CLOUD-OUTLINE?. 
245 
does it happen that the surface of these clouds and their general 
contour do not change more rapidly ? To this it may he replied, 
that if an observer were so placed as to obtain a distinct and 
near view of the component particles of the cloud, he might 
(and probably would) observe a very rapid motion among 
them. For, at a distance of several miles, appearances are 
very deceptive as to material composition,* movement of par- 
ticles, &c. ; and it is also plain that identity of visible form at 
two different moments is no proof whatever of an identity of 
the component particles, or (perhaps I should say) of a con- 
tinuity of composition. Take, for example, the rainbow, which 
exhibits such exquisite symmetry and such apparent stability 
Fig. 2. 
FESTOON CLOUD. 
of form as altogether to preclude the idea of change or motion, 
while it is well known that the bow and its colouring are due 
to refraction of light through a constant succession of falling- 
drops, so that not only do two persons see different rainbows, 
but no one can see the same rainbow two moments together. 
In the symmetrical marking which the pyramid cumulus as- 
sumes at times, we have a remarkable resemblance to that which 
is occasionally taken by the whirling sand-pillar of the desert, 
when the latter takes a compound form. A kind of complex 
* A singular phenomenon may often be noticed in connection with the 
fall of rain, which has not been fully explained, viz. : the darting to and fro 
of small particles among the larger drops, like minute snowflakes flitting 
about in defiance of gravitation. 
