210 Mr. Prideaux's Experimental Contributions 



also have taken the copper, in preference to the tin, and the 

 needle would have deviated accordingly. 



But no deviation occurred. 



Mercury, which is pretty near tin in thermo-electric power, 

 was then substituted ; and being heated to the melting point 

 of tin, a warm iron wire was plunged in, which gave a devia- 

 tion of 5° : but iron not coming well into contact with mercury, 

 a tinned iron wire was tried, by which a deviation of 6° was 

 obtained as before (5.). In this slips of tin were successively 

 melted ; but without any manifestation whatever on the 

 needle. 



Here we have no evidence of thermo-electricity, on the dis- 

 appearance of caloric; in the preceding experiments we had 

 no evidence of the disappearance or abstraction of caloric, in 

 the production of thermo-electricity. To what, then, are we 

 to attribute the contrary evidence of the previous experiments, 

 as the apparatus was certainly air-tight? 



6. When the wires were separated, the whole heated surface 

 was exposed to the air in the bulb ; and the portion between 

 the heated ends would be most expanded, and least affected 

 by the cooling influence of the glass. When they were in 

 contact, this, the most effective part of the heated surface, was 

 covered up. Hence a momentary expansion upon opening, 

 and contraction upon closing the contact of the wires. 



But since when thus in contact they cool less quickly, in 

 proportion as they have less effect upon the air; so the whole 

 time of descent, for 50° of the scale, should be greater, rather 

 than less, when closed than when open. 



If this were the true explanation, homogeneous wires should 

 answer as well as those of different metals. Accordingly, two 

 copper wires were passed through a cork, bent and heated as 

 before (3.) ; and the descent of the fluid was found to be sud- 

 denly stopped and resumed, at the instants of opening and 

 closing the contact, just as when the wires were of two dif- 

 ferent metals. 



7. The evidence is therefore opposed to the expenditure of 

 caloric in the production of thermo-electricity, and to the 

 converse. Yet I thought the opposing experiments worth 

 stating, (as they were not unlikely to occur to others,) to show 

 that they had been tried, and shown to be inconclusive. 



III. How far is the Radiation of Heat, or any collateral pro- 

 perty, concerned in the Development of Thermo-electricity ? 



8. The result of M. Becquerel's exact researches, above 

 quoted from the Ann, de Chim. et de Phys. 9 Aug. 1829, is, 

 that the only known property of heat in which the metals take 



