between Metallic Masses having different Temperatures. 451 



ensue : this manifestly depends upon the same principle, the ad- 

 hesion between two perfect planes being well known to be great 

 in amount. I have not met with so strong a case in the course 

 of my experiments. 



44. We may conclude this head by noticing, that the inter- 

 ference of any foreign matter between the metals (with the ex- 

 ception of the metallic pellicles already mentioned) seems fatal 

 to the experiment. Dust, amalgam, a coating of oxide, or even 

 oil-gilding, stops the vibration. The action of mercury is pro- 

 bably by increasing the adhesion. 



3. Influence of Temperature. 



45. We have seen that the metal of greatest conducting 

 power must have the highest temperature in the combination of 

 two required to produce a vibratory motion. Not merely is there 

 no action between two metals, when the temperature of both is 

 the same with air of an apartment, but likewise when both are 

 raised to any higher temperature, for example, that of boiling 

 water. 



46. I have not ascertained what is the smallest difference of 

 temperature requisite to produce the effect. It varies, of course, 

 with every different pair of metals. With lead and copper, for 

 example, the vibrations will continue much longer than with lead 

 and tin, although in the former case the temperatures tend more 

 rapidly to an equilibrium. 



47. A difference of temperature of 1 50 seems to be sufficient 

 for all practical purposes. 1 icing anxious to investigate the pro- 

 perties of some metals at a definite higher temperature, I heated 

 several bars in a cast-iron vessel full of sand, along with a ther- 

 mometer, having a very long scale ; this vessel was placed in an- 

 other containing oil, and when the temperature had risen to 350 

 the bars were placed upon cold lead. On one occasion I em- 



VOL. XII. PART II. 3 M 



