140 Prof. C. I. Burton and Mr. W. Marshall. [June 18, 



crystallising of the substance. The crystals melted rapidly at the 

 temperature of the room. 



MonocMor ethane. The whole apparatus was cooled to C. for this 

 series of observations. 



Propyl Bromide and Isobutyl Bromide. A slight " kick " of the 

 galvanometer image in the opposite direction to that indicating a rise 

 of temperature was observed on the first stroke of the pump. 



Iodides. The free iodine was extracted by shaking with mercury. 



Carbon Disulphide. The temperature rose slowly to within a 

 degree of maximum ; then there was a sudden rise to maximum, 

 succeeded by a rapid fall. 



Looking merely at the rise of temperature produced by pressure, 

 it is impossible to deduce any general laws from these figures. We 

 find that, as a rule, in comparable series, the higher the molecular 

 weight, the less the rise of temperature. This is best seen in the 

 cases of the series of acetates and the halogen substances ; but there 

 are several perfectly distinct exceptions such as the fatty acids 

 which come in the inverse order to that stated. It is worthy of 

 note that these exceptional series all contain large proportions of 

 oxygen. 



We would draw special attention to the effect of pressure on par- 

 aldehyde and tertiary butyl alcohol, both of which are caused to 

 solidify at temperatures above their normal melting point. We think 

 it probable that all substances which follow the ordinary law of ex- 

 pansion by heat could be solidified by pressure if tried at tempera- 

 tares not far from their melting point. The apparatus which we 

 used serves extremely well for observing the course of events in such 

 cases. 



When a liquid which does not solidify is compressed, the image on 

 the galvanometer scale travels almost steadily and rapidly to a point 

 at which it remains fixed for a few seconds ; it then begins to move 

 back towards zero, as the substance is cooled by the water outside 

 the compression tube. In the two cases mentioned above (paralde- 

 hyde and tertiary butyl alcohol) the behaviour of the galvanometer 

 was entirely different, indicating a rapid rise of temperature on 

 application of the pressure, and then a continuous slow increase as 

 long as the pressure was maintained constant. The method of 

 closing the gun by means of a ram renders it easy to remove and 

 inspect a substance within half a minute after letting off the pressure, 

 so that there is not time, for the crystals to melt. 



Experiments with Solids. 



A number of specimens of metals were prepared and tested, with 

 the results given in tabular form below. Of each metal two small 



