VOL. LXXVI.] PHILOSOPHICAL TRANSACTIONS. 13Q 



of injuring the surface by paring or rubbing, I mark on the ends the degrees 

 which they respectively exceed or fall short ; which degrees are accordingly t6 

 be subtracted, or added, in all observations of heat made with those pieces* 

 It may be proper to take notice of an irregularity in the apparent diminutions of 

 the pieces, which was sometimes observed to happen from another cause, their 

 bending a little in strong fire, so as to be prevented from going so far in the 

 gage as they would have done if they had continued perfectly straight. But as this 

 takes place only in pieces of considerable length, and as they derive no advan- 

 tage of any kind from that length, the remedy is too obvious to need being here 

 mentioned. 



Another fallacious appearance arose, not from any imperfection in the pieces 

 themselves, but from a deception with respect to the heat in which the com- 

 parison of them had been made. In one period of the course of my experiments 

 I employed, for firing them, a small, shallow, cylindrical vessel, setting the 

 pieces on end, close together, on its bottom, and placing it in the middle of 

 the fuel, in a common air furnace, with care to keep the fire as equal all round 

 it as possible. It was expected, that all the pieces would receive an equal heat ;, 

 and as they were found, after the operation, to differ in their dimensions, some- 

 times considerably, from each other, these differences proved a source of much, 

 perplexity, till it was discovered that the pieces had really undergone unequal, 

 degrees of heat. 



In the paper on the comparison of this thermometer with Fahrenheit's, I 

 have taken notice of the great difficulty of obtaining, in small furnaces, a per- 

 fectly equal heat, even through the extent occupied by a few of these little 

 pieces : and how ditFerent the heat may be in different parts of one vessel, we 

 may be satisfied by an easy experiment, viz. setting a cylindrical rod of clay, of 

 the length of 8 or 10 inches, upright in the middle of a crucible, and urging it 

 with strong fire in a common small furnace ; the rod will be found very dif- 

 ferently diminished at different parts of its height ; and if its height be sufficient 

 to reach some way above the fuel, nearly the whole range of the thermometric 

 scale may be produced in one rod ; an ocular proof not only of the diversity of 

 heat within a small compass, but likewise of the peculiar sensibility of this ther- 

 mometer, every part of the mass expressing distinctly the degree of heat which 

 it has itself felt. It will be proper, in this experiment, to have a tube fixed in 

 the bottom of the crucible, for keeping the rod steady. By this means the heat 

 of my air-furnace renders a rod of the thermometric clay tapering, from about 4 

 parts in diameter at top to 3 at bottom, which are nearly the proportions be- 

 tween the width of the piece when unburnt, or but just ignited, and when it 

 has suffered a heat of 1 6o degrees. By due attention to the circumstances 



T 2 



