ON SEISMOLOGICAL INVESTIGATION, 



67 



and though we could formulate a law which would make through-earth 

 waves reach the different points of the surface in times proportional to 

 the arcs, yet it would be a most complicated and improbable law. There 

 is no doubt the surface run fits in admirably. I always had great diffi- 

 culties about it, but " facts are chiels that winna ding," and the most 

 obvious interpretation of your curve is your old view of surface waves. I 

 never could bring myself to believe in the transition of such small oscilla- 

 tions through the heterogeneous crust. It is conceivable, however, that 

 what we observe may be the outcrop of waves running over the surface 

 of the inner more homogeneous nucleus. This may not be necessary. 

 Meanwhile I am compelled to withdraw my antagonism to the surface 

 wave. 



' The preliminary tremor curve fits to within the errors of observation 

 the formula time oc chord. 



' The chord is proportional to sine of half the angle. Tabulating we 

 get:— 



' The fourth column gives the same comparison for the second curve, 

 what you call the second phase. In calculating the ratios of the times 

 I take 23 minutes to be the time to the antipodes in the case of the first 

 phase and 33"5 minutes for the second phase. 



' You will see how admirable the agreement is for the first phase. 

 But it is not so good for the second phase. The first phase runs at 

 practically the same rate along the chords. The second phase seems to 

 run a little more quickly the deeper they go.' 



Tables of apparent velocities based on observations prior to the end of 

 1896 are given in the 'British Association Report' for 1897, p. 173. 

 These are discussed in the Report for 1898, p. 221, where a table shows 

 that the apparent arcual velocity for preliminary tremors varies with the 

 square root of the average depth of the chord. In the report for the 

 following year, p. 231, the preliminary tremors are referred to as passing 

 through the earth with an average velocity which increases with the 

 square root of the average depth of the chord along which they are 

 assumed to travel. Whilst pointing out the inaccuracy of this reference, 

 it must"^ also be observed that the variable velocity it implies is not 

 sustained by observations here published. 



f2 



