HYLIOTA. 231 



Let us, in conclusion, offer a few remarks on the 

 more general results which the preceding investiga- 

 tion tends to establish, while the facts are as yet 

 fresh on the memory. 1. We perceive that all 

 natural suh-genera, as such, imitate, as it were, or 

 represent one or other of the primary forms of 

 nature, and that their variation is regulated by the 

 same order of succession. The rasorial forms, for 

 instance, is followed by the tenuirostral ; and this, 

 again, by the Jissirostral. 2. That this law of vari- 

 ation is the same, also, which regulates the varia- 

 tion of the species : one example of this we have 

 already given in a sub-genus of quadrupeds, and in 

 the sub-genus of Todus we have a second example 

 in the class of birds. Lastly, we perceive that the. 

 rank of any one of these types can only be known by 

 analysis ; and then only when our materials are very 

 ample. Hence arises the uncertainty respecting 

 the rank of every type ; because, without the most 

 refined analysis, it is impossible to determine whe- 

 ther a bird is a specific or a sub-generic type of 

 form. Let us now proceed to the fifth and last 

 group of this family. The 



EURYLAMIN7E, OB BROADBILLS, 



as before observed, constitute one of the primary 

 divisions of this family. They are a most remark- 

 able group of birds, whether we regard their ap- 

 pearance, or the extreme interest which attaches to 



