Total Eclipses of the Sun. 419. 
but when the vapour forms in the upper current, its motion will 
probably be contrary to that of the lower one; this latter point, 
however, is very uncertain. 
With regard to the third motion, nothing can be laid dawis 
theoretically respecting it. In a general way the consequencé 
is that the apparent motion, considered as forming a part of the 
three mentioned above, will, both in amount and also in direc- 
tion, be accidental. Nor are we able to lay down anything 
more definite relative to the changes of form originating in the 
motion. Ifin a cloud of uniform density (fig. 3) the motion 
takes place in the direction cd, the elevation on the moon’s limb 
will remain tolerably steady, with the exception of its extension, 
which alters with the breadth of the cloud. A projecting point 
(fig. 4) apparently disappears behind the moon’s limb if the 
motion is from a towards 4; a projection becomes converted 
into a mountain-crest when the mass of vapour approaches 
towards the moon’s ceutre. 
Some kind of motion, however, is always to be looked for; for 
it would be a most improbable case that all the three motions 
mentioned above should just counteract each other. 
If we compare the theoretical views hitherto promulgated with 
observation, it will be seen that, generally speaking, observers 
have not directed their special attention to the circumstances 
that were essential to arriving at a decision; nevertheless many 
particulars in confirmation of my explanation may be derived 
from the great number of reports and sketches that we have 
before us. The sketch by Fearnley of the solar eclipse of the 
28th of July, 1851, as observed by him at Rixhoft, seems to me 
to be especially characteristic. See fig. 6 for a copy thereof. 
The form proper to clouds is here so distinctly pronounced, 
and extends so far away from the moon’s edge, that it seems to 
me impossible to look for the origin of the phenomenon any- 
where but in our atmosphere. 
I select the following circumstances from the very instructive 
report of Schmid, who observed the same eclipse at Rasten- 
berg. 
in the first seconds of the totality, he at once perceived in the 
corona “ several brighter places of imperfectly defined form, some- 
thing like separate, white, very nebulously-blurred little clouds.” 
When the sun broke forth in the form of a fine crescent of light, 
he examined carefully the remaining part of the moon’s edge, 
which disappeared altogether after 1 minute and 55 seconds, 
after being 40 seconds previously partially interrupted, so that 
there yet remained only separate portions of the are which repre- 
sented the curvature of the moon’s edge. 
Schmid did not remark any protuberances at all during the 
