THE SEISMOGRAM AND ITS ELONGATION. 113 
The waves approacht Taschkent and Jurjew making a small angle with the meridian. 
No difference can be made out for the two components during the first and second pre- 
liminary tremors at Taschkent, but at Jurjew the east-west component was larger for 
both. The waves approacht Mauritius exactly from the north; and it is the most 
distant station from the origin (162°). The earlier part of the motion was distinctly 
stronger on the north-south component; and this preponderancy lasted during the first 
part of the second preliminary tremors; but it must be remembered that the time of 
beginning of this phase is somewhat doubtful. On the other hand, we find the east-west 
motion, at Tacubaya, stronger for the first two phases, altho the direction of propaga- 
tion was practically symmetrical with respect to the two components. At Upsala the 
east-west movement was slightly more marked during the first preliminary tremors and 
the north-south during the second preliminary tremors, tho the opposite would have 
been expected. At Potsdam, Jena, and Gottingen the north-south movement was 
slightly the stronger during the second preliminary tremors, also contrary to expectation. 
This is the very meager evidence which the records of the earthquake offer regarding 
the direction of motion during the preliminary tremors. It is not entirely consistent, but 
it indicates on the whole that longitudinal vibrations were preponderant during the first 
preliminary tremors, and transverse during the second preliminary tremors; but that 
both kinds of motion existed practically during the whole of the preliminary tremors; 
and therefore the evidence can be said to favor the theory advanced to explain the draw- 
ing out of the record.' 
We must remember that transverse vibrations may have any direction around the 
direction of propagation, and in particular may lie in the vertical plane thru this 
direction; the horizontal projection of their motion would then lie in the direction of 
propagation of the disturbance along the surface, and they would be recorded as tho 
they were longitudinal waves. This may explain the longitudinal direction of the 
strong motion at Mount Hamilton, tho the movement on the fault-plane would lead 
us to expect transverse waves more nearly in a horizontal plane. 
THE PRINCIPAL PART AND THE TAIL. 
It is generally believed that the surface waves are also drawn out more and more as the 
distance of the station is greater; but an examination of the seismograms of the California 
earthquake does not support this view. It is very difficult to determine what should be 
considered the principal part and what the tail portion of the seismogram; but on mak- 
ing the best estimate we can of the principal part, we find no regularity in its duration; 
and we also find very different results according to the type of instrument recording. 
For instance, at Taschkent, one would estimate about 2.5 hours for the principal part 
from the Repsold-Zéllner instrument, and 15 minutes from the Bosch-Omori. At Balti- 
more (distant 35.7°) a Milne pendulum makes the duration of the principal part about 
47 minutes; whereas Bosch-Omori instruments at Washington (35.4°) and Cheltenham 
(35.6°) indicate a duration of only 6 to 8 minutes. At San Fernando (85.25°) a Milne 
pendulum gives a duration of 45 minutes; at Krakau (85.98°) a Bosch-Omori, subject to 
some solid friction, gives 5 minutes; and at Vienna (86.37°) a Wiechert inverted pen- 
dulum gives 13 minutes. The following table, in which the duration of the principal part 
is given in minutes, is made up from the records of Milne pendulums alone, and shows 
that even the same type of instrument does not yield consistent results. 


1 Prof. C. F. Marvin (Monthly Weather Review, 1907, vol. xxxv, p. 5) obtained very interesting 
results regarding the direction of vibration at Washington at the time of the Jamaican earthquake, 
January 14, 1907. The longitudinal vibrations began earlier and were much the stronger during the 
preliminary tremors; the transverse vibrations were much the stronger during the principal part. 
I 
