22 CORRELATION OF PHYSICAL FORCES. 



by a given mechanical action. His mode of experimenting 

 is as follows : An apparatus formed of floats or paddles of 

 brass or iron is made to rotate in a bath of water or mercury. 

 The power which gives rise to this rotation is a weight raised 

 like a clock-weight to a certain height ; this by acting during 

 its fall on a spindle and pulley communicates motion to the 

 paddle-wheel, the water or mercury serving as a friction 

 medium and calorimeter ; and the heat is measured by a 

 delicate mercurial thermometer. The results of his experi- 

 ments he considers prove that a fall of 772 Ibs. through a 

 space of one foot is able to raise the temperature of one pound 

 of water through one degree of Fahrenheit's thermometer. 

 Mr. Joule's experiments are of extreme delicacy he tabu- 

 lates to the thousandth part of a degree of Fahrenheit, and a 

 large number of his thermometric data are comprehended 

 within the limits of a single degree. Other experimenters have 

 given very different numerical results, but the general opinion 

 seems to be that the numbers given by Mr. Joule are the 

 nearest approximation to the truth yet obtained. 



Hitherto I have made no distinction as to the physical 

 character of the bodies impinging on each other ; but Nature 

 gives us a remarkable difference in the character or mode of 

 the force eliminated by friction, accordingly as the bodies 

 which impinge are homogeneous or heterogeneous : if the 

 former, heat alone is produced; if the latter, electricity. 



We find, indeed, instances given by authors of electricity 

 resulting from the friction of homogeneous bodies ; but, as I 

 stated in my original Lectures, I have not found such facts 

 confirmed by my own experiments, and this conclusion has 

 been corroborated by some experiments of Professor Erman, 

 communicated to the meeting of the British Association in 

 the year 1845, in which he found that no electricity resulted 

 from the friction of perfectly homogeneous substances ; as, 

 for instance, the ends of a broken bar. Such experiments as 

 these will, indeed, be seldom free from slight electrical cur- 

 rents, on account of the practical difficulty of fulfilling the 

 condition of perfect homogeneity in the substances themselves, 

 their size, their temperature, &c. ; but the effects produced 

 are very trifling and vary in direction, and the resultant effect 



