448 ORNITHOLOGICAL EXPLORATION OF CARIBBEE ISLANDS. 



Roseau, aud passed the jieriod of convalescence on the Caribbean coast, 

 procuring, in my intervals of strength, the birds breeding in the clifls. 

 It was in the Atlantic Mountains that I procured the first specimens of 

 the Imperial Parrot ( Chrysotis angusta) ever sent to America. 



The 4th day of July, 1877, 1 arrived at Antigua, in latitude 17<^ north, 

 and examined that island, and also Barbuda, 30 miles farther north. These 

 islands are of different formation from the others of the chain, though 

 owing their origin, i^erhaps, to the same cause. The rocks of the west- 

 ern portion of the island, according to Davies, are of igneous character, 

 denoting violent action and protrusion from beneath, akin to the vol- 

 canic, but without actiTal enii^tion or explosion. Eastern rocks belong 

 to the " tertiary aqueous formation, being chiefly calcareous freestone and 

 limestone, containing organic remains." "In the intermediate space 

 are exhibited both igneous and aqueous action ; the former in the indu- 

 rated clays and siliceous cherts, forming the skirting western declivities; 

 the latter in the numerous petrifactions, varieties of wood or coral, prin- 

 cipally, if not entirely, siliceous." In Barbuda are many deer, wild guinea 

 fowl, and horses, the result of stock introduced from England many years 

 ago. 



In Antigua I found one species of owl, since declared to be new^ the 

 Speotyto amaura, but few birds of any kind. In September large flocks 

 of x>lover, and in October teal and duck, visit this island in large num- 

 bers. Its full avv-fauna will be found in the catalogue recently prei)ared 

 by Mr. Lawrence from my notes. 



In September, I sailed for St. Vincent, stopping two weeks, on my way 

 south, at Dominica, to procure further specimens of the species declared 

 new and rare by letter from Mr. Lawrence. I was there fortunate in 

 securing the owl since identified as a local variety of Strix JIammea, with 

 three eggs. Two specimens were shot, male and female. Beaching St 

 Vincent in October, I early set to work, but was hindered by the rains. 

 I not only searched all around the coast, but penetrated to the mount- 

 ains. Here I was i)eculiarly fortunate, securing seven neic species. The 

 names, as appearing in the catalogue of the birds of St. Vincent, since 

 published, are as follows : Turdus nigrirostris, Myiadestes sihilanSy Thry- 

 othortis musicus, Cei'thiola atrata, Certhiola saccJiarma, Leucopeza hisliopi., 

 Calliste versicolor. My adventures in securing the Myiadestes sihilans 

 were most interesting. I found that the only way to reach it was to camp 

 upon the crest of the crater of the volcano, which I did, making my home 

 in a cave for five days and nights, 3,000 feet above the sea. It rained 

 continually during my stay, in my exposure to which I contracted a severe 

 cold which culminated in a fever several weeks later. I had the ex- 

 quisite pleasure, however, of securing there the wonderful bird named 

 above, known to the natives as the "invisible souffriere bird." Barely 

 a hint of the existence of this bird is contained in scientific annals, and 

 that only in Goss's Birds of Jamaica, in a letter to the author of that 

 work. 



