350 Prof. Trouton and Mr. Andrews on the 
the name of “ British Pitch” are plotted in fig. 3, and — 
the two effects mentioned above. 
On first applying the torque there is a rapid flow or move- 
ment, which oradually diminishes and finally reaches a 
steady state. This steady state of rotation would apparently 
go on indefinitely but for the bending produced in the 
cylinder by its own weight, which finally introduces com- 
plications. Indeed, by judiciously keeping the cylinder hori- 
zontal by slight upward pressure occasionally applied, it is 
possible to twist the cylinder for hours together. 
Ifa white line be painted along the specimen before twisting, 
a beautiful spiral line of many twists is ultimately obtained. 
Probably with arrangements to make the experiments under 
a fluid of the same density as pitch so as to remove the 
vertical component of force on the pitch, the rotation could 
be kept going as long as one pleased. 
Elastic Viscous Recovery. —This second effect is shown at 
the end of the curve, where on removing the turning couple 
the cylinder of ee tums back some little distance, at first 
rapidly, then slowing down gradually to rest. 
Evidently, to do this, energ gy must have been stored in the 
substance in the form of elastic strain. The effect may 
perhaps be looked upon as complementary to the initial state 
when the rate of displacement is abnormally great in com- 
parison to the final or steady rate of rotation. In this initial 
stage a store of elastic energy is gradually accumulated, 
which is preserved intact during the state of steady rotation, 
and is given out on removal of the stress to produce the 
return flow. 
Different Torques applied to the same-sezed Cylinders.— 
Hxperiments were made to ascertain if the connexion between 
the torque and the rate of rotation was that deduced from 
the simple theory given above. These experiments show that 
though for practical purposes the theory is sufficient, yet in 
reality the phenomenon is more complicated than therein 
assumed. ‘The results are shown in fig. 4, where the final or 
steady angular velocity is plotted against the weight pro- 
ducing it. It will be seen that for small torques the rate of 
rotation is proportionally less than that for larger ones, but 
that finally above a certain value the curve develops into a 
straight line. For all values of the torque above this value, 
on subtracting a small constant quantity we get pro- 
portionality. This is probably partially due to the purely 
elastic strain, and will be better understood from later 
consideration in this paper. 
Determinations with different-sized Cylinders. —Determina- 
tions made with different-sized cylinders were in satisfactory 
