t 



28 Canon H. B. Tvktrdm— Oniilholo(/ical 



night. I saw for several days just before the close season 

 numbers of Partridges for sale in the market of Las Palmas, 

 and only of this species. 



When I first obtained the French Partridge at San Bar- 

 tolomeo I was struck by the size of the bill, much larger, I 

 thought, than in any specimen I had seen elsewhere. That 

 this was no accidental variation I ascertained by comparing 

 my skin with twenty-nine other birds 1 found for sale in Las 

 Palmas on the day of my return. All agreed in dimensions 

 both of bill and tarsus, as well as in coloration. On com- 

 parison with the series in the British Museum, and with 

 Lord Lilford^s and my own, I find the following marked 

 differences. The Canarian bird has a band of reddish brown 

 on the nape and hind neck, brighter than in French and 

 English, but not brighter than in Spanish examples. But 

 whereas in European birds the whole of the rest of the upper 

 parts are reddish brown, in the Canarian the back and upper 

 tail are slaty grey. I have seen an example from Andalusia 

 which is intermediate between the Northern and the Cana- 

 rian characters. The black collar round the fore neck is very 

 much wider than in European birds. So marked is the dis- 

 tinction that Mr. Godraan felt disposed to describe his 

 specimen from the Azores as a new species, had it not been 

 that it was a wretched and mutilated specimen in moult. 

 But on comparing my Canarian with his Azorean there can 

 be no doubt of their specific identity. But the marked 

 structural distinction is in the beak and tarsi. In French 

 specimens the beak measures 1-33 inch from the gape, and 

 the Canarian 1-82 inch. Depth from the ridge of the 

 culmen : French specimens '27 inch ; Canarian, '33 inch. 



Length of tarsus in French specimens 1-33 inch, Canarian, 

 1*82, while both bill and tarsus are comparatively far more 

 massive in the insular bird than even their measurements 

 would show. I propose, therefore, to distinguish the latter 

 as 



Caccabts rufa, var. australis. 



C. rostro^ quarta parte robustiore et longiore quam in 

 C. rufa: tarsis robustioribus et dimidio pollicis longi- 



