1920 .] Jaggar.—New Zealand Volcano Research. 167 
a bath of ordinary automobile oil. Upper and lower suspensions are both 
of steel piano-wire. The writing-stile at the end of the boom is pivoted 
directly to the boom Avithout magnification levers. The tip rests on a 
drum-surface that moves only 1 mm. per hour, or about 1 in. per day. 
This motion need not be continuous, and so may be actuated by the time- 
clock contact, a magnet and ratchet taking effect once an hour. Three 
weeks or longer may easily be registered without resmoking. The free 
period of the instrument at the Hawaiian station is seventy-two seconds, 
and. the damping aperiodic. Tilt is measured from a fixed line of hour 
dots stamped on the sheet by an arm external to the drum-carriage. The 
registered line sIloavs the diurnal temperature tilt curve, and a longer period 
wandering in large measure volcanic. Two such instruments should be set 
up respectively in the meridian and at right angles to it. These instru¬ 
ments have proA 7 ed unexpectedly serviceable for strong local shocks, the 
absence of magnification keeping the apparatus from derangement, and the 
maximum amplitude appearing directly on the tracing. 
Tremor and local earthquakes require a magnification of 100 and a high¬ 
speed drum moving at least. 30 mm. per minute. This means a change 
of sheets every twelve hours. In Hawaii we have rebuilt the Bosch 
100 kg. instruments, damping them with rigid vanes in sets of four immersed 
in oil-tanks under the booms. The vanes lie in the oil edgewise to the 
directions of swing. The transmission from boom to magnifying-lever is 
a light T bar with polished steel pins at each end of the cross-bar ; these 
pins rest in watch-jewels sunk in sockets in the boom and lever-arms 
respectively. Guards of light spring brass prevent the T bar from being 
dislodged during a strongish shock. Quick local earthquakes with such 
apparatus register the P and S tremors very well when these are present, 
and the L waves are strong and smooth, without dotting. Time is marked 
every minute by electro-magnets which lift the stile-tip, the latter being of 
steel magnetized with polarity opposite to the lifting-magnet. A gap in 
the scratched line results, just as in the optical records. A pendulum 
period of from seven to ten seconds is quite long enough for local quakes. 
Volcanic tremor appears in all the sheets of the HaAvaiian record as fine 
undulations of the line, best seen with a magnifier. We have arranged 
the N.-S. and E.-W. pendulums so that both stiles register on a single 
drum, showing both records side by side on the same sheet, and greatly 
reducing the labour. 
Notes on our experiments in teleseismic registration will be found in 
recent numbers of the Bulletin of the Seism,ological Society of America 
The volcanologic instruments are chiefly thermometers, thermo-couples, 
cameras, transits, plane-tables, vacuum-tubes, steel pipes, and Seger cones, 
along with much apparatus devised as occasion demands, and so requiring 
shop tools. All of this material is the ordinary equipment of a physical 
laboratory, and each volcanic field requires experimental study in order 
to determine Avhat is needed. The bulletins issued by Professor Ornori 
in Japan and by the Hawaiian Volcano Observatory* give evidence of what 
experimental adventures the workers have met with in two different fields. 
Incidentally the ordinary meteorological records should be kept at a volcano 
station, in order to check the volcanic events against pressure., temperature, 
&c. Fairly accurate time-keeping is, of course, an essential, and if 
subsidiary stations are established, uniform time-keeping among the stations 
is imperative. 
* See references in Nature, 21st March, 1918. 
