May i6, 1872] 



NATURE 



47 



ON MEASURING TEMPERATURES BY 

 ELECTRICITY* 



'■PHE truth revealed to us by one of the younger 

 -'■ branches of physical science, which has been cul- 

 tivated and expounded nowhere more effectually than 

 within these walls, has divested heat and electricity of 

 their mysterious character, and has taught us to regard 

 them simply as '• modes of motion." 



Light also has been shown to be identical in its nature 

 with heat, and the only remaining physical agency, 

 " chemical affinity," has been recognised as a force 

 differing only in "quality of action" from the others. 

 According to these views, force, in whichever type of ac- 

 tion it presents itself, is as indestructible as matter 

 itself, and is therefore capable of being stored up and 

 measured with the same certainty of result. We have 

 a unit of force, or the foot lb., and a unit of heat, or the 

 heat necessary to raise the temperature of i lb. of water 

 through 1° Fah., and it has been already proved that 

 772 units of force are the equivalent value of one unit of 

 heat. Again, the chemical force residing in i lb. of pure 

 coal is equal to about 14,000 heat units, or 14,000 X 

 772 = 10,808,000 ft. lbs. = 4,825 tons lifted one foot high. 

 Questions regarding the quantitative effects of heat 

 present themselves, however, much less frequently for our 

 consideration than questions regarding its intensity, upon 

 which depends the nature of the phenomena surrounding 

 us at every step, both in science and in ordinary life. The 

 instrument at our command for determining moderate in- 

 tensities or temperatures, the mercury thermometer, leaves 

 little to be desired for ordinary use ; but when we ascend 

 in the scale of intensity, we soon approach a point when 

 mercury boils, and from that point upv.'ard we are left 

 without a reliable guide. The result is that we find in 

 scientific books on chemical processes statements to the 

 effect that such or such reaction takes place at a dull red 

 heat, such another at a bright red, a cherry red, a blood 

 red, or a white heat — expressions which remind one rather 

 of the days of alchemy than of chemical science of the 

 present day. 



There are pyrometers, it is true, but these are either of 

 a complex nature, or little reliance can be placed on them. 

 It is my purpose this evening to place before you an 

 instrument by which I hope to fill up to some extent the 

 existing gap. It is the result of occasional experimental 

 research, spread over several years, and it aims at the 

 accomplishment of a double purpose, that of measuring 

 high temperatures, and of measuring with accuracy the 

 temperatures of inaccessible or distant places. 



But before entering upon the details of my subject, 1 

 propose to place before you an instrument which fulfils, 

 in principle, all the conditions essentially necessary in 

 thermometry, and is at the same time the very first in- 

 strument that was ever proposed for measuring tempera- 

 tures. I speak of the air thermometer by Galileo. It 

 can be shown on theoretical grounds that the expansion 

 of a permanent gas at constant pressure is the most per- 

 fect index of temperature. It is, in fact, the degree of 

 energy of the atomical motion in an elastic fluid which 

 determines its volume, and which constitutes at the same 

 time its temperature. 



The air thermometer consists simply of a bulb of glass 

 with a long tubular stem, open to the atmosphere at its 

 extremity. If I heat the bulb (by dipping it, for instance, 

 into boiling water) and put it into a holder, with the hollow 

 stem reaching downward into a cup of mercury, the air 

 within the bulb will no longer communicate directly with 

 the atmosphere, because the mercury is interposed. If 

 now I cool the air within the bulb, by the external appli- 

 cation of iced water, its heat motion will diminish, and 

 its volume would be reduced proportionally, if the external 



* Lecture delivered at the Royal Institution of Great Britain, March ist, 

 1S72, by C. Wiili.im .Siemens, F.R.S. 



atmosphere could enter freely to fill up the vacancy thus 

 created. But inasmuch as the external air cannot enter, 

 a reduction of pressure will take place, which, according 

 to the law of elasticity by Boyle, must be proportionate to 

 the reduction of volume at constant pressure. The 

 difference of pressure thus created between the bulb and 

 the external atmosphere will be balanced by the column 

 of mercury rising up into the tube, and the elevation to 

 which the mercury attains is a true index of the tempera- 

 ture to which the air in the bulb had been previously 

 heated. This is true with regard to all temperatures, from 

 the lowest to the highest, and the instrument mav be 

 termed a universal thermometer. If the bulb could be 

 cooled down to 273° Centigrade below the zero point, it 

 would follow by the law of Charles that the elastic 

 pressure of air would be reduced to nothing, that is to say, 

 the motion of the particles of air, which we call heat, 

 would have ceased, and we should have reached the point 

 of an absolute zero, a point which has been theoretically 

 established also by other means. 



Practically, such an instrument would be most incon- 

 venient ; its indications would have to be corrected by 

 calculation for baroinetrical variations ; the capacity of 

 the descending tube, which contains air not subjected to 

 variation of temperature, would have to be taken into ac- 

 count, and no reliable observations could be arrived at, 

 without taking special precautions, such as are only within 

 reach of the experimental physicist. 



[Tlie other knovi'n methods of measuring ordinary and 

 furnace temperatures were here passed in review, and 

 the limits of their application pointed out. They were 

 classified into — 



Thermometers, by expansion of liquids. 

 Thermometers, by the expansion of solids. 

 Pyrometers, by chemical decomposition of solids, 

 comprising Wedgwood's and Deville's pyrometers. 

 Pyrometers, by observing the melting-point of metals. 

 Pyrometers, by thermo-electricity. 

 Pyrometers, by exposing a copper or platinum ball of 

 known heat capacity to the heat to be ascertained, 

 and of quenching it in a measured quantity of 

 water.] 

 The instrument which forms the subject matter of my 

 di-course presents many points of analogy with the air 

 thermometer, if we substitute '■ electrical resistance in 

 conductors" for "expansion of gases." Both these effects 

 are functions of temperature, increasing with the tempera- 

 ture according to progressive laws, which in the case of 

 the gases we call the law of Charles, and in the case of 

 conductors, the law of " increase of electrical resistance 

 with ternperature." The latter law, which is of recent 

 origin, had already been partially developed by Arndsen, 

 Swanberg, Lenz, and Werner Siemens, when my attention 

 was directed in i860 towards an application of the same 

 to the measurement of temperatures at places inaccessible 

 to the ordinary thermometer. By means of the contriv- 

 ance which 1 shall describe presently, I was enabled to 

 tell, in the testing cabin of a cable ship, the increasing 

 temperature of the interior of the mass cable in the hold, 

 and to prove the necessity of trans-shipment of the same 

 into a vessel fitted with water-tight tanks, which have been 

 resorted to ever since, to avoid the danger of softening the 

 gutta-percha covering. 



I have arranged an apparatus for proving to you in the 

 first instance that the conductivity of a wire of platinum 

 or other metal is greatly influenced by its temperature ; 

 for this purpose 1 direct the current of a galvanic battery at 

 will through two branches of equal resistance, each branch 

 comprising a free spiral wire of platinum and one of the 

 coils of a differential galvanometer. By throwing the power- 

 ful light of an electric lamp upon the face of the dilferen- 

 tial galvanometer, and by throwing the image by means of 

 a mirror :!nd lens upon the screen, the auJience will see 

 any movement of the needle to the right or the left that 



