THE EQUATORIAL FORESTS 39 
There is a type of trees in the Amazon valley — 
sparingly represented near Para — with the stems 
either quite simple or emitting only a few long 
wand-like branches, which are naked except at the 
continually lengthening apex, where they bear a 
few crowded leaves, often of such enormous length 
as to give them at a distance the aspect of palms. 
In their season, flowers spring from the naked 
trunk or branches, generally in clusters, and often 
noticeable from their size and beauty, as in Gustavia 
fastuosa^ which has the large rose-like flowers some- 
times 7 inches across. Some of the handsomer 
Melastomes (Bellucia, Henriettea) are of this type. 
Nearly every shade of green is observable in 
the hues of the foliage, the deepest being usually 
in the large glossy leaves of Guttifers and the 
opaque leaflets of Ingas. There are no autumnal 
tints on the Amazon, for although some leaves turn 
reddish or brownish with age, the change is very 
far from being simultaneous in all the leaves of the 
same tree, and is often entirely hidden from view 
by the unceasing growth of new leaves. But the 
absence of that periodical ornament of our northern 
woods is almost compensated for by the delicate 
hues of rose and pale yellow-green assumed by the 
young leaves at the growing point of the branches, 
contrasting admirably with the deep green of the 
rest of the tree. 
Many leaves are grey or hoary beneath, as in 
the Cecropias (or Imba-iibas, as they are called by 
the Indians), and some leaves are clad on the 
underside with a fine down of a lustrous metallic 
hue — silvery, coppery, or bronze — especially notable 
in various Laurels and Chrysobalans. When my 
