206 The East Anglian Earthquake.
phenomenon to a vorticose movement of the ground. In view
of the complicated character of earthquake-motion as revealed
by seismograph tracings, it seems not improbable that the
rotation may in some cases be caused by such a twisting
motion. Of the causes tending to produce vorticose motion
may be mentioned the interference of direct and reflected
waves moving transversely, the interference of the normal
and transverse waves, &c.; but whatever cause may be the
true one, it would be quite impossible to determine in such
cases the direction of propagation from the direction of
rotation of the chimney. It may here be pointed out that
Mr. Kinahan's remark (p. 66, note) with reference to the
chimney-twists at Langenhoe is fully in accordance with
the conclusions at which we had arrived on other grounds
(p. 202), vit., that this place suffered by the interference
of direct and reflected waves. At Wivenhoe, where the
general direction appears to have been much confused, the
chimney-twists may possibly have been also due to a vorticose
movement.
But although a twisting motion of the ground may in
some cases be admitted, it has been shown by Mallet (Proc.
Roy. Irish Acad., vol. xxxi., part 1, 1846) that a simple
rectilinear movement would be quite competent to produce
the effects observed. Were this straight line motion the
cause of the twists observed by us, it might have been
expected, as pointed out by Milne,108 that the rotation in the
same locality would have been very frequently in opposite
directions, as it is certain that the centre of friction could not
be expected to bear the same relation to the centre of gravity
in all chimneys. As we have already stated, however (p. 65,
note 30), the effect of this opposite rotation was only occasion-
ally observed during our visit of inspection, the majority of
the chimneys in each district having been twisted in the
same direction. It is probable, therefore, that a simple
rectilinear movement played but a subordinate part in
producing the twists observed after the present earthquake.
The best dynamical treatment of rotated columns, &c, that
108 Trans. Seism. Soc. Japan, vol. i., part 2, p. 33.