Ill 
1921-22.] On Models of Ferromagnetic Induction. 
members of an infinite row, and rupture took place somewhere about the 
middle, with a field which, for every azimuth, was found to be veiy nearly 
twice as strong as that which produced the rupture of a single pair. The 
observed ratio was as exact as could be expected, having regard to the fact 
that there were unavoidable small differences among the eight magnets. 
0 50° 100“ 150“ 200° 
Fro. 9. — AAA, rupture of end members of a long row of magnets. BBB, rupture of a 
single pair of the same magnets with the same pitch. 
18. The same device was used to produce another condition, namelj^, 
practical endlessness in one direction only. Taking a row of five uniformly 
spaced and equal magnets, one end was left free, and at the other a block 
was placed and adjusted so as to maintain there a state equivalent to 
indefinite extension, while the deflecting field was increased until rupture 
took place, the break-up beginning at tlie free end and travelling towards 
the other. Fig. 9 shows what was observed when the row was tested to 
rupture under these conditions. The five magnets used in this experi- 
