94 nuPORT=-1858: 
so as to support the balls B and B, upon their upper ends, which are slightly 
hollowed to the same curve as the surface of the balls, as seen at full size in 
fig.7. The balls, when in this position, rest against and are steadied by the 
hollow stop over the axis of the vertical pillar, 6 in figs. 4, 5, and 6. 
The balls may be common cast-iron cannon shot, chosen of good sphe- 
rical form and of equal weight ; and each ball is in metallic connexion at one 
point of its surface with a galvanic-circuit wire, of which it forms one pole, 
marked e¢,—the supports s, s, and the stop b, being all of hard wood or other 
insulating material, as pottery or glass. The height of the central column 
should be such, that the centre of gravity of each of the two balls, when on 
their supports, may be some submultiple of 32 ft.=g (say 8 feet =i) for 
facility of calculation. 
The shallow basin ¢¢ is subdivided in two semi-circular separate areas, by 
a wood division, d, equal in depth to the outer rim, this division crossing in 
the diameter which lies at right angles to the plane of the supports s, s,—7. @. 
being east and west for the north and south balls, and vice versd in the other 
instrument. Each segment of the shallow basin is lined within its outer rim 
and bottom with sheet-lead, which is at one point of each in metallic con- 
tact with the other pole of the galvanic circuit marked E,—. 
The two segments of the dish are filled up to the level of the surround- 
ing rim, with a bed of damp sand, pressed uniformly and “struck off” level 
to the rim by a straight edge, so as thus to present a uniform bed 9 inches 
deep, the balls B, B, being 6 inches in diameter and 8 feet above it. While 
the instruments (i. e. that N.S. and E.W.) are thus prepared, the galvanic 
circuit remains constantly broken, the poles formed by the balls being in- 
sulated from the other poles formed by the sand-beds, the lead lining, &e. 
Suppose now, in fig. 4, an earthquake-wave to emerge from S. to N. in the 
direction of the arrow; the ball B, is deft behind as in the former instrument, 
topples off its slender support s, and commences to fall to the surface of the 
sand. The moment it strikes the sand, it makes contact with its own circuit, 
and as the time of its fall can be exactly calculated and is constant (neglect- 
ing the small resistance of the air), this ball (as before) marks the precise mo- 
ment of the arrival of the shock at the instrument. The other ball B is 
urged forward by the movement of the whole instrument in the direction of 
the arrow, or that of the wave’s emergence, being supported by s and 8, until 
the instrument acquires its maximum velocity vas before. This ball is then 
thrown off from its support with this velocity, and, describing a small trajec- 
tory in air, falls to the bed of sand, and inits turn makes contact with its own 
galvanic circuit. The ball partially buries itself in the damp sand at the 
spot it falls upon, without change of position from any elastic effort, all such 
being absorbed by the “deadness” of the sand. If the shock has been in 
the plane of the meridian, the place where it shall land on the sand-bed will 
also be in that plane, say at BY. 
Then the horizontal distance from the centre of its support s to the centre 
of the ball, measures the horizontal component of the velocity, this space 
being described by it during the time of its descent through eight feet. The 
difference in time (as shown upon the ruled paper by the pencil-tracers and 
clockwork as before) between the instant of B, and of B leaving their sup- 
ports, is almost exactly = > or half the time of the wave. 
The same explanations will apply to the other, or E. and W. instrument ; 
and if the azimuth of emergence 6 be somewhere between N.S. and E.W., 
all four balls will be displaced, and the obliquity of throw of each of the balls 
il Og 
bisa ee ee a 
