Experiments and Ohfervations. 599 
hermetically, and on the 20th of May, 1779, two of 
them were fufpenaed to a nail out of a window, and the 
thiud was wrapped up in paper, and was inclofed in a 
box, where not the leaft glimmering of light could enter. 
In this fituation they were left for above a year, and 
laft week I broke one of thofe that had been kept out of 
the window, and that which had been in the dark, in 
prefence of Mr. rirwan, f. r. s.; but the pyrophorus 
of each tube feemed to be equally good, taking fire 
within about half a minute after being taken out of the 
tubes, and expofed to the air upon pieces of paper, which 
fhews that neither the prefence or abfence of light had 
injured it. 
June 7, 1780. 
P.S. Having mentioned to fe veral perfonsmy intention 
of making fome experiments upon the temperature of 
the atmofphere with a new metalline thermometer, and 
of giving an account of them in this paper, I muft hero 
mention, that as the inftrument was not finilhed in. 
proper time, I fhall defer giving a defcription of it, as 
well as an account of the experiments, 8ec. to another 
opportunity. 
•zus*- 
