16 
Proceedings of Royal Society of Edinburgh. [sess. 
The Torsional Constants of Iron and Steel. 
By Dr W. Peddie. 
(Read January 22, 1900.) 
(Abstract.) 
This paper gave details of a series of experiments made on the 
same iron wire as was used in experiments described in previous 
papers. This series was made on the wire after it had been heated 
to redness and allowed to cool. A linear relation was again found 
to hold between log. b and n , where b and n are the quantities 
(constant in any one experiment) symbolised in the equation 
y n (x + a) = b, y being range of oscillation and x being number of 
oscillations of the wire which have taken place since the com- 
mencement of the experiment. It was further found that the 
line representing that relation passed (as did all other such lines 
previously obtained with this wire) through the point log. b — 2 '3, 
n— 1. Thus the quantity provisionally called the Oscillation Con- 
stant in the preceding paper, and regarded as characteristic of the 
material of the wire, retains its old value even after the wire has 
been heated to redness. 
The present paper contained also a description of a new 
apparatus now used for the investigation of the phenomena. 
It further contained an account of two series of experiments 
made upon a steel wire. In each series a linear relation held 
between log. b and n , and the lines representing the relations 
passed through a point log. & = 3T2, n — 1. Thus the oscillation 
constant for steel has a larger value than that for iron. 
The theory sketched in last paper was developed a little further, 
and it was shown how numerical measurements of the elasticity 
of metals may he obtained from the observations. The deviation 
from perfectness of elasticity is about six times as great in iron as 
in steel. The theory shows also that, in all wires of the same 
material and yitch ( i.e ., ratio of length to radius), the Oscillation 
Constant has the same value. This indication of theory has not 
yet been tested experimentally. It is shown also from theory 
that the Oscillation Constant has an explicit connection with the 
distortion at which the strongest molecular groups break down. 
