Notes . 14 7 
and no large sieve-tubes ; but there is a relatively better developed pith 
than in the underground roots. 
Descending the aerial root towards the ground, all the intermediate 
stages between the two extremes described, occur. The change in 
the form of the cells of the velamen, and their gradual increase in 
number, is especially noticeable. 
Comparison of the Structure of the Aerial and Subterranean Roots . — 
The aerial roots exhibit a return to a simpler state. They are not 
roots which have merely stopped in development. The apex of an 
underground root shows a thick velamen forming at once close to the 
very tip of the root, whereas the apex of a root just commencing its 
aerial career shows already a diminution in its velamen ; and the apex 
of an aerial root just about to stop growth altogether is practically quite 
similar to that of the root of any ordinary terrestrial plant h It is still 
more remarkable that, when an aerial root dips down into the ground, 
it changes the character of its apex once more, and commences to form 
a thick velamen ; such a root has a thin basal part, with only a few 
layers of velamen-cells, which are not markedly elongated at right 
angles to the axis ; but nearer the apex the root is thicker and has 
a typical large velamen. Altogether it appears that, for some reason, 
the velamen is not fully developed unless the root be under the soil ; 
whether this peculiarity is occasioned by the action of light or the 
distribution of moisture must be left undecided till direct experiments 
be made. 
Conclusions . — Schimper has clearly explained the use of the nega- 
tively geotropic roots possessed by epiphytic orchids, and there is no 
reason to believe that they have any other function when they are 
developed in individual plants growing in the ground : that they act, 
for instance, as respiratory organs. The negatively geotropic roots 
are a sign of an epiphytic mode of life ; and their occurrence in 
specimens of G. speciosum growing in the ground shows clearly that 
this plant is a typical epiphyte which can, however, grow as a terrestrial 
plant — and not the converse. This evidence is strengthened by the 
fact that Grammatophyllum is essentially an epiphytic genus. Hence, 
in G. speciosum we have an epiphyte which can be found growing wild 
as a terrestrial plant 2 . In this latter position, the velamen is not only 
1 Excepting the fine reticulate thickening of the one-layered velamen. 
3 Mr. Ridley tells me that Cymbidium aloifolium is not to be found growing as 
a ground-orchid, although it can be cultivated as such. 
