109 
1921-22.] On Models of Ferromagnetic Induction. 
distances between the centres, namely, distances which made i equal to 
1‘21 and 1*29 respectively. The curves are symmetrical about the 180° 
line. It will be seen that they confirm the theory of rupture described 
above and illustrated in fig. 4. The change from one mode of rupture to 
the other was observed to occur when a differed from 180° by about 30° 
or 35°. The ratio of the field that would produce rupture at 180° to the 
field that would produce rupture in the most favourable position was found 
to range from about 2*35, when i was 1T8, to 2T, when i was 1'29. 
0 “ 20 ° 40 “ 60 ° 80 “ 100 “ 120 ° 140 ° 160 ° 180 ° 200 ° 
Fig. 8. — Experiments on the rupture of a pair of magnets, with various values oii. 
15. From the observed value of the field which produces rupture it is 
easy to calculate r the true half-length of the magnet. 
Applying equation (1) to a single pair of magnets we have 
which may be written 
( 10 ) 
where M is the moment of each magnet. The value of r which will satisfy 
this equation is readily found by trial. For example, in the above test a 
was 3‘965 cms., and when a = 90° the current producing rupture was 3’75 
amperes, making H^ = 2 20 c.g.s. If we try r = 3’36 cms. we have ^ = 1T8, 
