Systems of Scientific Units of Measurement. 179 



should be placed, though in my opinion the above order seems 

 the true one, and I think most people would agree in putting 

 Correlation at the top of the list. Its importance is so great 

 that it has in fact been more completely realized than any of 

 the others, most physical ideas having their due formulae 

 assigned and recognized. It has been the more easy to do 

 this, since the process is absolutely independent of the real 

 units employed. The ideas need only to have their physical 

 nature truly comprehended. 



In the second point — Simplicity — there has been much less 

 progress made. In electrical matters, for instance, though in 

 describing a magnitude it may be perfectly well understood 

 what units of length, mass, and time form the fundamental 

 basis of the measurement, it is further necessary to state 

 whether the Electrostatic or the Electromagnetic system is 

 contemplated. 



Again, heat is recognized as a form of energy, and yet the 

 two are not' referred to the same unit. If they were so, Joule's 

 equivalent would always be unity. The defects in Compre- 

 hensiveness have been lately discussed in this Society by 

 Prof. Riicker*. The one with which I have been brought 

 most in contact is the omission of temperature from the 

 systems in use. I think that if generalizations as great as 

 the Electromagnetic Theory of Light are to be made by the 

 help of systems of units, the Thermometer ought at once to 

 be brought in out of the cold region of unsympathetic isola- 

 tion and independence. 



In chemical physics are to be found some notable omissions 

 in atomic heat and electrochemical equivalence. When we 

 examine the units to note how many are founded upon natural 

 magnitudes, we recognize that the second is not a subdecimal 

 part of the solar day, but only the 8^400 of that natural unit. 

 To set against this, however, it is a unit which is in common use. 



The centimetre is decimally connected with the earth^s 

 quadrant, and the unit of density is that of water, the gramme 

 merely being derived from the centimetre and density of water 

 by proper correlation. 



In the quadrant-volt second, or practical system, the length 

 unit is brought up to practical coincidence with the earth- 

 quadrant ; the second, as before, is the unit of time ; and the 

 other quantities are only decimally connected with the C.G.S. 

 I am not aware that any accidental coincidences occur between 

 other units of this system and natural ones, the nearest ap- 

 proach being probably the volt to theE.M.F. of a Daniell cell. 



But it is when we compare the units with those of an 

 arbitrary nature perhaps, but in common use, that the failure 

 * See p. 104 of the present Number. 

 N2 



