ON SEISMOLOGICAL INVESTIGATION, 233 
actual emergence, which distances will increase from P, towards P3. 
Directly d, emerges, not only will c, be eclipsed, but also cy c3, coming 
from the direction P, P3, will also be hidden. 
At some point like P;, when the duration of the preliminary tremors 
reaches a maximum on towards the origin, the quantity will decrease, if only 
on account of the fact that the velocity along the brachistrochronic ray 
differs less and less from that of the distortional wave within the crust. 
Such a view may possibly explain the rise and fall in the values of our 
last column. 
The growth in amplitude of the groups of tremors may be due to the 
fact that the first group has travelled on the path OP,, whilst the second: 
has travelled OP, P,, &c., whilst the crests of these groups, especially of 
those immediately in advance of the large waves, should roughly agree 
with the impulses which these represent. 
VII. On Certain Disturbances in the Records of Magnetometers and the 
Occurrence of Earthquakes. By Joun Miwe. 
In the ‘British Association Reports for 1898,’ pp. 226-251, a large 
number of records were brought together, showing what has happened at 
or about the time of large earthquakes to magnetic needles at various 
Observatories. These records may be classified as follows : 
1. Those which show that magnetographs have very frequently been 
disturbed at the time when their foundations have been moved by the 
large but unfelt waves of earthquakes originating at a great distance. 
Examples of such movements are to be found in the registers from 
Utrecht, Potsdam, and Wilhelmshaven. For the particular kind of earth 
‘movement referred to, magnetic instruments at these places furnish 
records of value to the seismologist. 
2. Those which show that magnetographs are seldom, and then only 
very slightly, or in some instances apparently never disturbed at the time 
of large earthquakes. This appears to be the case at Greenwich, Kew,, 
Falmouth, Stonyhurst, Pola, Vienna, Copenhagen, and Toronto. 
3. Those which show that magnetic needles have exhibited perturba- 
tions, frequently of considerable magnitude, a short time before the occur- 
rence of large earthquakes. As illustrative of such observations, reference 
may be made to the registers from Zikawei, Mauritius, Utrecht, and 
Greenwich. Similar observations have been made in Japan.! 
On pp. 248-251 of the above-mentioned report, an attempt is made to 
explain these observations, whilst to extend the same I append the 
following table received from P. Barrachi, Director of the Melbourne 
Observatory. 
Declinometer Disturbances observed at the Observatory, Melbourne. 
P. Barracut, Esq., Director. 
The magnetographs at Melbourne are of the same form and dimensions 
as those at Kew. The value of an ordinate of 1 inch in the curves is very 
nearly 29’, and the time scale corresponds to 14:7 inches for twenty- 
four hours. 
In dealing with the curves for Observatory purposes—as, for instance, 
taking mean values, &c.—oscillations whose amplitudes are less than 2’ 
? See Seismology, Int. Sci. Series, pp. 225, 226. 
