270 Mr. 0. Salvin on the 



skin shows markings of a dull flesh-colour, that part has been 

 coloured in the living bird with some shade of red. There are 

 cases, however, where actual colouring matter is to be noted, 

 in addition to the usual horn-black, as in Phaethomis adolphi, 

 where the basal half of the lower mandible is straw-colour. 



18. Amazilia riefferi. 

 Coban. November. 



This Amazilia is found also at Yzabal. It is far from common 

 at Coban. All my specimens appear to be males. 



19. Amazilia dumerillii*. 

 Duenas. August. 



During the months of July, August, and September, one 

 of the most favourite resorts of this Humming-bird was the 

 western boundary of the llano of Duenas, which, starting from 

 the village, and bounded to the eastward by the river Guacalate, 

 extends, sweeping by the base of the Volcan de Fuego, almost to 

 the Hacienda of Capertillo, its southern extremity. Dispersed 

 all over this plain is found, in groves, patches, and isolated trees, 

 a Tree-Convolvulus, bearing a white flower, and attaining an 

 average height of about 25 or 30 feet. During the above 

 months, this elegant species might be seen in almost every tree, 

 some feeding among the flowers, some settled quietly on a dead 

 branch, uttering their low, plaintive, hardly to be called musical, 

 yet certainly cheering song ; others less peacefully occupied in 

 a war of expulsion, driving out, by vehement cries and more 

 effectual blows, the tenant of a tree, which in its turn wreaks 

 vengeance on some weaker or unexpectant antagonist. 



Of this species I have skins, of which the sexes are in the pro- 

 portion of four males to one female. 



20. Thaumantias candidus. 



This species, which is very abundant about Coban, is found 

 also at Yzabal. Many species of Humming-birds in Guatemala 

 extend through a great range of temperature, the same species 



* This Amazilia I have previously called^, arsino'e; but upon com- 

 paring my specimens with Mr. Gould's numerous examples of the Mexican 

 A. arsino'e (which I have been enabled to do through his kindness), I find 

 it distinct, and correctly referable to A.dumeriUii. 



