THE EXTINCTION OF SPECIES AND OF GROUPS 233 



the regular Urchins of the family of the Cidaridae in 

 the ancestral forms very near akin at least to the 

 Archceocidaridce of the end of the Primary, though 

 the multiplicity of their rows of interambulacrary 

 plates have led to classing these latter zoologically 

 in a sub-class of the Palechinida, which is in ap- 

 pearance quite separate from the Euechinida. 

 Thus we are beginning, thanks to the researches of 

 Hyatt and other palaeontologists, to recognize, 

 in the varied types of the Palaeozoic Goniatites, the 

 origin of several branches of true Ammonites, till 

 then classed in a sub-order of Prosiphonata quite 

 distinct from the Retrosiphonated Goniatites. 

 Finally, to quote a last example among the Verte- 

 brates, the Crocodilians were separated by Huxley 

 into two sub-orders, the Mesosuchians of the 

 Jurassic and the lower Cretacean, characterized 

 by bi-concave or amphiccelian Vertebrae and the 

 Eusuchians of the upper Chalk and of the Tertiary, 

 whose vertebrae are convex behind or proccelian. 

 But Lydekker has shown that it would be more 

 rational to separate the true Crocodilians into two 

 great parallel branches easy to characterize and to 

 trace starting from the Lias, and called the Brevi- 

 rostres and the Longirostres, each of them comprising 

 Mesosuchian forms in the Jurassic and the lower 

 Cretacean, and Eusuchian forms from the upper 

 Cretacean onward. The separation proposed by 

 Huxley was, therefore, artificial, and rested on 

 a stage of ossification of the vertebral column less 

 advanced in the Jurassic Crocodiles than in their 

 Cretacean, Tertiary, or existing descendants. 



Notwithstanding these restrictions, it is quite 



