22 
NOTES OF A BOTANIST 
granite region of the Rio Negro and Orinoco, and 
saw there tall trees growing on perfectly bare rock, 
or where the earthy covering was only a few inches 
thick, so that the roots were necessarily either 
wholly or in great part above the surface, I under- 
stood how sapopemas might have originated ; for 
those peculiarities or seeming anomalies of struc- 
ture, which we (to hide our ignorance) are too often 
contented to call freaks of nature," have no doubt 
arisen, in the first instance, from the adaptation of 
organisation to the accidents of existence, and have 
been continued through descendants of the original 
stock even when no longer exposed to the influ- 
ence of such accidents. 
A few sorts of trees, including some palms, are 
supported on exserted or superterraneous roots, 
which differ only in that particular from ordinary 
subterraneous roots, that is to say, they are round 
or cylindrical, and not flattened and dilated vertically 
like the sapopemas. In England, an old willow or 
other tree standing by a river, whose floods have 
washed away nearly all the earth from its roots, 
may give an idea of this form ; which, however, is 
constant in many Amazon trees whose roots have 
never been exposed to denudation by the action 
of water, whatever may have been the case with 
the prototypes of those trees. These examples led 
me to conjecture, at first, that the sapopema form 
itself might have taken its rise from denudation, 
in the remote ancestors of the existing types of 
trees ; or at least that sapopemas were at first a 
sort of scaffolding to raise the crown of the root 
above the reach of inundations ; and I am still 
willing to believe that to this cause their origin 
