292 A COMMENTARY ON NOS. VII. AND VIII. OF THE NATURALIST.” 
the Pratincoles, the Sheatlibill, &c. If such be admitted, confessedly?* as a mat¬ 
ter of convenience, along with other genera, such an arrangement might, provi¬ 
sionally, be adopted as at least possible; but when - the position assigned to 
them is vaunted as equally satisfactory with those of genera concerning which 
there is no difference of opinion, it behoves those who pursue the inductive me¬ 
thod of investigation to analyze their claims, as Mr. Allis has so ably done 
those of the ColumhidcE^ to a place among the Hasores. For my part I greatly 
admire that gentleman’s method of procedure, vastly preferring the positive to the 
imaginative^ in matters of this kind; for it rarely happens that a deviation in 
any fundamental character, such as those on which Mr. Allis has founded his 
conclusions, is unaccompanied by equivalent diversities to bear out the distinc¬ 
tion. 
To give an example; and I select it on account of its having been so over¬ 
looked. The crania of the Albatross and Petrels, on the one hand, and of the 
Gulls and Terns on the other, present essential distinctions, such as those to 
which Mr. Allis has alluded in the case of the Columhidce^ and true Rasores ; 
and to recur to another of the leading characters mentioned by that gentleman, 
we find that the former lay, in every instance, a single white egg, whereas the 
latter produce three, which are blotched with black on a coloured ground. Again, 
the mode of flight, manner of progression on the ground, texture of plumage, and 
a variety of other particulars, afford corresponding indications, which are surely 
sufficient, taken collectively, to warrant their separation into two equivalent 
divisions, of the rank of families. It would, indeed, be vain to endeavour to 
point out analogous diversities between the Pavonidce and Tetraonidce, which 
have been admitted as such. They might subsequently be brought together, as 
a superior division, to be termed Longipennes^ after Cuvier, which would be 
subordinate to the higher group Nataiores. The Yagers (Lestris) would apper¬ 
tain to the same family as the Gulls, at least they are considerably more allied tu' 
these than to the Albatrosses and Petrels; though, at the same time, as they-are 
considerably less intimately connected with them than the Terns are, it becomes 
necessary, in order to represent duly the subordinacy of characters, to institute a 
minor supergeneric division among the haridae^ or what have been termed sub¬ 
families, thus separating the Gulls and Terns from the Yagers. The Albatrosses 
and Petrels being considered as a separate group, equivalent to the entire Laridce. 
Of course, the next step is into genera and sub-genera, which latter I should 
like to see specified in the following manner:— Larus-xema minuta^ Sterna- 
anous stolida^ Sterna-rhgncops nigra^ S^c.; these pertaining to the order and 
suborder Natatores Longipennes^ and family and subfamily Laridce Larince. 
It is only thus, I conceive, by establishing a long series of successively subor¬ 
dinate divisions, that any approach to a satisfactory arrangement is at all prac- 
