INSESSORES. 103 



death : the bill is reserved, as a knife, to separate the parts. But, among- the 

 Dentirostres, this process is partially reversed : the bill, and not the feet, is the 

 instrument of capture ; and that the struggles of its prey may not injure the 

 face of the bird, this organ is either much lengthened, or the base is so defended 

 from the possibility of mischief, by rigid feathers or stiff bristles directed forwards, 

 that the face becomes secure from all injury. The cere, so general among the 

 Raptores, is entirely wanting. The feet are less perfectly constructed than those 

 of the Conirostres, as the middle and exterior toes are frequently connected ; but 

 in such birds the tarsi are sufficiently elevated to remove them from the Fissirostral 

 order ; although, from want of due attention to this circumstance, the genus 

 Todus of Linnaeus has been lately transferred from the Dentirostres to the latter 

 tribe. 



Agreeing, therefore, in these general characteristics, we find the groups of 

 this tribe are naturally distributed in the three following circles : — 



DENTIROSTRES. 



1. ( Bill abruptly bent, 'strong, and deeply emar- 1 

 Typical group. \ ginate; claws much curved and very acute : J- LaniaDjE. 



I carnivorous or insectivorous. J 



2. r Bill moderate, compressed, gradually bent ; 1 

 Subtypical group.} the notch less developed : insectivorous or > Merulidje. 



[ frugivorous. ) 



3 - t Bill weaker ; the base generally broad ; the -j Sylviadye. 



Aberrant group. J feet either very slender, or short and L Ampelidje. 

 feeble. J Todid/E. 



The difficulty of framing characters for a group so diversified as the last, is suffi- 

 ciently obvious to every naturalist acquainted with their economy ; and it had been 

 better, perhaps, to have assigned to each of the three, by which it is composed, 

 a distinct character. But as this will be done in the sequel, we have considered 

 it more expedient to preserve, even in this slight sketch of the tribe, an indi- 

 cation of those primary groups into which we shall hereafter show it is first 

 resolved. 



In entering more into the details of the above families, it will be my object to 

 demonstrate, that the whole constitute a circular group. That the Laniadce are so 

 intimately connected with the Mendidce, that the two families have been blended 

 together by no less authorities than MM. Cuvier and Temminck ; that the aberrant 



