THE MEASUREMENT OF TEMPERATURE. 



IN these days of specialisation the methods of investiga- 

 tors differ as widely as their aims, and apparently they 

 have little in common except the desire to do their utmost 

 for the advancement of science. Let us suppose an un- 

 scientific but intelligent visitor to pass, say, from a physical 

 to a chemical, and thence to a biological laboratory. Except 

 that in each are many instruments built up of glass and 

 metal their contents would present but little similarity, nor, 

 with one exception, would repetitions of the same apparatus 

 be noticeable. One instrument, however, would be com- 

 mon to all, would be found in every room, possibly on 

 every table, and inquiry as to its name would, I need 

 scarcely say, receive the reply : " A mercury thermometer ". 

 Our visitor would be led to the natural conclusion that such 

 an essential instrument to which so much attention must 

 have been devoted would have arrived at a high degree of 

 perfection and that its history would be one of progress 

 and evolution. 



Such being the case it is strange to reflect that the mer- 

 cury thermometer has remained practically unchanged from 

 the time (17 14) when Fahrenheit first suggested the use of 

 certain fixed points to the present day. 1 True, Cavendish in 

 1780 investigated the conditions under which observations 

 with this instrument must be made, such as the effect of the 

 immersion of the stem, etc.,' 3 and in recent times the labours 

 of Regnault, Rowland, Crafts, Pickering and Guillaume have 

 added much to our knowledge of its imperfections ; although, 

 with the exception of the last named, they have done but 

 little to help us to remedy them. 



1 The origin of the Fahrenheit fixed points is so little known that I may 

 be excused a reference to it. His lowest "natural cold" was that of his 

 freezing mixture, his " highest natural heat " that of the human body. Thus 

 the freezing and boiling points of water are, on that scale, consequent,, 

 not original ones (see Gamgee, Cam. Phil. Soc. Proc). 



2 Thorpe, Essays o?i Historical Chemistry, p. 75. 



