John Hogg, Esq., on the Classification of Birds, 61 



order to assign it a station more consistent with nature, 

 placed it in a separate family in my Cuneirostral tribe. 



Subtribe 2, Clamatores, Criers or Screamers, I have limit- 

 ed to only a few groups ; one of which, the Coraciadidae, or 

 the European Boiler family, I prefer placing in the Coniros- 

 tral tribe, and next to the jays, which in many respects it 

 resembles, rather than among the Fissirostres, as some of 

 our modern naturalists do. Although the wide gape, with 

 which the common roller is furnished, may give it a claim to 

 that place ; still I am inclined to divide the present family 

 Coraciadidw, and station the Australian and African kinds, es- 

 pecially those of the latter, which are long -tailed, and strong- 

 ly approach the bee-eaters in form and appearance, next after 

 the family Meropidse, among the Fissirostres. And this di- 

 vision would then constitute a new family, and stand in my 

 subtribe Allodactyl% and just before the Hirundinidse. I 

 have observed the common roller in Sicily, and think it 

 clearly more allied to the Corvidae than to any other group. 

 Like the jay, it is restless, makes a loud chattering cry, 

 and seeks its food upon the ground, which consists of in- 

 sects, caterpillars, worms, &c. But it even resembles the 

 woodpeckers in breeding in decayed trees, and having eggs 

 of a beautiful shining white colour. In fact, the eggs of 

 the roller in their shape more exactly correspond with those 

 of Picus minor. 



Family 2. Corvidw. It will be remembered that Cuvier 

 classes the genus Fregilus, with the hoopoe, amongst his Te- 

 nuirostres, which is perhaps its proper place, if we regard 

 the beak alone. To place it in the Conirostral tribe seems 

 incorrect ; but its general appearance and habits must decide 

 its station to be with the Corvidae. Fregilus is, consequent- 

 ly, an aberrant genus. 



The third subtribe, or Cantatores, Singers, is very exten- 

 sive, comprising, in strictness, all the singing birds, and es- 

 pecially the true Warblers. 



Family 4. Loxiadw. Considering that the family of Frin- 

 gillidcB, as usually retained, is much too comprehensive, and 

 ought to bo divided into one or two more, I have adopted, for 

 the larger and thicker billed genera Loxia, Pyrrhula, Cocco- 



