[ 178 ] 



XIX. On some Facts connected loitk the Systems of Scientific 

 Units of Measurement. By T. H. Blakesley, M.A., 

 M.I.C.E* 



IT seems to me a not unfitting opportunity, as a sequel to 

 Prof. Riicker's suggestion, that we should modify some 

 of our scientific formulae, to inquire whether, with a change of 

 formulation, we may not advantageously effect a change in 

 some of the units themselves. That the C.G.S. and the prac- 

 tical system of units do not satisfy all the requirements of a 

 perfect system of mechanical units must, I think, have oc- 

 curred to others as well as to myself ; and I mention this, not 

 as an occasion for surprise, for indeed it would have been 

 surprising if any system brought into vogue twenty years ago 

 should still be found even to approach perfection, but rather 

 in order to draw attention to the points in which these systems 

 fail to give satisfaction, with a view to possibly rectifying 

 their shortcomings. 



The sense of imperfection may of course arise in the failure 

 to attain to any one of the various ends which it should be 

 the aim of a good system to reach. 



I imagine that the following would be generally considered 

 to be among those ends : — 



First. All the quantities considered should be so connected 

 by the equations representing the laws of nature, that no 

 coefficients are required in expressing any units in terms of 

 others. [Correlation.] 



Secondly. Where quantities are recognized as being essen- 

 tially of the same nature, they should be measured by the same 

 unit. [Simplicity.] 



Thirdly. Such a system should embrace all of the physical 

 ideas which occur. [Comprehensiveness.] 



Fourthly. Such a system should, so far as is compatible 

 with other requirements, have its units closely and decimally 

 connected with natural units. [Naturalness.] 



Fifthly. The units should, subject to other considerations, 

 agree with established though arbitrary units of measurement, 

 actually or decimally. [Convenience.] 



These five heads may be briefly described as: — 

 Correlation. 

 Simplicity. 

 Comprehensiveness. 

 Naturalness. 

 Convenience. 



I lay no particular stress upon the order in which these ends 

 * Communicated by tlie Physical Society : read December 8, 1888. 



