Fuly 16, 1885] 
NATURE 
” > 
+. 
a. 261 
Records were at first made by reading a scale of millimetres 
placed beneath the end of the pointer. Experience showed this 
method to be inconvenient and without satisfaction. What 
occurred between the hours of observation was unknown, whilst 
the records which were made were liable to greater or less 
errors due to the observer. This led me to seek for some 
method which would render the observations automatic. To 
attach a hair to the end of the pointer and let it be dragged 
across the surface of a smoked glass created too great friction. 
‘The necessary appliances for photographic registration were too 
costly and too troublesome to be employed as I was situated in 
Japan. A very near approximation to frictionless registration 
was obtained by sending a current down the pointer, the end of 
which trailed on the surface of a thick film of iodised starch 
covering a strip of paper. The strip of paper, which was on a 
metal tray, moved slowly by clockwork beneath the lower end of 
the pointer. On taking out the paper I found that the film of 
starch, with its blue line, could be dried down to form a brown 
line on the paper. The process was troublesome and the line 
subject to distortion by the flow of the starch. The next idea 
was to discharge a spark from the end of the pointer and per- 
forate a band of paper moving beneath the end of the pointer at 
the distance of about 5mm. This feature in the apparatus, M. 
d’ Abbadie writes me, is an essential feature in the apparatus of 
M. Bouquet dela Grye. To avoid losing a record, should the 
pointer move parallel to the length of the paper, two bands of 
paper, gg, moving at right angles (by means of the clock, H), 
are employed (Fig. 2). One band passes beneath the other over 
the surface of a brass plate, 4. The paper used is the ordinary 
paper employed in a Morse telegraph instrument. By allowing 
the hand of a clock to pass every five minutes across a wire the 
current from two of Thomson’s tray-cells is sent through an in- 
direction coil which yields the sparks to perforate the paper. 
Every hour the hand of the clock makes a long contact by 
passing across a small strip of platinum. In this way a large 
hole is made in the moving bands of paper and the hours are 
recorded. 
To secure myself against error the same secondary current 
which perforates the paper of one machine is carried by wires to 
perforate the paper of a second instrument of slightly different 
construction placed on a stone column in a distant room. 
The only work required is to wind the clocks which pull the 
paper and the clock which makes the contacts. This being 
done, records are automatically made every five minutes. 
Up to the present the records which have been obtained have 
not been analysed, but certain of the results which they indicate 
are evident. These are as follows :— 
I. Sometimes for days the pointers remain stationary, as is 
eas by the sparks being regular and in a straight line (see 
ig. 4). 
2. Sometimes the pointers are in a state of tremor, and the 
sparks perforate the paper at many points, giving a line of 
several millimetres in breadth (see Fig. 5). These tremors may 
continue for ten or twelve hours. From the diagram, the dura- 
tion of these tremors and the range of motion can be accurately 
measured. The instruments in both rooms agree as to the 
occurrence of tremors and periods of rest. 
_ 3. Sometimes the pointer will slowly wander from the straight 
line, and then slowly return. This usually takes place two or 
three times in succession. It would seem as if the ground had 
been slowly tipped through one or two seconds of arc, the period 
of each tip being from fifteen to sixty minutes (Fig. 6). 
In regard to the occurrence of these tips, the instruments in 
the two rooms only occasionally coincide. 
As to whether they are really to be regarded as true disturb- 
ances of level, or simply as movements due to local causes, I 
shall be better able to speak after a more carefal examination of 
the records. 
4. Sometimes I find the bands of paper perforated over their 
whole breadth by sparks in all directions. This indicates that 
an earthquake has occurred and the pointer has been swinging 
(see Fig. 5, about 9.15 p.m.). All these figures have been pro- 
duced by pricking through from the original diagrams. The 
clocks which I have used are made from small American spring 
clocks costing in Japan about 12s. each. The total cost of the 
portion of the apparatus figured, including the case, M, the doors 
of which and the parts which come in contact with the column 
are edged with flannel, is about 25 yen, or £4 10. In Europe 
an instrument of better construction would cost more. One of 
the columns on which an instrument is placed measures 6 feet by 
3 feet and 5 feet high. It is constructed of brick and rests on 
concrete, The other column, which also rests on concrete, is 
made of stone. It measures 2 feet 2 inches X 2 feet 2 inches 
and is also 5 feet in height. 
This latter column is rather too slight, as I found that even 
the pressure of my thumb is sufficient to cause the pointer of the 
instrument to move several millimetres. 
Amongst those who may possibly have a practical interest in 
this matter are those who have to deal with mines—especially, 
perhaps, coal-mines. 
In the columns of the Hafan Gazette, in NATURE, in the 
Mining Fournal, and other papers, references have been made 
to the attempt to observe earth-tremors and other phenomena in 
the Takashima Colliery near Nagasaki. At the conclusion of a 
4 Dk G3 
February 24th. 2-4. A.M. (No movement ) 
Fig. 4 
Ae Se CSR a a  N 
a 7 8 
March 28th. 1885. 6&7. P.M. (An earth tip ?) 
Fig. 6, 
March 16th. 1885. 8-10. P.M. (Tremors & Earthquake) 
Fig. 5. 
report to the British Association, 1884, on the earthquakes of 
Japan, a letter from Mr. John Stoddart, the chief engineer at 
that mine, tells us that, owing to the working of the mine and 
other causes, he finds it impossible to make observations with 
delicate instruments. He therefore proposes to move the in- 
struments to som? distant station, assuming that any natural 
cause which would cause tremors in the mine will be generally 
felt over a considerable area. As to whether there is a connec- 
tion between earth-tremors and the escape of gas in collieries, 
we do not yet know. Mr. Walter Browne, in a paper to the 
North of England Institute of Mining and Mechanical Engineers, 
thinks it desirable that investigation on this subject ought to be 
made, and quotes what is being done at Takashima. Mr. Gallo- 
way, writing in NATURE of February 5, if I read him correc:ly, 
does not encourage Mr. Brown’s suggestions, and enters into an 
argument about the possibility of an earth-tremor forming a 
fissure, Earthquakes often form fissures on the surface, but these 
effects in mines are usually nothing. I make this statement on 
the authority of personal inquiry in many mining districts. 
With the exception of the disturbances near an epicentrum, the 
movements due to ordinary earthquakes are so superficial that the 
range of motion at the depth of 10 feet is sometimes only one- 
fortieth of what it is at the surface. Earth-tremors are pheno- 
mena usually lasting many hours, and they certainly occur with 
low barometers. That they could by any possibility form fissures 
it is difficult to imagine. 
