July 2 2, 1875I 



NATURE 



235 



ELECTRICAL RESISTANCE THERMOMETER 

 AND PYROMETER''' 



THIS paper consists of three parts. The first treats of the 

 experiments made by Dr. Siemens, with a view of deter- 

 mining the law of the variation of electrical resistance in metallic 

 conductors, with variation of temperature, through a greater 

 range than had been before attempted. Tlie second describes 

 certain instruments, by whose use this law is applied to the 

 measurement of temperature. The third treats of a simple 

 method of measuring electrical resistance by means of the 

 differential voltameter. 



Our author first refers to the previous experiments made by 

 Arnsted, by his brother, Dr. Werner Siemens, and by Dr. 

 Matthiessen, and to the law deduced by Clausius, " that the 

 electrical resistances of metals are directly proportional to their 

 absolute temperatures." The maximum range of these experi- 

 ments was 100° C. Dr. Siemens's experiments were made upon 

 copper, iron, steel, silver, aluminium, and platinum ; the last of 

 these has received the most attention at his hands, as, having the 

 highest melting point, it is the most valuable from a practical 

 point of view. 



The method employed in one series of experiments was to 

 wind metal wire upon pipe-clay cylinders, having helical grooves 

 to prevent contact between the convolutions of the wire, and to 

 place these, together with three delicate thermometers, in a 

 copper vessel enclosed in a larger one 

 containing linseed oil, and having 

 hollow sides packed with sand to 

 diminish sudden variation of tempera- 

 ture. The bath was gradually heated 

 by means of Bunsen's burners to 340° 

 C, or close to the boiling point of 

 mercury, and the readings were made 

 with a Wheatstone's bridge and deli- 

 cate galvanometer. A second series 

 of experiments was made in a heated 

 air vessel having a metallic screen to 

 prevent irregular losses of heat by 

 hation or by atmospheric cun-cnts, 

 (,' other conditions being similar to 

 , ;se m the first series. The results 

 obtamed were found to accord gene- 

 rally with those of Matthiessen and tie 

 other observers within the limits of 

 their experiments, but pointed to a 

 different law of increase beyond those 

 limits. The formula hitherto known 

 as Matthiessen's is — 



^ ^. ^.__ 



* I - -0037647^+ -00000834^' 

 and was the mean of the results ob- 

 tained on various metals. This for- 

 mula is shown to give discordant 

 results at the higher temperatures, as 

 the calculated resistance at 300° C. 

 is i-6l nearly of what it is at 0° C, whilst at 2000° C. it is "0373, 

 showing clearly that the formula is reliable only between very 

 narrow limits. . •,. , , 



We quote the author as to the law of resistance which he pro- 

 poses : "Now, if we apply the mechanical laws of work and 

 velocity to the vibratory motions of a body which represent its 

 free heat, we should define this heat as directly proportional to 

 the square of the velocity with which the atoms, or may be the 

 molecules, vibrate. 



"We may fmther assume that the resistance which a metallic 

 body offers to the passage of an electrical impulse from atom to 

 atom, or from molecule to molecule, is directly proportional to 

 the velocity of the vibrations which represent its heat. In com- 

 bining these two assumptions, it (oUow's tliat the resistance of a 

 metallic body increases in the direct ratio of the square root of 

 the free heat communicated to it. Algebraically, if r represent 

 the resistance of a metallic conductor at the temperature T, 

 reckoning from the absolute zero, and o, an experimental coeffi- 

 cient of increase peculiar to the particular metal under considera- 

 tion, we should have the expression— 

 r = oxi. 

 This purely parabolic expression would make no allowance for 

 * Abstract of a Paper read at the Society of Telegraph Engineers by C 

 William Siemens, D.C.L., F.R.S., &c. 



Fig. I. 



the probable increase of resistance, due to the increasing distance 

 between adjoining particles with increase of heat, which would 

 depend upon the coefficient of expansion, and may be expressed 

 by /3 T, which would have to be added to the former expression. 

 To these factors a third would have to be added expressing an 

 ultimate constant resistance of the material itself at the absolute 

 zero, and which I call 7. The total resistance of a conductor at 

 any temperature, T, would, therefore, be expressed by the 

 formula — 



Diagrams are given in which this hypothetical law is graphically 

 represented, and in which its results are compared with those 

 obtained by the experiments already cited, and by this means 



the 'following formulae are arrived at for the diiTerent metals 

 named :— 



For platinum 



•0021448 



r = -039369 



r = -092183 



For copper ... r= -026577 



,, iron ... r = -072545 



,, aluminium r — '0595 1436 li 



,, silver ... r = -0060907 T-'' 



th + 

 xi -H 

 Xi-f- 

 xi + 



•0024187 X 

 •00216407 X 

 •00007781 X 

 •0031443 X 

 •0038133 X 

 •00284603 X 

 •0035538 X 



•30425 

 •24127 



•50196 



•29751 

 •23971 

 •76492 

 •07456 



Dr. Siemens, however, has not been satisfied with limiting 

 his experiments to temperatures within the boiling point of mer- 

 cury, but compared the law he had deduced with experimental 

 results at higher temperatures obtained by the use of the metal 

 ball pyrometer shown in Fig i. Its principal parts are a 

 metal ball, whose heat capacity equals one-fiftieth of that 

 of an imperial pint of water, a copper vessel containing a pint of 

 water, and a thermometer having a fixed and sliding scale with 

 divisions of equal size, but each division in the latter being equi- 

 valent to fifty in the former. The zero of the sliding scale is 

 fixed to coincide with the position of the mercury level in the 

 thermometer. The ball, having been heated, is dropped into the 

 water, whose temperature is the sum of those indicated on the 

 fijced and sliding scales. By the use of this instrument, whose 

 readings were comparal with those of the mercury thermometer 



