1911-12.] Transverse Induction Changes in Demagnetised Iron. 307 
made by each double zone towards the total transverse effect, when the 
angle 0 (fig. 1) = 22 0, 5, 45°,. and 67 0, 5 — continuous, dotted, and dash-line 
curves respectively. As each zone has been reduced by the cosine of its 
angular distance from the equator, belts with equal molecular surface 
distribution are being considered. 
© 
None of the diagrams represent the experimental results. But as they 
are successively summed up, so that broader and broader equatorial belts are 
taken, the results improve. The curves for the three angles of 6 change their 
relative positions and approximate to the experimental results in diagrams 
(8) and (9). The former deals with an equatorial belt between 40° north and 
south parallels, the latter with a belt between 50° north and south parallels. 
Between these positions the curves for the complementary angles 0 = 22 0, 5 
and 67°*5 will practically coincide, the maximum values of both falling 
considerably below the maximum of the curve for 0 = 45°, in very much 
the same way as in the experimental diagrams figs. 5 and 7. 
So far as the transverse effects are concerned, therefore, the experi- 
mental results may be co-ordinated with a broad equatorial belt of equally 
spaced molecules in reference to the demagnetising (polar) force. 
But on the assumptions which have been made this conclusion does 
not allow a sufficiently large saturation value for the induction co- 
directional with the subsequent magnetising force H c . The number of 
molecules in the equatorial belt between 10° north and south parallels 
having been taken as 12, their number in a broad belt, say between 45° 
north and south parallels, may be considered as approximately 98, i.e. 
proportional to area. This number will also represent the saturation 
value of the co-directional induction (B c ) when all the molecules are in 
line with H c (98 x cos 0°). From fig. 8 (8) and (9) the transverse component 
in terms of the summation of sines may be taken as 5‘75 when 0 = 45°. The 
ratio of these two theoretical effects is thus 17, while the same ratio as 
determined experimentally is B c /B h = 27, the saturation value of the co- 
directional induction being B c = 15,000 and the transverse induction being 
B l = 560 when d = 45°. 
The reason of this discrepancy is obvious. In the absence of any 
known law connecting the distribution of the molecules in demagnetised 
iron with latitude, and as a first trial only, zones between 0° and 50° north 
and south latitude have been considered to the exclusion of zones between 
higher parallels. If these latter had been taken into account, and as their 
transverse effect is opposite to that for zones at lower latitudes, as can be at 
once seen from the curves of fig. 8 (1) to (5), or even by inspection of 
diagram fig. 1 , it is evident that there must be a closer distribution of 
