208 SCIENCE PROGRESS 



that we need for the present well-established body of physics 

 five kinds of fundamental quantity, the ones usually chosen 

 being those of length time, mass, permeability and tempera- 

 ture. The reason for the sufficiency and necessity of five 

 seems to be the fact that physics is at present considering five 

 fundamentally different kinds of " thing," viz. time, space, 

 matter, electricity and entropy — the latter being the " degree 

 of run-downness of a system." We can, in fact, arrange the 

 physical sciences in a hierarchy such that each successive 

 member introduces the consideration of one additional kind of 

 thing." Thus geometry introduces space ; kinematics intro- 

 duces, in addition, time ; mechanics, matter ; electrodynamics, 

 electricity ; thermodynamics, entropy. 



If we accept as a criterion for the choice of fundamental 

 quantities that they should have extensive magnitude, we will 

 have to reject temperature and permeability (or dielectric 

 inductivity) from the customary list. The desirability of this 

 criterion consists in the fact that we can then measure all our 

 fundamental kinds of quantity by a simple process of fitting 

 unit quantities together until we reach the magnitude of the 

 quantity we are measuring ; it would accord with the criterion 

 of simplicity also on the ground that we have to measure all 

 quantities having intensive magnitude by correlation with some 

 quantity having extensive magnitude. 



On these grounds Dr. Tolman then suggests that we should 

 employ as fundamental quantities, length, time interval, mass, 

 quantity of electric charge and entropy. No statements are 

 needed to justify the choice of the first three. As regards the 

 fourth, it is superior to the usual choice of permeability not 

 only because it has extensive magnitude, but also because of 

 greater psychological simplicity and a more direct relation to 

 the fundamental " thing," electricity. It is interesting to note 

 that experimental work seems to confirm the view that electric 

 charge is in the first group of quantities having extensive mag- 

 nitude — that is, it has discrete magnitude, and the most natural 

 way of measuring a charge would be based on the counting of 

 electrons. 



Entropy, although justifiable on the ground that it has 

 extensive magnitude, may not be justifiable on grounds of 

 simplicity, since temperature certainly seems to most people 

 the simpler idea. Entropy, however, does bear a simple rela- 



