204 



SCIENCE 



[N. S. Vol. XXVII. No. US ; 



kindred, which are present and even abun- 

 dant, are more and more shark-like in den- 

 tition, fins and dermal defenses the further 

 back we go. To maintain in the face of 

 such evidence that a ceratodont existed in 

 the Urzeit would be, it seems to me, as diffi- 

 cult as, for example, to maintain the prob- 

 able presence of man in the Jurassic. 



The puzzles of primitive sharks have 

 been considered several times during the 

 past year, and I think it is quite safe to 

 say on cumulative evidence that "if the 

 earliest true fish could be found, it woiUd 

 almost certainly fall within the subclass 

 Elasmobranchii. "^ From a recent study 

 on the structures of the lowly aeantho- 

 dians* we have reason to believe that they 

 are allied more closely to the more typical 

 sharks. Their dentition was quite shark- 

 like, with (a few) successional rows of 

 teeth, and their fin structures conform 

 more typically to the plan known in elado- 

 donts. In this connection the evolution of 

 their curious finfold type of fin in different 

 members of this group has been indicated 

 lately by Smith Woodward. Of other 

 structural features in acanthodians we 

 know even details, e. g., in sensory canals 

 and ear structures. As vertebrate mor- 

 phologists some of us wish we could be- 

 lieve that the acanthodians, earliest of 

 sharks, had a great number of gill slits, 

 but for the present we shall have to content 

 .ourselves with the typical selachian num- 

 ber, five: we wish also that we could feel 

 assured that the mandibular arch in acan- 

 thodians was segmented dorsally after 

 the fashion of a typical gill arch, as 

 Reis and Jaekel have shown in the Permian 

 AcantJiodes bronni, but unhappily certain 

 earlier genera (the lower Devonian 

 Ischnacanthus gracilis, Cheiracanthus mur- 



^ Smith Woodward, " Natural Science," Vol. VI., 

 p. 38. 



* Dean, " Notes on Acanthodian Sharks," Am. 

 Jour, of Anatomy, Vol. VII., pp. 209-226. 



chisoni) do not have this interesting sub- 

 division of the palatoquadrate, and it may 

 well be, therefore, that in later forms the 

 subdivision is due to fossilization, the car- 

 tilage subdividing, owing to weakness from 

 nutrient vessels, etc., at certain definite 

 spots. 



An up-to-date classification of the se- 

 lachians by Tate Regan should here be 

 mentioned^ which gives interesting notes 

 as to the evolution of the sharks and con- 

 siders the paleontological evidence. 



The chimgeroids have been made the sub- 

 ject of comparative study during the past 

 year," and from an examination of their 

 fossils, anatomy and embryology the con- 

 clusion is reached that they are to be 

 classed not as ancestral sharks, but rather 

 as a group highly divergent from some 

 early shark stem. The few undeniably 

 primitive features which they possess are 

 heirlooms from some Paleozoic selachian 

 ancestor— features which modern sharks 

 have not as well conserved owing, among 

 other causes, to the elaboration of hylos- 

 tylism. The nearest living kin of the 

 chimsroids are probably cestraciont sharks. 



Important in this connection is the dis- 

 covery that our pre-Permian "chimsr- 

 oids," i. e., those antedating menaspids, 

 may have to go by the board. Jaekel's 

 discovery of associated remains of Rham- 

 phodiis, as Dollo points out,'' makes it prob- 

 able that all ptyctodonts, hitherto classed 

 as chimseroids, are in reality highly modi- 

 fied arthrodires! 



Our knowledge of the descent of lung- 

 fishes has not progressed perceptibly dur- 

 ing the past year — i. e., if we admit that 

 the Arthrodira and early lung-fishes are 

 not related, the view which we have main- 



» Regan, Pro. Zool. Soc. London, 1906, pp. 722- 

 758. 



" Dean, " Chimseroid Fishes and their Develop- 

 ment," Monograph 34, Carnegie Institution, Wash- 

 ington, D. C. 



'Bull. Soc. Beige de Geol., 1907, pp. 97-108. 



