^io Dr. roebuck’s Experiments 
I heated a piece of iron, of nearly one pound weight, to 
a white heat, or what the fmiths call a welding heat, and 
found, by the moil accurate experiments which I could 
make, and which I again and again repeated, that the 
iron, when left feveral hours in the balance to cool, 
weighed nearly one grain lefs when cold than when hot; 
and that a piece of iron, of about five pennyweights, which 
was tried by the fmaller and more accurate balance, 
weighed, as appeared by an index which moved oppoiite 
to a quadrant, fomewhat more when cold than when 
hot. I tried the fame experiment on copper; but, to my 
great furprize, 1 found a piece of copper, of nearly one 
pound weight, four grains lighter after it had been left 
fome hours to cool in the balance than when it was put 
in. 1 repeated the experiment, and found the event the 
fame; but fufpedting this might poffibly arife from the 
copper calling fcales, I placed a fheet of white paper un- 
der the balance, and collected as many fcales as made up* 
nearly the deficiency of the weight. 
TO SIR JOHN PRINGLE, HART. P. R. S- 
London, 
^ -*• ■“? May g, 1776; 
9* T BEG leave to inform you of the following 
experiments, which were made to afeertain 
the variation of the weight of bodies when hot and cold. 
April 
