AMPERE-TURN vs. GILBERT 267 



turns and the exciting current of windings. Thus, one would 

 have to deal with two units of magnetomotive force, the gilbert 

 and the ampere-turn, one being about 0.8 of the other. In the 

 second place, with the assumption p= 1 for non-magnetic materials 

 B becomes numerically equal to H, which is a grave inconvenience, 

 because B and H are different physical quantities. B and H 

 have different physical dimensions, because // has a definite 

 physical dimension, even though the numerical value of it is 

 assumed to be equal to unity for air. Therefore, to be sure 

 that proper physical dimensions are preserved, one has to remem- 

 ber where // is omitted in formulae, and for a physical interpretation 

 of results it is much more convenient to have it there, explicitly. 



Still another objection to using the gilbert and to putting 

 ft equal to unity for air is that the ratio of the ampere-turn 

 to the gilbert is equal to a quasi-scientific constant 4^/10. 

 To the author's knowledge, there is no simple, elementary way 

 of deducing the value of this constant, without going over 

 the whole mathematical theory of electricity and magnetism. 

 Thus, a constant is retained in practical formulae, the significance 

 of which remains a puzzle to the engineer all his life. It is true 

 that the value of //= 1.257 is equal to the same 4?r/10 after all; 

 but in this case there is nothing "absolute," mysterious, or 

 sacred about the value of 4^/10. The student is simply told 

 that 1.257 happens to be equal to 4^/10 because the value of 

 the ampere was unfortunately so selected. It is not necessary 

 to go into further details, because the historical reasons which 

 led to the selection of the values of unit pole and unit cimvnt 

 hardly hold at present. All calculations would be just as con- 

 venient if fi were equal to 2.257, or any other value, instead of 

 1.257. 



For these reasons the author un hesitatingly discards tin- 

 gilbert in teaching as well as in practice and uses the anip< n - 

 turn as the natural unit of magnetomotive forrr. Tin- value 

 of permeability becomes then an experimental quantity which 

 depends upon the units selected fur flux and length. 



