ON SEISMOLOGICAL INVESTIGATIONS. 



Table IX. 

 Suggested Anomalous Reflection of P. 



45 



and again the same arcs of 20°, 21°, &c., to arcs of 110°, 109°, &c., 

 as in Table IX. We start with 20° + 120°, which gives a combined arc 

 of 140° : succeeding cases give combined arcs of 130°, 120°, and 80°, 

 and let us look first at the last column. The time for the combined 

 arc of 80° runs up at first from 910^ to a maximum at 914^ ; then down 

 to a minimum at 905% and then pursues its original course upwards. 

 There must be a slight pause at the maximum and the minimum, though 

 our coarse tabulation to 1^ only and to 1° of arc does not put it in 

 evidence. These pauses make two anomalous reflections : but the 

 pauses being slight, the reflected waves are probably not noticeable on 

 the records. Look now at the first column, showing the results for 

 140°. The maximum and minimum have run together to make one 

 long pause at about 1248^ or 1249^ : hence we get a single anomalous 

 reflection, but much stronger; the two waves formerly separated com- 

 bining to reinforce one another. This combination is beginning to dis- 

 appear in favour of separation at 100°+20° = 120°, and the separation 

 is pronounced at 60°+20° = 80°. About 120° therefore this anomalous 

 reflection will die down : the precise distance at which it separates into 

 two clearly depends upon a precise adjustment of the tables, which is 

 scarcely yet attained. (The study of this anomalous reflection may 

 possibly give effective help in attaining that precision.) 



It is thus fairly easy to see why these reflected waves should be 

 mistaken for the direct P at distances greater than 110°._ Firstly, it 

 must be remembered that the direct P is becoming fainter as we 

 increase A beyond 110°; secondly, the two anomalous reflections begin 

 to coalesce and reinforce one another; and thirdly, it must be remem- 

 bered that an anomalous reflection of this kind has an advantage over 

 the direct P, and even over a regular reflection, in that it has two 

 alternative paths by which to travel, viz., arcs of 20° -I- 120° and of 

 120° -t- 20°: it may make either the short or the big jump first. For 

 regular reflection there are only the two equal jumps. 



As regards the actual times of transmission, it will be seen that they 

 accord fairly well at first with the observed times deduced for PX on 

 p. 48. 



