126 OLIVER AND DORMAN [CHAP. 8 



indicating that the reason for the absence of these waves in the case of earth- 

 quakes is not an unknown structural feature, but rather the location of the 

 source. 



5. Evidence from Short-Period Surface Waves 



It is to be ex])ected that oceanic surface-wave trains include waves with 

 periods shorter than those of the crustal waves discussed above, which were 

 explained entirely as the fundamental modes of the Love and Rayleigh types. 

 In addition to the i)ortions of the fundamental-mode dispersion curves which 

 continue to high frequencies, it was noted previously that propagation in an 

 infinity of higher modes of both wave types is theoretically possible. A practical 

 limit to the frequency of coherent signals which can be projiagated through the 



it;^\yim'J!^i'*^^yf*^:*iimfi py A-.^:! cQ A = 3600km Ki^^^'^i'^^y^ 

 >^Wx..Aw;,/*vx;,v^^^A^f'W-. 190350 52 /^N Ibyw ^^,^,wawk^^w,v^; 



Fig. lU. Honolulu seisniograius of a shock in the Aleutians. Because of the direction, 

 almost directly north, from which the waves approached, the north-south component 

 gives a record of longitudinal motion while the east-west component gives the trans- 

 verse motion. Fundamental -mode Rayleigh waves are clearly visible on the vertical 

 and longitudinal instruments, whereas long-period Love waves are absent on the 

 transverse component. Short-period waves, however, are present on the transverse. 



sub-bottom and to the distance at which they can be detected is imposed by 

 attenuation. On the other hand, theory shows that propagation of short-period 

 waves is confined to shallow (le])ths and, therefore, crustal structure must be 

 more important in determining the properties of short-period waves. 



A train of short-period (G to 9 sec) waves, which is often fovnid on seismograms 

 for oceanic paths, was described by Coulomb (1952) and by Oliver, Ewing and 

 Press (1955). Exam})les of this wave train are shown on the Honolulu seismo- 

 grams of the shocks of 27 April, 1958, 23 September, 1958, and 13 July, 1958, 

 which all occurred in the Aleutian-Gulf of Alaska region (see Figs. 10, 11 and 



