THE EQUATORIAL FORESTS 23 
may be partly traced. But that it has not been 
the sole, or even the principal, cause is plain from 
the fact that of the trees flourishing at the present 
epoch in inundated grounds, where their trunks are 
under water for months to the height of 10, 20, or 
more feet, very many have no sapopemas at all. 
As suggested above, a rocky matrix, bare or thinly 
covered with earth, may have been the main origin 
of sapopemas, for it is in such sites that the most 
numerous and perfect examples of them exist at 
this day ; and if we suppose it combined with in- 
undations and denudations, I think we may thereby 
explain most of the modifications of exserted and 
dilated roots. 
Sapopemas exist on trees of many genera and 
families, but they seem to attain their greatest size 
in Bombaceae, Leguminosae, Lecythideae, Moraceae, 
and Artocarpeae. There is, however, one family, 
Lauraceae, consisting almost entirely of forest trees, 
yielding to no others in their noble aspect and the 
usefulness of their products, in which I have never 
seen more than a rudimentary development of 
sapopemas ; their roots, in fact, penetrate deeper 
than most others, and wherever laurels predominate 
it is a sure indication of a good depth of soil. 
There are instances in a single family of some 
species of trees having large sapopemas and others 
none at all ; as in Lecythideae, where the gigantic 
Bertholletia buries its roots almost entirely, and 
the species of Lecythis, some of which are trees of 
vast size, have the roots raised high out of the 
ground. 
