20 
Dr, Priestley’s Experiments, &c. 
my experiments in general ; but as they fpeak of the water 
produced as appearing both on the in fide of the veffel, and on 
the fur face of the mercury, it can be no or! ' n J me- experi- 
ment of the revival of iron from finerv cinder; n t; ter 
•/ 
that is found in this procefs was never iuppofed to co- e from 
the little that is contained in the inflammable air, but the 
much greater quantity contained in the binder. 
Before I conclude this Paper, I (hall juft mention a few cir- 
cumftances attending the many explofions 1 have made of in- 
flammable and dephlogifticated air in the long metallic and glafs 
veflels I have made ufe of, as they were pretty remarkable. 
The explofions were made by a fmall eledtric fpark at one end 
of the veffel, and the greateft force of the explofion was always 
at the other end. No tinned iron veffel could bear many of 
them before they fwelled out at that end, and at length burfb ; 
and even the flat end of the copper veffel, which was not lefs 
than one-tenth of an inch thick, was in time made quite convex, 
and the cylindrical part next to it was made very fenfibly 
wider than any other part of the tube. This muft have been 
effe&ed by mere force , and not by heat ; for the hotteft part 
of the tube, after every explofion, was never there, but always 
about the middle, though fomething nearer to that end than 
the other, and in the glafs veffel the denfe cloud was always 
formed at that end. 
The probability is, that the air where the ele&ric fpark 
is made taking fire firft, the inflammation does not extend 
itfelf fo rapidly but that the air at the oppofite end is firft 
condenfed, in confequence of the inflammation and expanfion 
of the air at the other end, fo that the air is there fired in a 
condenfed ftate ; and hence its greater force* 
