[ 447 ] 
have been at the moft, but about blood warm. For 
jlnce a difference of heat , between the two Jides , is ab~ 
folutely neceffary , it is of importance that this difference 
be incrcajed as much, and as feuddenly , as the nature 
of the circumjlances will admit. 
Jf you follow this rule, I apprehend you will be 
more likely to fucceed : but that you may be the bet- 
ter enabled to make my experiments aofwer, it may 
not be amifs to relate them again in the manner they 
were lately repeated, and frequently, with two different 
‘Tourmalins , before feveral members of the Royal- 
Society of London , who are well acquainted with en- 
quiries of this kind. 
EXPERIMENT. 
After the convex fide of the Tourmalin ( b , fig. r. 
Tab. XXII.) has been held for a fhort time, about 
one tenth of an inch, from the flame of a candle, 
both fides thereof are elettrified plus: and con- 
tinue fo for half a minute or more. And in a 
fhort time after, the fame Tourmalin , without being 
heated a frefh, or difturbed by any other caufe than 
that of the air furrounding, returns to its natural Rate ; 
as you have called it; that is, th z plain fide changes to 
minus, and the convex fide remains plus. Thefe appear- 
ances continue whilffc the Tourmalin is cooling. 
Upon heating the Tourmalin again, as before, ex- 
cepting that the plain fide was now next the flame, 
both fides thereof were ele&rified minus : and conti- 
nued fo, for half a minute or more ; and in a fhort 
Mmm 2 time 
