EFFECTS OF HEAT ON FLUIDS. 429 



All these facts show that the effect of intense heat applied 

 to liquids and gases is far more complex, and presents greater 

 interest to the chemist, than has generally been supposed. In 

 far the greater number of cases, possibly in all, it is not mere 

 expansion into vapour which is produced by intense heat, but 

 there is a chemical or molecular change. Had circumstances 

 permitted I should have carried these experiments farther, and 

 endeavoured to find an experimentum crucis on the subject ; 

 there are difficulties with such substances as bromine, phos- 

 phorus, &c., arising from their action on the substances used 

 to contain and heat them, which are not easy to vanquish, 

 and those who may feel inclined to repeat my experiments 

 will find these difficulties greater than they appear in narra- 

 tion ; but I do not think they are insuperable, and hope that, 

 in the hands of those who are fortunate enough to have time 

 at their disposal they may be overcome. 



To completely isolate a substance from the surrounding 

 air and yet be able to experiment on it, is far more difficult 

 than is generally supposed. The air-pump is but a rude in- 

 strument for such experiments as are here detailed. 



Caoutchouc joints are out of the question ; even platinum 

 wires carefully sealed into glass, though, as far as I have been 

 able to observe, forming a joint which will not allow gas to 

 pass, yet it is one through which liquids will effect a passage, 

 at all events when the wires are repeatedly heated. 



In some experiments with the ignited platinum wire her- 

 metically sealed into a tube of glass, the end of the tube 

 containing the platinum wire was placed in a larger tube of 

 oil to lessen the risk of cracking the glass. After some days' 

 experimenting, though the sealing remained perfect, a slight 

 portion of carbon was found in the interior liquid. This does 

 not affect the results of my experiments, as I repeated them 

 with glass tubes closed at the end and without platinum wires, 

 and also without the oil-bath, -but it shows how difficult it is 

 to exclude sources of error. When water has been deprived 

 of air to the greatest practicable extent it becomes very avid 

 for air. The following experiment is an instance of this : A 

 single pair of the gas-battery, the liquid in which was cut off 

 from the external air by a greased glass stopper, had one 



