330 OBSERVATIONS ON A SPECIMEN OF " SXILA BASSANA. 



or make different species for different regions, I somewhat incline to 

 number this among the imaginary ones. 



The habits of the Solan goose, especially at its breeding time, have 

 been so fully described in popular works, that I shall not dwell on the 

 subject ; but its systematic position and affinities may be worthy of 

 some consideration, and will give me the opportunity of offering some 

 details illustrative of views of classification, of which a general out- 

 line has been already submitted to you. 



The genus Sula is referred to the family Pelecanidae, among the 

 Natafores or swimming-birds ; and this family is distinguished by the 

 union of the hind toe with the others in a single membrane, great 

 power of wing, generally short legs, the nostrils being slits with a 

 scarcely perceptible aperture, the skin of the throat generally more or 

 less extensible, the tongue small, and the gizzard united with the sto- 

 mach. The admitted generic forms in the family are Atagen^ the 

 frigate-birds — (amidst much confusion about names, I have adopted 

 those of Dr. Geo. Gray, as given in the Genera of Birds) ; Phaeton, 

 the tropic bird ; Sida, the gannett ; Graculus, the cormorant ; Pele- 

 canus, the pelican ; and Plofus, the snake-bird. These genera are not 

 rich in species, and the family is a small one, and remarkably devia- 

 tive in its habits from the mass of Natatorial birds ; yet such striking 

 peculiarities are observable among the genera, that Dr. G. Gray has 

 placed them in three sub-families, and we may see reason for extend- 

 ing this sub-division. But before we proceed with the analysis of the 

 Pelecanidae, I may perhaps be indulged in a few observations on 

 classification generally, and especially on the best treatment of the 

 great class of Birds, designed in illustration and confirmation of a 

 method already submitted to the Institute. 



It is well known, that according to the views of McLeay and 

 Swainson, the plan of creation in every different group of organized 

 beings, is sub-division into three leading, — or, since one of them is 

 itself again divided, — into five secondary groups, which are charac- 

 terized as typical, sub-typical, and aberrant — the latter itself exhibit- 

 ing three variations. In applying this scheme to the various parts of 

 organised nature, so much knowledge and ingenuity v^ere displayed, 

 and there were so many striking instances in which a satisfactory 

 natural grouping seemed to result, that many were led to believe that 

 the secret of natural classification was detected, and that in order to 

 obtain a perfect system we had only diligently to work out in the 



