M. Agassiz on the Classification of Fishes. 137 



order, and even each family, follows a particular progression ; 

 that there is, in regard to each group, a beginning and an apo- 

 gee in its development ; that by turns they terminate by be- 

 coming extinct, if they go back to a remote antiquity, or by 

 acquiring a considerable extension in the present creation, if 

 their appearance dates only from a recent epoch. These re- 

 sults, which are so evident in the class of fishes, I have, in like 

 manner, established in that of Echinodermata ; and although 

 I have not yet given a detailed explanation of the general re- 

 sults of my studies respecting these animals, I can yet affirm 

 that I have recognised among them the same laws of develop- 

 ment. To be convinced of this, it is enough to remember in 

 what proportion we find the Crinoides and star-fish in the series 

 of formations, and what is the condition, in narrower limits, 

 of the diff*erent families of the order of Echinites. With such 

 results before us, we are naturally led to suppose that it is the 

 same with the other classes of the animal kingdom ; and that, 

 if we have not yet succeeded in seizing everywhere the thread 

 of their development, it is because we have not found out the 

 key to their connection. We already possess, in regard to all 

 the classes, positive indications of this preponderance, at de- 

 terminate epochs, of certain types, which change proportion 

 with their cotemporaries belonging to more recent eras ; for 

 example, among the Mammifera, the Pachydermata, the Eden- 

 tata, the Marsupialia, and the Quadrumana ; among reptiles, 

 the Ichthyosaures, the Plesiosaures, the Megalosaures, the 

 Ophidians, and the tailless Batracians ; among the Crustacea, 

 the Trilobites ; among the Cephalopoda, the genera with par- 

 titioned shells, whose development is most remarkable, from 

 the Orthoceratites and the Goniatites, down to the singularly 

 plicate enrolled or straight forms of the Scaphites, the Ancylo- 

 ceras, the Cyrtoceres, the l^tychoceres, the Turrilites, the He- 

 licoceres, and the Baculites. Among the Acephales, may we not 

 point out facts in every way similar between the Brachiopodes 

 and the Lamellibranches ? And is it not a very significant fact, 

 that we observe this regularity in progi-essive development shew 

 itself in a manner so much the more evident, as we endeavour 

 to find the marks of it^ in the best known classes \ From 

 this consideration, very powerful arguments may be advanced 



