ON SEISMOLOGICAL INVESTIGATIONS. 7 
to the ratio of the leverage about the mass acting as a steady-point, and 
magnifies and records the horizontal ground movements ; this sensitivity 
is termed its ‘ magnification.’ The constants of one of the instruments 
were then altered so that, while the magnification remained the same, the 
tilt sensitivity was raised very considerably. But alterations in sensi- 
tivity to tilt hadno effect upon the recorded amplitude of the microseisms, 
suggesting that microseisms are purely compressional waves. 
The constants used were :— 
Mons. Magel50) 5° vy. °. « Tilt I seo, = 110 mm. 
Wo: Ose~,,, 150) .. : z Se is = 26 mm. 
On May 4 a small earthquake shock was recorded at a moment when 
the constants of the machines were alike, and it was gratifying to find 
that two damped machines when properly standardised may be relied 
upon to give similar results. The Plate shows comparable sections of this 
record from each machine ; also part of the record taken at Shide (126 
miles distant), which shows the same characteristics as the other two— 
note the isolated movement C. (Up the film at West Bromwich compares 
with down the film at Shide, and the paper speed is slower.) The letters 
A, B, C, &c., identify corresponding movements. The discrepancy at 
G was due to a fault in the driving motor, which has since been remedied, 
otherwise the West Bromwich records are identical. 
On May 15 the maximum of another small shock was recorded when 
the periods were 40 seconds and 12 seconds respectively, and the damping 
7:14. 14:1. The longer period gave from three to seven times the 
amplitude according to the impressed earth period. As the dampings were 
unequal, the result was not strictly comparable ; but it was noted that the 
longer period, notwithstanding the advantage of less damping, was much 
slower to take up the earth wave. The long period pendulum showed a 
lag of from 5 to 9 seconds behind the other, according to the impressed 
wave period being short or long respectively. 
This points to the desirability of machines conforming to some standard 
period if the times from different records are to be strictly comparable. 
A further important observation was the fact—previously referred to in 
these Reports—that two sites comparatively near together may be quite 
different as regards daily wandering of the zero. 
In the Plate the difference in the spacing of the lines on the two 
machines shows this clearly. The one with the wider spaces was taken 
in the out-building, and the displacement corresponded to an elongation 
of the sunny side during the day and a contraction at night, and was 
greatest on hot days. Time did not permit of sufficient investigation 
to discover whether the pier, cast-iron column, or whole house were 
affected. If only one or both of the former, then, since the heat rays 
are from the infra red end of the spectrum, protection may be afforded 
by interposing some athermanous substance, such as glass or water. 
In the 1915 Report attention was drawn to two machines in the same 
azimuth, at Shide, behaving quite differently as regards wandering. The 
Milne-Burgess was not only less affected, but also showed a time lag 
behind the Milne-Shaw. The Milne-Burgess machine was fitted with a 
large glass cover. 1t seems probable that the athermancy of glass to dark 
heat may have been the cause of the observed effect. 
If the iron column only is affected, then a great advance might be 
secured by making the column of substances with very low co-efficients of 
expansion, such as silica or invar steel. 
