THE COLUMBID^. 61 



opportunities of gaining information which are not easily recovered. It is only as 

 the mind becomes familiar with one subject, that it is prepared to see the different 

 bearings of those that are related to it, and to appropriate them to its use. When 

 I commenced the pursuit, I had read nothing whatever on the subject, and had 

 no one to direct my labour. I was not even aware, that some birds had caeca and 

 others none, that the sclerotic membrane of their eye was furnished with, and 

 supported by, a ring of bony plates, or any other of the facts of comparative 

 anatomy. But to return. 



I now come to the points of agreement between the Rasores and the Columbidce. 

 Here I find myself quite at fault, and shall be glad of the friendly assistance of 

 my readers ; for I find a total absence of that kind of resemblance, either in habit, 

 function, form, or internal structure, which I think ought invariably to accompany 

 the different families of one order. Indeed the only points of similarity I can 

 discover betweeen them, are, that the one is principally and the other entirely a 

 vegetable feeder ; and that they each have a membranous crop as well as a carti- 

 laginous one, and a gizzard ; even these are shared by several other families 

 of birds, and the membranous crop of the Rasores and the Columbidce differs in 

 shape ; in the former it is globular, in the latter composed of two lobes. 



Pavonidce and Tetraonidce are said by Selby to constitute the typical forms 

 of the Rasores ; Cracidce, Struthionidw, and Columbidce, to be aberrant families of 

 the same group. To shew in how unequal a degree they are entitled to the term 

 aberrant, I would observe, that out of 17 distinct points of agreement existing 

 between the Pavonidce and the Tetraonidce, the Crucidee agree with them 

 in fourteen, the Struthionidce in seven, and the Columbidce in only one. 



To make the order Rasores consist of these five families, appears to me very 

 incongruous, and to arise more from a desire of adapting them to the exigences of 

 a preconceived theory, than from any natural affinities observable between them. 

 A circle* composed of families differing so greatly in the amount of their aberra- 

 tions, presents, according to my ideas, chasms so extensive and frightful between 

 the different families of which it is composed, as not at all to accord with the 

 beautiful order of Nature; whereas, if placed in separate orders, we find the ex- 

 treme species running into each other, and forming one harmonious whole, which 

 cannot be contemplated without feelings of admiration and delight. 



* I have written the preceding observations without being aware of the views of the Editor of 

 The Naturalist on Systematic Zoology, and of course without wishing to hurt the feelings of any 



one. I have, in fact, given my own candid opinion on the subject T. A. [This \g, precisely 



what we should wish all our Correspondents to do. Although we would by no means have our 

 readers suppose that Mr. Allis has, in the above paper, demolished the quinary or circular 

 theory, yet we consider his remarks well worth the attention of the quinary systematist.— Ed.] 



