740 REPORT—1890. 
9. We see from the tables and curves that each curve has a minimum distance 
from the line of abscissas, and that each comes to an horizontal asymptote, parallel 
to the line of abscissas, for = oc. By looking at the formula we see that there is, 
in fact, an infinite succession of minimums and maximums in the expression for © ; 
but it is only the first minimum and following maximum that occur within the 
range of variation of ©, which we regard as sensible. In the case of 7=1 the 
formula gives 6=4-7 for the first minimum. The curves show for the cases of 
t=2, 3, 4, respectively the first minimum at 160 =4}, 3, and 2°6 respectively. 
us 
The thickness which corresponds to 6=7 is the half-wave length of the electric 
disturbance, which, as we have seen, is for copper 2:244 centimetres when the fre- 
quency of the alternations is 80 periods per second ; and for this case, therefore, the 
thicknesses that give minimum generation of heat in the first, second, third, and 
fourth layers are respectively 11-22, 6°31, 4:21, and 3°65 millimetres. Anything 
more of continuous copper than these thicknesses in any of the layers would. be 
not merely ineffective or comparatively ineffective, but would be positively anti- 
effective. Even with so small a thickness as 2°8 millimetres, for copper and 
frequency 80, line 2 of the table (corresponding to a sixteenth of the wave length) 
shows, in the first, second, third, and fourth layers, losses of 0:3 per cent., 2 per 
cent., 5 per cent., and 10 per cent. in excess of that due to the true ohmic 
resistance of the copper were it all effective. When the size chosen for the trans- 
former and the amount of output required of it are such that a thickness of 24 
millimetres in the direction perpendicular to the layers is insufficient, a remedy is 
to be had by using braided wire, or twisted strand, with slight insulation of varnish 
or whitewash, crushed or rolled into rectangular or square form of the desired 
thickness and breadth. A very slight resistance between the different wires thus 
crushed together would suffice to cause the current to run nearly enough full bore 
to do away with any sensible loss from the cause which forms the subject of this 
communication. 
5. The Molecular Theory of Induced Magnetism (with exhibition of a Model). 
By Professor J. A. Ewine, F.R.S. 
In applying Weber's theory of molecular magnets to explain the phenomena of 
induced magnetism, it is not necessary to assume that the molecules are subject to: 
any other directional constraint than is supplied by their mutual magnetic forces. 
This is demonstrated by means of a model consisting of a group of small perma- 
nent magnets, each free to turn about a fixed centre. The manner in which the: 
configuration of the group changes when an external magnetic field is imposed or 
varied in any way is shown, by means of the model, to correspond exactly with. 
the known character of the corresponding changes of induced magnetism in 
iron and other susceptible metals. Hysteresis, of which magnetic retentiveness is 
one manifestation, occurs in virtue of the movements of the molecules through con- 
ditions of instability ; these movements, being mechanically irreversible, involve dissi- 
pation of energy. Such movements occur when metals are subjected to cyclic 
strains, apart from the existence of magnetisation. The author has developed his. 
views in a paper communicated to the Royal Society (‘ Proceedings, June 19, 1890) 
and republished in the ‘Philosophical Magazine’ for September 1890. The con- 
siderations adduced there lead to the following conclusions :— 
1, That in considering the magnetisation of iron and other magnetic metals 
to be caused by the turning of permanent molecular magnets, we may look simply 
to the magnetic forces which the molecular magnets exert on one another as the 
cause of their directional stability. There is no need to suppose the existence of 
any quasi-elastic directing force or of any quasi-frictional resistance to rotation. 
2. That the intermolecular magnetic forces are sufficient to account for all 
the general characteristics of the process of magnetisation, including the variations 
of susceptibility which occur as the magnetising force is increased. 
3. That the intermolecular magnetic forces are equally competent to account 
