yo SCIENCE PROGRESS. 



range. It is difficult to account for the fact that such an 

 instrument remains comparatively neglected, and I am glad 

 of this opportunity of bringing its claims under the notice 

 of so large a number of scientific workers. I propose to 

 briefly review the evidence we now possess concerning the 

 accuracy of this method of determining temperatures. 



The bibliography of this subject is small, and at the end 

 of this article will be found what I believe is a fairly com- 

 plete table of such communications as have been published. 

 I shall refer to these papers by the figures allotted to them 

 in that list. 



I do not propose to enter into any details as to the 

 method of making or using these instruments as full infor- 

 mation on such points will be found in papers (7), (8) 

 and (9) ; for my object is rather to present the evidence 

 in favour of their accuracy in such a manner that its 

 cumulative weight may be appreciated. It may be advis- 

 able to first clear the ground by disposing of some objec- 

 tions which I find are commonly entertained by (I may 

 remark) those who have had no experience in the use of 

 the instrument. 



1. The general scepticism as regards the constancy 

 of the platinum thermometer may, I think, be traced to the 

 adverse report delivered by the influential committee ap- 

 pointed by the British Association in 1873 (1). In 1874 

 they reported that the indications of the instruments ex- 

 amined changed by as much as 50 C. after heating in a 

 common furnace, and also that the fixed points altered by 

 mere lapse of time. It must, however, be remembered that 

 this committee was appointed to examine, and did examine, 

 but one form of apparatus, viz., the Siemens pyrometer, and 

 their adverse conclusion should in no way affect the reputa- 

 tion of the modern form of instrument. 



In Siemens's pyrometer a platinum wire was wound on 

 common clay, placed in an iron tube, and exposed to a 

 high temperature. Changes exceeding those found by the 

 committee would be anticipated by observers who have 

 studied the behaviour of platinum wires under such circum- 

 stances, and my own feeling is one of surprise that the 



