772 THE AMERICAN NATURALIST. [Vor. XXXVI. 
is represented in the case of the ancient ganoids (crossop- 
terygians); for in certain Devonian forms (cf also Mesozoic 
genera) the dorsal fins show clearly concrescence of radial and 
basal elements in a way which parallels strikingly the biserial 
archipterygium (of the paired fins)! This condition, although 
well known to the paleontologist, appears to have escaped the 
notice of the adherents to the Gegenbaurian school. It is per- 
fectly evident that if the biserial archipterygium arises in the 
unpaired fin from the concrescence of lateral fold-like support- 
ing elements, the similar structure in the paired fins must have 
arisen in a similar way, or, to say the least, the burden of dis- 
proof lies with those who believe that such closely similar 
structures arose by diametrically opposite processes. 
IV. CONCLUSION. 
The evidence of Paleozoic sharks, then, is most distinctly in 
favor of the Thacher-Balfour teachings. The oldest forms, 
acanthodians and cladoselachids, represented by the best- 
preserved fossils present lateral-fold fins. Moreover, there is 
evidence that as the series advances from the Lower Devo- 
nian, the structures of the biserial archipterygium are gradu- 
ally acquired. The Carboniferous Cladodus meilsomi, Fig. 1, 
of Traquair shows for the first time a definite segmentation 
of the supporting elements of the base of the pectoral fin? 
Then in the Permian Symmorium Cope, the fin bases not only 
are formed, but show apparent fusion in the metapterygial 
region, and even perhaps an extension from the body wall of the 
metapterygial terminal, a condition which would be best corre- 
lated with a change of function in the fin. And finally, in the 
xenacanthids, whose structures are known only in this later 
_ horizon, the pectoral fin acquires a biserial archipterygium. The 
conditions, however, of Cladodus neilsoni and of Symmorium 
1 Cf. also the anals of Xenacanthus. 
? Judging from the size of the shark and the character of the radials, the 
pectoral fin was probably larger, and there is thus ground to believe that the 
posterior radials of the fin have been lost. This is better understood by reference 
to the accompanying Fig. t, where the Scottish specimen is shown overlaid with 
an outline of Cladoselache of similar size. 
