448 BOTANICAL GAZETTE [JUNE 
generally soon seized upon, and the fact that these marginal 
ditches remain devoid of vegetation, not being even bordered 
with the usual water margin plants, suggested to the writer that 
there must be some cause which prevents the growth of vegeta- 
tion in that zone, and that this might be identical with the one 
which originally gave rise to the ditch. A rise of water, in 
conjunction with such a cause, might make a compound atoll. 
Atkinson, in his Lessons in botany, makes additional sugges- 
tions in regard to the origin of the ditch, one of which is that it 
is caused by the shade of trees and shrubs growing on the solid 
land, and thus giving low vegetation around the border of a 
pool a poor opportunity. However, the northern edge of a 
pond is quite exposed to the sun during the hours of midday, 
yet the ditch exists there as markedly as anywhere else. 
It was observed that formations of this character were found 
only in wooded districts, or on those recently cleared. Portions 
__ of the islands of the neighborhood are treeless, and in these 
_ localities the ponds were filling’ up from the margin in the well 
known manner. 
- Another fact was eventually noticed, namely, that the ditch 
varied in width, and that this variation bore a constant relation 
to the contour of the surrounding surface. The ditch is always 
Poses esate from the arrangement of the surrounding slopes, 
: os ial is washed in. That fact seemed to harmonize 
‘ill with ideas of erosion and deposit. 
: sidered, however, an explanation of the different facts began to 
_ The forest floor around is of humus, and covered with 
felt of roots, mosses, and mycelia. It does not take a 
matter | is washed into the pond. Such is brought in, 
om 
: hen the character of the material brought in was con 
: amination to convince one that ordinary erosive action 
of rain has literally ceased. Only fallen leaves and other 
ntities, and young plants which might start : 
‘ 
