Tendency of Elastic Fluids to Diffusion. 13 



Having now related all the experiments I made of any 

 importance to the subject, it will be proper to add, for the 

 sake of those that may wish to repeat some of them, that 

 great care must be taken to keep the inside of the tube dry ; 

 for if a drop of water interpose between the two gases, I 

 have found that it effectually prevents the intercourse. 

 Glass tubes should therefore be used, that one may be sa- 

 tisfied on this head, as the obstruction will then be visible. 



I shall make no further comments on the above experi- 

 ments bv way of explanation; because, to those who under- 

 staad my hypothesis of elastic fluids, they need none : and 

 I tbiak it would be in vain to attempt an explanation any 

 other way. I cannot however, on this occasion, avoid ad- 

 verting to some exj.eriinents of Dr. Priestley, which few 

 modern philosophers can be unacquainted with : I mean those 

 relating to the seemina; conversion of water into air. {Vide 

 Philos. Transact, vol. Ixxiii. p. 414; or his Experiments 

 abridged, vol. ii. p. 407.) He found that unglazed earthen 

 retorts, containing a little moisture, when heated, admitted 

 the external air to pass through their pores at the same time 

 that aqueous vapour passed through the pores the contrary 

 way, or outward ; and th^t this last circurastaTice was ne- 

 cessary to the air's entrance. The retorts are air-tight, so 

 far as that blowing into them discovers no pores ; but 

 when subjected to a greater pressure, as that of the atmo- 

 sphere, or even one much short of it, they are not able to 

 prevent the pas;-.age of elastic fluids. The fact of air passing 

 into the retort through its pores, and vapour out of them at 

 thesanie time, arc elegantly and most convinaingly shown by 

 Dr. Priestley's experiments, in which he used the apparatus 

 represented in Plate V^II. fig. 1. of the edition above referred 

 to. The doctor confesses his explanation of these remarka- 

 ble facts is very inadequate; and no wonder, for it is im- 

 possible for him or any other to explain them on the com- 

 monlv received principles of elastic fluids. But we will hear 

 what he says on the subject : — " At pre.->cnt it is my opi- 

 nion that the agent in this case is that principle which we 

 call attraction of cohesion, or that |'>ower by which water 

 is raised in capillary tubes. But in what manner it acts in 



this 



