4 2 4 EXPERIMENTAL INVESTIGATIONS, 



water by ignited platinum, it may be that the sudden appli- 

 cation of intense heat, and in some quantity, so forces asunder 

 the molecules that, not having sufficient nitrogen dissolved to 

 supply them with a nucleus for evaporation, the integral 

 molecules are severed, and decomposition takes place. If 

 this be so, and it seems to me by no means a far-fetched 

 theory, there is probably no such thing as boiling, properly so 

 called, and the effect of heat on liquids in which there is no 

 dissolved gas may be to decompose them. 



Considerations such as these led me to try the effect of 

 boiling on an elementary liquid, and bromine occurred as the 

 most promising one to work upon. As bromine could not be 

 boiled in contact with water, oil, or mercury, the following 

 plan was ultimately devised : A tube, four feet long and T %t ns 

 inch diameter, had a platinum loop sealed into one closed ex- 

 tremity ; bromine was poured into the tube to the height of 

 four inches ; the open end of the tube was then drawn out to 

 a fine point by the blow-pipe, leaving a small orifice ; the bro- 

 mine was then heated by a spirit-lamp ; and when all the air 

 was expelled, and a jet of bromine vapour issued from the 

 point of the tube, it was sealed by the blow-pipe. There was 

 then, when the bromine vapour had condensed, a vacuum in 

 the tube above the bromine. The platinum loop was now 

 heated by a voltaic battery and the bromine boiled ; this was 

 continued for some time, care being taken that the boiling 

 should not be too violent. At the end of a certain period 

 from half-an-hour to an hour the platinum loop gave way, 

 being corroded by the bromine : the quantity of this had 

 slightly decreased. On breaking off, under water, the point 

 of the tube, the water mounted and showed a notable quantity 

 of permanent gas, which on analysis proved to be pure oxygen. 

 As much as a quarter of a cubic inch was collected at one ex- 

 periment. The platinum wire, which had severed at the 

 middle, was covered with a slight black crust, which, suspect- 

 ing it to be carbon, I ignited by a voltaic spark in oxygen in a 

 small tube over lime-water ; it seemed to give a slight opa- 

 lescence to the liquid, but the quantity was so small that the 

 experiment was not to be relied on. No definite change was 

 perceptible in the bromine ; it seemed to be a little darker in 



