OF THE ANNULOSA. 397 



fined to the Insects breathing by tracheae, and to the ana- 

 logies which they in particular may display. 



It has been already hinted that a very beautiful analogy, 

 although disguised, reigns between the corresponding co- 

 lumns of Ametabola and Mandibulata. To prove that 

 there was ground for the assertion will now be my aim. 

 The subjoined columns represent part of the preceding 

 table of affinities. 



Branchiopoda . Vermes . . . Larvae of Hymenoptera 

 Isopoda .... Chilognatha . Larvae of Ti ichoptera 

 Laemodipoda . . Chilopoda . . Larvae of Neuroptera 

 Amphipoda . . . Thysanura . Larvae of Orthoptera 

 Decapoda . . . Anoplura . . Larvae of Coleoptera. 



There is so strong an affinity in general structure be- 

 tween Ricinus and certain Coleoptera, that it has disposed 

 me to believe that the circles of Ametabola and Mandi- 

 bulata must here meet ; a further proof of which may be, 

 that the larvae of Coleoptera, as being the nearest to the 

 Ametabola, do not imitate any one group in particular, 

 but the whole five, in precisely the same order in Avhich 

 they occur in their own circle. But if there should be 

 any doubt on the subject, I am sure that the perusal of 

 the 26th and 27th pages of Savigny's first Memoire must 

 remove it entirely. When the reader has well weighed 

 the comparison therein made and the nature of relations 

 of analogy, he will perceive that these two pages support 

 not only the affinity of the Coleoptera to Ricinus, but 

 many also of my other observations. The Hexapod larvae 

 of Orthoptera, with the setiform appendages to their tail, 

 meet with their prototypes among the Thysanura; the 

 flat carnivorous larvae of Neuropterous Insects, furnished 

 as they are with victorious mandibles, bear an obvious 



