188 Mr. G. Hookham on Permanent Magnet Circuits, 



proportional to the length of the bars. Hence, in order to 

 overcome a considerable air-space resistance, the actual dimen- 

 sions must be such that the ratio of the total cross-section of 

 the steel magnet-bars to their length must be very small com- 

 pared with the ratio of the cross-section of the air-space to 

 the distance between the pole-pieces. Obviously^ too, the 

 best proportions would depend on the nature of the steel 

 magnets — an entirely unknown quantity — so that the first 

 experiment had to be made almost at haphazard. Curiously 

 enough I have never been able to improve on the proportions 

 used in this first experiment. I have had some fairly accu- 

 rate tests made lately, and find that the best relation between 

 the two ratios, as stated above, is about 1 to 70. Thus, if A 

 equal the area of cross-section of the air-space, L the distance 

 between the pole-pieces, a the cross-section of the steel magnets, 



I the length of the bars, then ^ should equal about 70 x r. 



For example, if the cross-section of the steel magnets is 3 

 square inches, and their length 6 inches, then the air-space 

 may be 9 square inches, and the pole-pieces ^ inch apart, or 

 4^ square inches with ^ inch distance, or 2^ square inches 

 with jig inch, and so on, for in each case the relation remains 

 the same, viz. 1 : 72. 



The object of making such tests w^as to get the best possible 

 return for a given weight of steel. On the analogy of the 

 voltaic cell, this object would be attained when the internal 

 and external resistances were equal. With the proportion at 

 1 to 70 my tests showed that this equality practically existed ; 

 and, consequently, that the resistance of tungsten steel, as I 

 use it, is about one seventieth that of air, and therefore about 

 twenty times that of soft iron. The actual intensity obtained 

 is also very high, being in the ordinary pattern of my meter 

 nearly 4000 in the air-space, or at least half that in the air- 

 space of a modern dynamo, and represents a magnetizing 

 force of 1000 ampere-turns on a soft-iron core of full sectional 

 area. 



Some very simple, but practically important, results follow 

 from the assumed analogy to the voltaic cell. In indicating 

 them I neglect the resistance of the cast iron in the circuit, as 

 this can always be made negligible at a trifling extra cost. 

 Thus we can increase the intensity in an air-space of constant 

 area in the proportion of the square root of the weight of steel 

 magnets. We can, for instance, double the intensity by 

 doubling the length, and at the same time doubling the 

 cross-section of the magnet-bars, for then the magneto- 

 motive force is doubled, and all resistances remain the 

 same. Again, we may double the intensity in the air-space by 



