260 s 
NATURE 
[| Fuly 16, 1885 
For a long time es ae were installed beneath cases on a | stand, and so arranged that its stile could be viewed in the field 
column kindly lent to me by the Professor of Natural Philosophy 
in the Imperial College of Engineering. Here they remained 
for over a year, after which they made many journeys. At one 
time they were nearly 13,000 feet above sea-level on the top of the 
conical Fujisan. At another time they were some hundreds of 
feet below sea-level at the bottom of the Takashima mine. 
These observations attracted the attention of the authorities at 
the Imperial Observatory, who, recognising the bearing they 
might have upon work which was going on in the observatory, 
they supplemented my observations with a second set of levels. 
All the usual precautions were taken to guard against the effects 
of temperature, and observations were carried out every three hours 
both day and night for more than a year. As the books of 
records accumulated, and the curves grew until some of them 
were 30 to 40 feet in length, experience showed that the errors 
chiefly due to changes in temperature might be equal to and 
even exceed the effects which were being -sought. Now, I am 
inclined to an opinion communicated to me in a letter from 
M. d’Abbadie, who remarked that two levels upon the same 
column"might be parallel, and yet their bubbles ‘might move in 
a 
iy 
opposite directions. Notwithstanding this, the observations of 
levels have led to some interesting results. 
First, there is the fact that level bubbles may wander without 
there necessarily being a change of level. Second, that level 
bubbles continue to move long after the sensible motion of an earth- 
quake has ceased, thus giving us a means of observing the move- 
ments of long periods which usually bring the phenomenon of an 
earthquaketoaclose. Afterthe earthquake the bubble will some- 
times take up a position slightly different to that which it had 
before. Changes in the position of bubbles have been observed 
a short time before some of our earthquakes. Another result 
is the fact that the greatest irregularities in the curves showing 
the position of the bubbles of a level occur when earth- 
quakes are most numerous. This is during the winter months. 
A last result is the fact that during a typhoon, or when the 
barometer is unusually low and fluctuating, a level bubble may 
be distinctly seen to pulsate through a small range, as if there 
were continuous changes of level going on. 
While working with the levels another kind of instrument 
which I employed was a pendulum suspended from an iron 
of a microscope. By placing a prism beneath the end of the 
stile, the image of its end could be looked at horizontally, and 
the motion of the pendulum could be seen in any azimuth. At 
first I employed two microscopes placed at right angles, but by 
the adoption of the prism one microscope became sufficient. 
With these instruments, which are similar to those employed by 
Messrs. Bertelli, M. de Rossi, and other Italian observers, [ 
verified for myself some of the more important results which had 
been noted in Europe. 
For instance, it was seen that the pendulum was seldom at 
rest. Storms of tremors would take place with a low barometer. 
The pendulum did not always vibrate over the same point. It 
appeared as if there had been a tip in the soil, and the stand of 
the apparatus had been slightly inclined. These, together with 
other results which in many respects are little more than repeti- 
tions of results obtained by Bertelli, Rossi, and other observers, 
I have already published. Without attempting to describe 
other experiments which I have instituted, I will now give a 
brief description of an instrument which has been reached 
gradually, and which has given me the greatest satisfaction. 
From a letter received from M. d’Abbadie, whose researches 
regarding the change of vertical are amongst the most important 
yet instituted, I learn that my instrument has many points in 
common with one employed by M. Bouquet de la Grye. When 
I first set up this instrument, it was simply as a contrivance 
intended to make electrical contact, and set certain machines in 
action at the time of an earthquake. I next employed it as an 
instrument to record the occurrence of slight earthquakes. In 
its third form it was used to indicate earth-tremors and devia- 
tions in the vertical. It will be readily understood from the 
accompanying sketches, Figs. 1 and 2: AA is a circular disk of 
cast iron about } inches in thickness, resting on levelling screws. 
Bolted to this is a tripod of angle iron about 5 feet high, B B. 
This forms the support for a pendulum, c. ‘The bob of this 
pendulum weighs about 7 Ibs. It is made of a brass tube (3m. 
diameter and 24m. long) filled with lead. This is carried by a 
fine iron wire 3 feet 3} inches long, soldered into a small hole in 
a plate at the top of the tripod. A spike, C, projects from the 
base of the bob (see Fig. 3). As the bob with its spike were 
turned in a lathe, the end of the spike, the point of support and 
the centre of figure of the bob are fairly in a straight line. _ 
A long, light pointer, D, made of a strip of bamboo which has 
been varnished, is kept in contact with the base of the pointer, 
as shown in Fig. 3. At the top of the pointer there is a light 
brass ring, ¢ ; at the top of this there are two fine needle-points, 
aand 4. The point @ is kept in contact with the base of the 
pendulum by turning the screw T, which raises the flat spring s 
on which 4 rests. T is carried by a strong stand, E, which rests 
at three points on a ; /f is a disk of lead which is nearly equal 
in weight to that of the pointer below 4. : 
In one instrument a@ é is 6mm. whilst the total length of D is 
415mm. With these dimensions we may suppose that if the 
base pendulum moved, say, Imm., then the lower end of the 
pointer would move about 68mm. The values to be given to 
the defections observed in the pointer have also been estimated 
by giving a slight turn to one of the levelling screws of the base 
plate, and thus tipping the plate through a known angle. 
