to determining tlie magnitude of material particles . 13 



the liquid being retained, the reverse took place. In process 

 of time, varying from two to four months, every drop of the 

 water was forced out, so that when the vessel was opened it 

 was as dry within as when it came out of the furnace ; yet the 

 gas was retained highly condensed, and made its escape with 

 a loud report. 



On one occasion a violent explosion took place from a blow 

 which I had reason to suppose would not have been sufficient 

 to cause it, had alkaline solution been present to restrain the 

 elasticity of the gas. In one case die outer surface of the 

 ware was frosted over with the efflorescent alkaline carbonate, 

 which had passed through it in a state of solution : the water 

 was all gone, the vessel being as before perfectly dry within, 

 yet the gas apparently all present. 



The power that effected the transmission of the water and 

 alkali being the elastic force of the gas, I could conceive it 

 to act in no other way than mechanically, and to indicate 

 therefore some difference in the size of the elastic and the 

 liquid molecules ; the former being too large to find passage 

 through the pores of this closer v/are, while the latter, both 

 aqueous and alkaline, all escaped. So far the gaseous w^ere 

 heterogeneous to the liquid particles; and the difference of 

 size indicated might be supposed to depend upon a difference 

 of their original magnitude, and not of magnitude arising 

 from their liquid and gaseous condition. But it will be ob- 

 served that the alkaline carbonate travelled out with the water. 

 Here therefore was carbonic acid in a liquid state, forced out 

 through apertures which refused exit to the same matter in 

 the state of gas. By this it would appear that the atoms of 

 matter are indeed invested, as has been supposed, with an 

 sethereal atmosphere elastic in itself, but forcibly drawn around 

 the atom, so that it can find passage only where there is space 

 for its ^ther to accompany it, and that this aethereal invest- 

 ment being much larger around gaseous than around liquid 

 particles, apertures through which the latter may be forced 

 may nevertheless be too small for the passage of the former. 

 It appears to me difficult otherwise to explain the phseno- 

 mena of this experiment. 



Again, in that case in which the reverse took place, namely, 

 an exit of ihegas and a retention of the liquid, and which was a 

 result most persons would expect, inasmuch as it is commonly 

 more easy to make vessels water-tight than air-tight, a suffi- 

 cient explanation may be given which shall not be incompa- 

 tible with the case just recorded, and which may remove the 

 apparent incongruity of the two experiments. In this case, 

 though the pores were free enough to give passage both to 



