128 THE LIFE OF JAMES D. FORBES. [CHAP. 



intensity for 3,000 feet of vertical ascent. If, as Hum- 

 boldt states, the dip diminishes in*ascending, the dimi- 

 nution of total intensity will be somewhat greater. You 

 will judge of the extent of the inductions upon which 

 this is founded when I mention that the sum of the 

 heights to which I have carried Hansteen's apparatus 

 exceeds 160,000 feet, or thirty vertical miles, twelve 

 lieues. 



' I have lately been engaged in procuring thermometers 

 similar to those at the Observatory at Paris, to be sunk to 

 different depths in various soils. I have three sets from 

 three to twenty-six feet long ; one set to be sunk in 

 trap-tufa, a second in sandstone, a third in pure loose 

 sand. . . The observations in the Lead Hills are being 

 continued/ 



To the REV. DR. WHEWELL. 



'Jan. 31^, 1837. 



' . . . I feel gratified by the prominent place you 

 have given to my experiments as bearing upon the theory 

 of Heat, in w T hich you have done me full justice. . . . 

 But I must mention for yourself, if not for your 

 book, that the discovery of the polarization of heat was 

 not the necessary consequence of applying the thermo- 

 multiplier to the investigation, which would have been 

 a poor achievement, seeing it was another man's invention ; 

 but that Melloni had first applied the instrument to the 

 tourmaline question, and answered in the negative (Ann. 

 de Chimie, vol. 55); then Nobili, the inventor, attempted 

 to repeat Berard's experiment with the most improved 

 piles, and with results quite null (Bib. Universelle). So 

 that I conclude that, when I published my experiments, 

 the question of polarization was negatively answered by 

 persons operating with every advantage which I possessed, 

 and indeed seemed to be set at rest. My discovery was 

 the application of mica as a polarizing substance, first by 

 transmission, then by reflection ; and I have shown that 

 repeating Nobili's experiment the same as Berard's and 

 Powell's the quantity of heat reflected from glass is so 

 excessively minute that the errors might well equal the 



