ITINEEARY. xlix 



much smaller and poorer in growth than in the more humid atmosphere 

 higher up, though probably the larger growths had been alreadj' carried 

 off. On the stunted trees along the streams and dried-out channels, 

 small plants were numerous. 



One of the most interesting of all the orchids was an Epidendrum, 

 apparently E. sckomburgkii, which is very widely distributed on rocks 

 and stumps, often much exposed to the effect of wind and sun. Its beau- 

 tiful and abundant groups of red flowers were remarkable in the variation 

 of the shade of colour, passing at times into pale crimson and magenta, 

 pink or pale orange, or even to nearly a pale yellow or white, more or 

 less of the variation being often seen on the same inflorescence. At 

 times it was so striking that it almost seemed, at a first glance, as if 

 two or three different species were growing together in one clump,, 

 though the extremes were mostly in individual flowers of different 

 bunches. The range in size of plant and flower in this hardy species 

 was even more noticeable than in the Cattleya. They often grew on 

 bare rock in such exposed situations it was astonishing not that they 

 were stunted, but that they could survive. The humidity at night no 

 doubt gives the explanation in this and in other cases, hardly less 

 remarkable, such as Gyrtopodium criatatum, a specimen of which was 

 found growing amid small bare stones on a high hill, its golden, 

 brown-spotted flowers being very fragrant. The commoner species, 

 C. andersoni, which goes by the Jocal name of " sugar-cane orchid " 

 from the appearance of its long and thick, apparently jointed pseudo- 

 bulbs, which may reach about six feet in length, is often found on 

 bare rock fully exposed to the sun, as at the rapids of Akra-mukra, 

 Essequibo river, and in the Kanuku Mountains ; but in this case theie 

 are very large reserves of moisture, though the peculiarity remains in 

 the starting of the young plants originally, unless some protection has 

 been lost in some way from the site. 



It was noteworthy that, while many forms such as the yellow 

 Jussiseas, the blue-belled gentian {Lisianthus uliginosus),iinA the crimson 

 campanulats Centropogon surinamensis, which were plentiful in the 

 valleys of tJie upper plateau, were also met with in the lower savan- 

 nahs, some of the most striking of the lower forms did not reach the 

 Roraima district. Among these were such forms as the fragrant 

 purple Rhynchanthera, the deep-blue Stachytarpheta, the sky-blue 

 Jacquemontia, and the new purple Ipomcea irengana, which are so 

 common at times either along the streams or in the open savannah 

 the last apparently flourishing best on the ant-hills. 



A rather small spiny Mauritia palm (if. aculeata) is somewhat 

 common in the vicinity ; and not far away the great CEta palms are 

 plentiful in the valleys, with other such useful plants as the giant reed 

 or wild cane {Gynerium sacchar aides), and the Agave form {Furcrcea) 

 of the plants called " Crowia." The CEta is not only of use to the 

 natives as furnishing sandals (which are fairly lasting if allowed to 

 dry before being used), thatch and fibre, from the leaves— the fibre 

 being a strong one for string, rope, and hammocks,— but the fruit is 

 gathered for food, especially in times of scarcity, when the provision- 

 fields may have failed. The pulp is pressed into large balls, and is 

 mostly used in the form of a rather thick drink. The wild cane yields 



