64 
SIR WILLIAM THOMSON" ON THE 
magnetizing force is less than a critical value of 280 or 290 ; and greater susceptibility 
in the unpulled than in the pulled wire when the magnetizing force exceeds the 
critical value; and from the fact that the difference in one direction of the suscep¬ 
tibilities in the pulled and the unpulled wire when the magnetizing force is the 
Glas gow vertical force, is about three times as much as the difference in the other 
direction when the magnetizing force is 80 times the Glasgow vertical force. The 
effect of the —M includes a reversal of the natural vertical force. That of the + M 
is merely an addition to it. 
§§ 214-222.- Preliminary investigation by the direct Magnetometric Method of the effects 
of transverse stress on the Magnetization of an Iron Tube. 
214. In order to test qualitatively, in the first instance, the effects of transverse 
stress on the magnetization of iron, experiments were made on a smooth gun-barrel, 
said to be made of tolerably soft iron. The barrel was fitted at its muzzle with a piston 
working watertightly in a Bramah stuffing-box, and served round with a magnetizing 
coil of silk-covered copper wire, separated from the barrel by a copper tube, and con¬ 
taining within it an induction coil, in order that the ballistic method might be used 
if this was deemed advisable. The barrel was then fixed rigidly to a stone pier in the 
laboratory, with its breech end resting on a large block of stone which formed the base 
of the pier. On a shelf, also attached to the pier, and at a convenient distance due 
magnetic west of the barrel’s axis, a small reflecting magnetometer was placed, with its 
needle on a level with the top of the barrel. A dead-beat galvanometer, the resistance of 
which was only a small fraction of an ohm, was used to measure the magnetizing current. 
The barrel having been filled with water was subjected to hydrostatic pressure, applied 
by means of a lever carrying a weight, the lever itself being counterpoised by a weight 
attached to a cord passing over a pulley above. The pressure (the friction of the piston 
in its collar being neglected) was measured by the amount of the applied weight and 
the multiplication of the lever. The effects of the application and removal of the 
pressure were measured by the deflections of the galvanometer needle, read on a scale of 
half millimetres, placed at a distance of one metre from the mirror. 
215. The first experiments after the apparatus had been got into working order 
brought out the remarkable result that the effects of transverse stress on the magneti¬ 
zation of iron are, as to quality, the opposite of those of longitudinal stress ; that is to 
say, when the magnetizing force is less than a certain critical value, the effect of 
applying transverse stress is to diminish, and of removing it to increase, the induced 
magnetization ; and when this critical value has been exceeded, the effect of the 
application of transverse stress is to increase, and of its removal to diminish the 
induced magnetization. 
o * 
The curves (1) of Diagram VII. show (after the manner of Diagram II. above) the 
effect for both fi-M and —M on the magnetometer needle, when placed on a level 
