168 
FIGURES UPON ICE, 
by any sub- 
stance lying 
there. Hence 
the theory of 
the figure. 
Objections to 
the Editor's 
theory of pu- 
trefaction in 
the body, &c. 
where they would be intercepted by it, and obliged to pass 
obliquely under it on every side, till they could rise with perfect 
perpendicularity, and thus in the act of freezing, give an accu- 
mulated whiteness to a perfectly distinct outline of the figure 
in the ice, as appears to have been the case in this instance. 
The ice forming within this outline would not be agitated, and 
therefore would be clear, hard, and slippery, according to the 
reasons you have given, and the test of universal experience, 
whereas the*whole remaining ice would be full of white bub- 
bles, and the confused crystallization would include the impu- 
rities of the water. The chief objection to this theory is, that 
the figure' of the hat should have been formed, as well as that 
of the body ; but if the delineation be correct, it appears to 
me, that the whole length of the figure with the hat on is 
given j and that the head, resting on the back of the hat, has 
formed the thick appearance represented, aided, perhaps, by 
some derangement of the frock. The objections that occur to 
me respecting your theory, are — first, that there was scarcely 
time for putrefaction to have commenced so as to have generated 
any heat. Secondly, that if such hot streams had risen, they 
would probably have spread on all sides under the ice, leaving 
no abrupt outline, much less an accumulated one. And 
thirdly, that if heat confined to this space had been the cause, 
the ice must have been thinner over the body than in other 
parts of the pond, which was not the case. On the supposition 
of bubbles from the bottom being the cause of the phenome- 
non, the precision of the figure is satisfactorily accounted for, 
excepting the alleged want of the hat, which, if it be not a 
mistake, militates equally against the other theory, as no good 
reason can be given why the wet hat should be a w'orse con- 
ductor of heat, than the rest of the wet clothing, particularly 
where consisting of many folds. The plate expresses a greater 
breadth of outline at those places where the figure is the thick- 
est, and this would be a necessary consequence of bubbles, as 
their accumulation at any point must correspond with the mag- 
nitude of surface at the bottom, intersected by that particular 
part of the figure. Upon the whole, therefore, I am inclined 
to think, that this curious fact is more likely to have been 
paused by bubbles than by heat, although if it could be clearly 
made out that there was no farther representation of the head 
than 
