Platinum Temperatures. 



455 



R. 



t observed. 



t calculated. 



v. 



27-56 



o 



1151 



o 



114-5 



+•6 



23-13 



139-1 



138-2 



4-9 



21-87 



145-5 



144-9 



+ •6 



so that (12) is preferable to (13). 



For the agreement between hydrogen-temperatures and 

 temperatures on the air-thermometer, they give 



— 189°'9 hydrogen-temperature =— 190 o, 6 air-temperature, 

 -186°-4 „ „ „ =-186°-9 „ „ 



or, near the temperature of liquid air the reading of their 

 hydrogen-thermometer was -6 of a degree higher than the 

 reading of the normal air-thermometer. 



Mr. E. H. Griffiths* s Observations. — In two papers pub- 

 lished in the Transactions of the Royal Society for 1891 (A), 

 the first (Gr.) by Mr. E. H. Griffiths, and the second (C. and 

 Gr.) by Messrs. Callendar and Griffiths, a series of most 

 careful observations on the change of the resistance of pla- 

 tinum with temperature has been recorded in connexion with 

 " platinum-temperatures.'''' 



In (Gr.) t — -or is plotted as a function of to- by means of a 

 formula giving t— -sr in powers of -ur as far as the fourth 

 power. For example, for a particular platinum-thermometer 

 called G, the formula is 



t-v= --022,915,5 w + -000,260,768^ 



- -000,000,401,3 *r 3 + -000,000,000,85 ™\ 



(14) 



This is the equation of a curve somewhat like a parabola, but, 

 with the given constants, a little flatter in one part than a 

 parabola would be, without the curvature at any point ceasing 

 to be continuous. 



Mr. Griffiths was led to this curve by the consideration of 

 eight observations, namely, the temperatures of melting ice, 

 and of the boiling-points (at 760 millim. pressure) of water, 

 aniline, naphthalene, methyl salicylate, benzophenone, mer- 

 cury, and sulphur. An attempt was made to " draw a smooth 

 curve " through these points when plotted ; but it was found 

 that while such a curve could be drawn through five of the 

 points, those representing mercury, methyl salicylate, and 

 aniline lay quite off the curve. The five points were there- 

 fore used as " fixed points " from which formulas like (14) 



