364 M. L. Agassiz on the Primitive Diversity and 



fill relations wliicli exist between the changes animals now living 

 undergo during their growth, and the order of succession of en- 

 tire classes of animals during successive geological ages. As 

 long as the natural position of Trilobites remained doubtful in 

 the animal kingdom, the characters of the prototypes of the class 

 of Crustacea could not be appreciated. Who does not see how 

 impossible it was for those who classified the Trilobites with the 

 Chitons to arrive at any sound results respecting the gradation 

 and order of succession of these animals ? whilst now they are 

 beautifully linked to the Macrura of the Trias, by the gigantic 

 Entomostraca of the Devonian and Carboniferous periods. Again, 

 the knowledge of the embryology of Crustacea gives us a key to 

 a correct appreciation of the early appearance of the Macrura and 

 the late introduction of the Brachyura. The removal of the 

 Bryozoa from among Polypi to the class of mollusks, will entirely 

 change the aspect and relations of the faunae of the palaeozoic 

 rocks. How different, again, would the order of succession of 

 mollusks appear, were we to adhere to Cuvier's view of sepa- 

 rating the Brachiopods, as a class, from the other Acephala, to 

 which they are now more correctly referred. The vexed question 

 of the period of appearance of Dicotyledonous plants in the geo- 

 logical series would have been settled long ago, had it been 

 placed upon its real foundation. It is not in reaUty to be argued 

 upon palseontological evidence chiefly, for it resolves itself in the 

 main into a botanical question, and the definite answer must de- 

 pend upon the position finally assigned by botanists to the fami- 

 lies of Coniferse and Cycadese. If these natural orders of plants 

 are really allied to the Dicotyledonese, then this type begins with 

 the palaeozoic rocks in the Devonian system, and there is no gra- 

 dation in the order of succession of plants during geological 

 times. But if the view of Brongniart is more correct, if the 

 Coniferse and Cycadese have to be separated from the Dicotyle- 

 donese as Gymnospermse, and if moreover these latter should 

 prove, as I believe they are, inferior even to the Monocotyle- 

 donese, then we may at once recognize in the vegetable kingdom 

 a similar gradation of types as among animals. These examples 

 may suffice to show what is required for a proper investigation 

 of the order of succession of organized beings in the course of 

 time, and how little confidence the investigations in this field 

 deserve, which have not been made with due reference to all the 

 points mentioned above. It is indeed only in the classes, the 

 structure and embryology of which are equally well understood, 

 that we are able to discover the laws regulating the succession of 

 animals and plants in geological formations, and our knowledge 

 is still too imperfect to carry the investigation into all families 

 of the animal kingdom. Yet enough is known to leave no 



