240 TlMEHRI. 
tated, because of its dangerous position and slender form 
to send a man up it for specimens ; and when he 
ascended and it swayed with his weight out over 
the awful chasm beneath, I trembled for what then 
seemed to me the great folly of my conduct. However, 
he got down safely, and indeed declined to return till he 
had secured an armful of the branches. The flowers 
have long tubes like those of Posoqiieria longiflora, but 
stiff enough to be erect. One of the highest trees of the 
open savannah, which I had cut down to get at the 
bloom, proved to be a probably new species of Plumieria. 
Three or four members of this genus * are among the 
principal ornaments of western tropical towns and 
gardens, and this new species is hardly inferior to them 
in plentiful and showy bloom ; but it is a much larger 
growing tree. Its milk-like juice, which flowed so copi- 
ously as to spurt into my face as I stood near while it 
was being felled, I found intensely acrid. I saw only this 
individual tree. 
Among the smaller shrubs which fill the copses, hang 
on the skirts, or straggle loosely over the more open parts, 
of the savannah, are several species of Gomphia. I 
gathered about a dozen, the greater part of which, 
strange to relate, proved new to the best herbarium in 
the world — that at Kew. They are mainly woody plants, 
a few feet high, with flowers of a whitish or pale yellow 
colour. That so many of them had never before been 
collected in the colony, is further proof of the exceptional 
character of the Kaieteur savannah flora. I gathered 
also two forms of Turnera rupestris, Aublet, which ap- 
* Horn mor.lv called frcvngvpa/nM. 
