PBESIDENTIAL ADDBESS — SECTION A. 27 



small variations of temperature or excessively mimite amounts 

 of heat -radiation. Ten years ago a good thermopile ^Yith a very 

 sensitive galvanometer was considered almost beyond being 

 improved on. This apparatus, with the assistance of a large 

 refractor or reflector, was delicate enough to indicate the heat 

 received from some of the brighter stars, but, when you know 

 that at present we have instruments more than one thousand 

 times as sensitive, you will understand how wonderful the 

 advance has been. 



We always measure change of temperature by the resulting 

 change in one or other of the properties of a piece of matter. 

 The resistance of a conductor to an electric current is always 

 increased if its temperature be increased, and we can measure 

 any variation in the resistance of a conductor wdth enormous 

 accuracy. 



Making use of this principle. Professor Langley in 1881 

 invented an instrument which he called a bolometer. It con- 

 sists essentially of a Wheatstone's bridge with a very delicate 

 galvanometer. The arms 7-1 and r.2 of the bridge are made of a 

 substance whose resistance varies considerably with tempera- 

 ture (usually thin strips of iron). A balance is made when both 

 o'l and '}\ are unexposed, or at the same temperature. If now 

 one arm, ri, be screened while the other, r.,, is exposed to radia- 

 tion, the galvanometer will be at once deflected, due to the 

 increase in the resistance of r., disturbing the balance ; and a 

 modification of the usual formula for the bridge will enable you 

 to calculate from the observed deflection of the galvanometer the 

 change of resistance, and hence the increase in temperature, of 

 rj, the exposed arm. This is, in general terms, the theory of 

 the instrument ; but of course very great precautions must be 

 taken in its manufacture and arrangement so as to exclude 

 disturbing influences. With a good modern galvanometer this 

 instrument will measure a variation of temperature of xotjVtoT) 

 of a degree centigrade. 



With this instrument its inventor, Professor Langley, has 

 conducted a most interesting and valuable investigation, extend- 

 ing over four years, about the temperature of the moon. His 

 most interesting conclusion was that the mean temperature of 

 the sunlit lunar soil is very much lower than was formerly sup- 

 posed, and cannot be very much above the freezing-point of 

 water. In addition, he discovered many facts of great meteoro- 

 logical interest. 



Wonderful as this bolometer of Langley is, it has been 

 superseded by another instrument still more wonderful, the 

 radio-micrometer of Professor Eoys. The working of this 

 instrument depends on properly utilising the fact that if a 

 complete circuit be made of two or more metals, and if one 

 junction be at a different temperature from that of the rest of 



