38 THE ROYAL INSTITUTION. [CHAP. I. 



tive warmth of various substances used in making 

 artificial clothing ; relative quantities of the same 

 substance ; different qualities of substance chemically, 

 as charcoal, ashes, dust. All his experiments indicated 

 that the air which occupies the interstices of substances 

 used in forming coverings for confining heat acts a very 

 important part in that operation. Air is a perfect non- 

 conductor of heat. These experiments were chiefly 

 made in 1787. They were not read before the Royal 

 Society until January 19, 1792. 



Early in the winter of 1787, as soon as the cold was 

 sufficiently intense, he began to repeat the experi- 

 ments of Dr. Fordyce ('Transactions of the Royal 

 Society,' vol. Ixxv.) on the weight said to be acquired 

 by water in the act of freezing ; and, being possessed of 

 a most excellent balance belonging to the Duke of 

 Bavaria, he soon came to the conclusion that all 

 attempts to discover any effect of heat upon the 

 apparent weights of bodies would be fruitless. 



He had previously, in April 1785, convinced himself 

 of the errors that arose from currents of air and from 

 the drying of the cords by which the scales were hung. 



These experiments were made into a paper entitled 

 6 An Inquiry concerning the Weight Ascribed to Heat.' 

 This was read before the Royal Society, May 2, 1799. 



In May and June 1786 he made experiments on 

 the production of air from water exposed to light. 

 These were read before the Royal Society, Febuary 15, 

 1787. 



When engaged in his experiments on the conducting 

 powers of various bodies with respect to heat, and 



