I2O THE ADDRESSES, LECTURES, ETC., OF 



measurement of work, here exhibited, are not numerous, but are 

 interesting. Among these may be mentioned Professor Colladon's 

 Dynamometrical Apparatus constructed in 1844 ; Eichard's Patent 

 Steam Engine Indicator, an improvement on Watt's, and Mr. G. A. 

 Hirn's Flexion and Torsion Pandynamometers. 



Eighth. The Measurement of Electrical Units of electrical 

 capacity, of potential, and resistance, forms a subject of vast 

 research, and of practical importance, such as few men are capable 

 of doing justice to. It may be questioned, indeed, whether 

 Electrical Measurement belongs to the province of mechanical 

 science, involving, as it does, problems in physical science of the 

 highest order ; but it may be contended on the other hand that at 

 least one branch of Applied Science, that of Telegraphy, could not 

 be carried on without its aid. I am happy to say that this branch 

 of the general subject will be brought before you by my esteemed 

 friend Sir William Thomson, than whom there is no one more 

 eminently qualified to deal with it. I may, therefore, pass on to 

 the next great branch of our general subject, the ninth, Thermal 

 Measurement. The principal instrument here employed is the 

 thermometer, based in its construction, either upon the difference 

 of expansion between two solids, or on the expansion of fluids 

 such as mercury or alcohol (the common thermometer) or upon 

 gaseous expansion (the air thermometer) ; or again, it may be 

 based upon certain changes of electrical resistance, which solids 

 and liquids experience when subjected to various intensities of 

 heat. With reference to these, the air thermometer represents 

 most completely the molecular action of matter which is the 

 equivalent of the expansibility. I shall not speak of the different 

 scales that have been adopted by Reaumur, Celsius and Fahrenheit, 

 which are based upon no natural laws or zero points in nature, 

 and which are therefore equally objectionable upon theoretical 

 grounds. Would it not be possible to substitute for these a 

 natural thermometric scale ? One commencing from the absolute 

 zero, of the possible existence of which we have many irre- 

 futable proofs, although we may never be able to reach it by 

 actual experiment. A scale commencing in numeration from this 

 hypothetical point would possess the advantage of being in unison 

 throughout with the physical effects due to the nominal degree, 

 and would aid us in appreciating correctly the relative dynamical 



