222 
PROFESSOR J. A. EWING AND MR. W. LOW ON THE 
must have formed (in such cases) a very approximately accurate measure of the force 
acting on the metal. In other cases, when the cones were more blunt, the force in 
the external field was somewhat greater than the mean force within the metal. 
§ 3. Figs. 1 and 2, copied from our earlier paper, show the forms of bobbin originally 
used, and the pole-pieces of the magnet by which they were magnetised. With 
bobbins of the type of fig. 1, the magnetic induction in the neck and the field in the 
surrounding air were measured by suddenly turning the bobbin round, end for end ; 
in bobbins of the type of fig. 2, the measurements were made by suddenly withdrawmg 
the bobbin from its place between the pole-pieces. In the latter case, the induction 
measured was the excess of the whole induction (53) above the residual induction (iS,) 
which persisted when the bobbin was drawn out. In iron bobbins the residual 
magnetisation was found to be sensibly constant from the lowest to the highest value 
of the inducing field employed in these experiments, but the form of the bobbin inade 
the amount of this residue small. It was measured in bobbins of the type of fig. 1, 
by comparing the result of turning the bobbin round with the effect of drawing the 
bobbin out; and, in the first instance, its value in iron bobbins of the type of fig. 2, 
was estimated to be about the same as in bobbins of the type of fig. 1. In later 
experiments, when other more retentive metals were being examined, and the residual 
magnetism consequently formed a more important part of the whole, its value was 
directly determined by using built-ujD bobbins which allowed one conical end to be 
withdrawn; the residual magnetism was then determined after the bobbin had been 
removed from the field by slipping oft’ (in one operation) one of the conical ends, along 
with an induction coil which had been wound for this purpose upon a loose ring over 
the central neck. 
Wrought Iron. 
§ 4. In the early experiments solid bobbins of the form and dimensions shown in 
fig. 1, were tested, one of Lowmoor, and another of Swedish wrought iron, vfith 
