238 TlMEHRI. 
study alone would occupy one's time on the ground for 
several weeks together. Resembling each other closely 
in leaf and general habit, they yet exhibit both remarkable 
unity and diversity in other characters. In flower and 
fruit the different species vary much one from another. 
On the form and character of the fruit, good distinguishing 
characters might be based ; but as they are rarely 
found in flower and fruit at the same time, the field is the 
only place in which they can be studied satisfactorily. Of 
about a dozen species which I gathered in bloom, only 
two were in fruit. They prevail largely on all the dry 
rocky or gravelly parts of the savannah, and no one would 
suspect from the hold they have of the ground, that in 
situations more favourable to general plant development, 
they prefer to live on the trunks and branches of trees. 
A species of Moronobea, another genus of this same- 
family, has perhaps the most striking flowers of the many 
interesting plants I came across on my journey. It is a 
new species, and is common on the higher ground on the 
edge of the forest, growing erect and tall, from ten to 
thirty feet high. A tree seen in bloom in external ap- 
pearance is not unlike Magnolia grandiflora. The 
flower, a pure white, but with an offensive odour, is the 
largest known to me of any western tropical plant except 
only the Victoria regia. The wood when cut exudes a 
yellow juice of an adhesive nature. 
The savannah is crossed diagonally by a rather faint 
trail which the Indians use-, on the few occasions they 
pass up or down the river, to avoid the Kaieteur, and 
about halfway down this path, among the low bushes 
which are there dispersed over the ground, occurs a 
