ment is represented in fig. 10. 



These scales are very small, the examples figured measuring only -^ inch in long 

 diameter. The crown is rounded in front, pointed behind, where it projects backwards 

 over the basal part, — convex above, and longitudinally ribbed in such a manner as to 

 remind one of some small umbelliferous seed. The usual mode of ribbing is this: — Each 

 side of the upper surface is constituted by a marginal ridge, inside which and at a some- 



