I ALLAN HANCOCK ATLANTIC EXPEDITION REPORT 



noon during the passing of a heavy thunder-storm on the coast of the 

 mainland. This mist was barely perceptible and of course not sufficient 

 to fui'nish water in any quantity. No streams or springs exist on the island. 

 The unweathered condition of the rocks suggests the great lack of rain, 

 while the presence of the sand in the hollows rather than on the hilltops 

 may be accounted for by the wind movements. The rounded hills are 

 wind worn, and there are no gullies due to rushing water." 



Since Cubagua is less than ten miles of? from Coche and is even 

 smaller, we can surmise that in climate and flora it is similar to Coche. 

 Johnston collected 21 families in ZZ genera and in 37 species on Coche. 

 Two of the species were described as endemic. Doubtless Cubagua shares 

 endemic plants with Coche, but that is not at present determinable with 

 our limited knowledge and the material at hand. Figure 1 shows a view 

 of the interior of the island. Judging from this, the dispersed but spotted 

 groupings of the arborescent plants, the mixture of mesophyllous shrubs, 

 the thick-stemmed dwarf tree, and the succulent element, the vegetation, 

 perhaps, should be classed as thorn forest, but the physiognomy also sug- 

 gests desert shrub. 



Of the other islands visited, Trinidad and Tobago are the best known. 

 Collections here reported from Trinidad were obtained mostly at the 

 Royal Botanic Gardens at Port of Spain. Many of them represent useful 

 or ornamental plants from other countries. 



Curasao has been better collected than other islands off the Venezuelan 

 coast. The floras of Tortuga and Aruba remained little known until 

 Boldingh published his work on them in 1913 and 1914. Urban (1902) 

 reports A. Ernst as having collected on Tortuga in 1874, Jacquin and 

 Suringer on Aruba, the former between 1755 and 1757, the latter in 

 1885. The environmental relations of the native plants of these islands 

 are still scarcely known. 



One hundred and fifty numbers were secured in Panama. Of the 

 localities visited, Bahia Honda and Taboga Island are on the Pacific 

 coast. These were actually visited during the 1939 Pacific voyage, the 

 plant collections of which are to be enumerated in a report now in prepara- 

 tion. In order to avoid duplication of citations and geographical unity, the 

 Pacific collections from Panama are reported here with the rest of the 

 Panama material. Bahia Honda has rarely been visited by botanists and 

 several discoveries there provide new distribution records. Figures 2 and 3 

 show some of the vegetation in Bahia Honda. 



