ON THE ANIMAL KINGDOM. 203 



the least perfect of the V ertebrata, or passes imperceptibly 

 into the neighbouring groups. 



■ Thus in the fishes which compose the genera Ammo- 

 ccetus Dum. and Gastrobrarichiis Bl. all those parts which 

 ought to have constituted their skeleton as vertebrated 

 animals become so soft and membranaceous, that they 

 may be considered as having no bones. The organs of 

 respiration and of manducation, the absolute want of the 

 sense of sight, the general habits and external form of these 

 singular fishes, all prove to us that they are connected 

 with the Annelides, and that by them nature passes to 

 the structure of the Annulosa. On the other hand, on exa- 

 mining some of the Echinoderma of Cuvier, such as 

 those composing the genus Comatula, we may trace the 

 articulated texture of the ^H/<^<^5a into the division of Ka~ 

 diata, many of whose external forms are also exactly imi- 

 tated by the sessile Cirr/iipedes. Of the Radiata, again, 

 the stellate form and the gelatinous semi-transparent sub- 

 stance are observable among the Acrita. So that the chain 

 Avhose links we have endeavoured to unfold returns into 

 itself, and we find that all animals form a circle composed 

 of the following great divisions, viz. 

 ACRITA, 



mollusca, 

 Vertebrata, 

 Annulosa, 

 Radiata. 



This arrangement of animals is, it is true, quite distinct 

 from that generally adopted ; but it will be seen that it is 

 not only conformable to nature, but that it removes many 

 of the discrepancies which shock the naturalist in the com- 

 mon systems. For instance, there is an acephalous ani- 



