110 The N.Z. Journal of Science and Technology. [April 
be equipped with Milne-Shaw machines, and that the data gained should 
be sent to England, to be worked up there . I hope, however, that some of the 
work based upon seismological records may be done here (although the records 
should be regularly sent to England also). A seismogram can be properly 
interpreted only by one who understands its full scientific import, not by a 
mere routine collector of records. Moreover, the B.A.A.S. Committee has 
made mistakes in the past: e.g., correspondents were for many years invited 
to send details only of P 1? P 2 , P 3 (that is P, S, L), no notice being taken of 
PR 1? SR 15 &c. ; yet it may be that in the solution of the problems still 
before us these may contribute quite as much as P and S. I suggest that 
in the periodical bulletins the details of the published earthquakes should 
include every distinct phase : this would still allow the exclusion from the 
bulletins of inferior observations that would otherwise fill up space to no 
purpose. All the chief facts would then be available to workers all over 
the world, and their enhanced interest would still further increase the 
accuracy of the observers and the value of the whole work. 
A Seismograph Detector. 
Dr. Klotz, of Ottawa, has published a very useful diagram of the time- 
curves of the waves of the chief phases of an earthquake ( Publications of 
the Dominion Observatory , Ottawa , vol. 3, No. 2, 1916). The horizontal 
scale gives the time (apparently about 10 seconds to 1 millimetre) ; the 
vertical scale gives the value of A for every 60 kilometres—that is, the 
arcual distance from the epicentre to the observatory. So that if we know 
the distance of the origin we can find at once from the curves the time 
taken from the origin by the P waves, or the S waves, &c. ; vice versa , if we 
know the value of S-P we can find the corresponding value of A. If the 
time scale of the diagram is the same as that of the seismogram, then S-P 
will be of the same length in each; and if, also, the diagram is made 
transparent, then when we place it so that P and S of the seismogram are 
immediately below points on the P and S curves of the diagram all the 
other phases ought to correspond approximately, and the arcual distance A 
is that given at the left of the same horizontal line. (We could use the 
interval between any two phases, instead of S-P— e.g., PRi-P, SRj-S, &c.) 
The normal time scale of the Milne seismograph is about 240 millimetres to 
the hour. The Wellington instrument runs 246 millimetres to the hour 
(4'1 millimetres = 1 minute). Dr. C. E. Adams has redrawn the diagram to 
this time scale with A in degrees as well as in kilometres (Plate II). 
As has been stated above, the distance of the origin from the place of 
observation may be found, assuming the correctness of the tables for P and 
S, from S-P—that is, the interval between the arrival of the P and of the S 
waves. For instance, in our typical seismogram, S-P — either 711 seconds, 
or 669 seconds. Hence the distance of the origin from Christchurch was 
101°*2 or 92°‘2. A large number of reliable stations show it to have been in 
a region about 104 o, 25 from Christchurch. The discrepancy points either 
to an error in the observation or its interpretation, or to an error in the 
tables—probably the former in this case. It is easy to see that accurate 
returns from a large number of stations would enable us to correct the 
tables. 
