M. Agassiz on the Classification of Fishes. 139 



it will be of importance to notice how this class is connected 

 with the Vermes by certain genera of the order of the Holo- 

 thurias, and with the Polypes by means of the Crinoides. If 

 we wish to divide the Crinoides in the most natural way, we 

 must insist on the analogy of the Echinocrines, for example, 

 with the true Echinides, and on that of the Comatules with the 

 Asterioe, while the true Encrines will form the central type of 

 the order ; and so on in succession. And in order to combine 

 the indications relating to the affinities of one class with those 

 which we possess respecting their succession, it will be neces- 

 sary to add to these zoological charts, as they may be called, 

 genealogical trees, on the trunk of which the most ancient ge- 

 nera will be inscribed, while the branches will bear the names 

 of the most recent types. By properly managing the propor- 

 tions of the trunk and branches, and making them of suitable 

 dimensions, we may even indicate exactly the period when 

 each group appeared, by giving to the different branches of 

 each order a degree of thickness proportionate to the import- 

 ance of the part which the types they represent have occupied 

 in each geological formation. 



It is in accordance with these principles that I have con- 

 structed the annexed table, which represents the history of 

 the development of the class of fishes across all the geological 

 formations, and which expresses, at the same time, the degrees 

 of affinity which the different families have to each other.* At 

 the top of the figure are inscribed the names of the four orders 

 into which I divide this class, and the characters of which are 

 discussed in the Becherches sur les Poissons Fossiles. These are, 

 the orders Cycldides, Ctendldes, CauQlflcs., and Placoules. Below 

 these appear the names of the families which have represen- 

 tatives in the presently existing creation. They are arranged 

 vertically, to correspond to the ascending lines, more or less 

 strongly marked, which indicate, by their lower extremity, the 

 point of departure in the development of the families, and, by 

 their breadth, the degree of importance which they possessed at 

 each epoch. On the sides of the table are inscribed the names 

 of the principal formations, in order to indicate the geological 

 Jievels from which all the families take their origin, as well as 



* The Table will be given in our next number. 



