HOW TO SKETCH THE MOON. 
23 ^ 
surface, in connection with the varying temperature of differ- 
ent parts of the lunar day. Apparent variation may not un- 
frequently he perceived among minor details ; but nothing 
satisfactory would result from the precipitate inferences which 
have been sometimes drawn from it, and a little further ex- 
amination of this subject may be permitted, as it has not 
always been sufficiently considered in its full extent. 
When there is evidence of want of correspondence in draw- 
ings made by ourselves or others at different dates, we have 
first to bear in mind the possibility of omission through de- 
fective attention — a not infrequent source of error, and one 
which is subsequently irremediable. Some observers, no doubt, 
are more liable to it than others, but none could claim abso- 
lute immunity ; to photography alone is reserved the triumph 
over this form of error. Then as regards the minutest visible 
details, the state of our own atmosphere may exercise con- 
siderable influence. But besides these potent sources of dis- 
crepancy, others remain in reserve of a very different character. 
It need hardly be said that the true relief of any surface not 
viewed in profile is decided by the known, and unalterable laws 
of light and shade. These laws indeed admit of no variation, 
but their application sometimes involves cases liable to mis- 
construction, from peculiarity of form in the projecting or 
receiving surfaces, and if the position of those surfaces is itself 
liable to change, either with regard to incident or reflected 
light, or both, it is e\ident how much the chances of some 
illusory effect maybe multiplied. A few instances may be of 
use in the way of illustration. 
No misapprehension could arise in the case of shadows pro- 
jected from regular forms upon an uniform sphere. But it is 
needless to observe how far these conditions are from being 
fulfilled upon the moon ; and not only are the shadows very 
irregular in their shape, but they are liable to many incidental 
causes of disturbance, partly of a real, partly of an optical 
character. Independent of their normal variation in actual 
length according to the height of the sun above the lunar 
horizon, they may also vary in actual length from the uneven- 
ness of the recipient surface. For instance, a shadow will be 
gradually shortened, or suddenly blunted, if its extremity falls 
on a rising slope, or against a rapid elevation ; or it will be 
correspondingly extended under opposite circumstances ; and 
thus the edge of the shadow of a long precipice of uniform 
height, instead of being bounded by a straight line, as it would 
be on a plain, may become curved in various directions, or even 
jagged, if falling on very irregular ground. An apparent 
lengthening or shortening may also take place in, obliquely 
viewed regions, when illuminated objects are brought by change 
