Geology and Natural History. 289 



mixed in nature as well as in duration. Such mixing is, however, 

 doubtless a phenomenon attending the passage of the waves 

 through the outer shell of the crust only, so that clear distinc- 

 tions between the first and second preliminary tremors and the 

 main waves are not lost. Plotting of the times of beginning of 

 the first and second preliminary tremors and of the principal part 

 indicate clearly that the first two are transmitted through the 

 earth, the latter around the earth. As the times of transmission 

 through the earth are dependent upon the changes in the nature 

 of the centrosphere dependent upon depth the preliminary tre- 

 mors are of great interest. Unfortunately at distances greater 

 than 100° for the first preliminary tremors and 125° for the 

 second preliminary tremors, the observations of the phases be- 

 come extremely doubtful. This further emphasizes the impor- 

 tance of installing instruments recording the vertical component 

 of motion, and instruments with high magnifying powers, not 

 less than 100. 



Following Wiechert's method, the curves representing the 

 noi-mals to the wave fronts and the velocities at various depths 

 are computed from the data of the seismograras. As shown in 

 fig. 28 of the report, the velocity of the longitudinal waves 

 increased rapidly with depth, but with decreasing rapidity, from 

 /7.2km pgj. ggg_ at the surface to 12-5'<^'" per sec. at 2170"^™ from the 

 surface, "66 of the radius from the center. Below that depth 

 the velocity is nearly constant. The velocity of the transverse 

 waves is 4'8'"° per sec. at the surface and increases almost lin- 

 early with depth, reaching a velocity of about '7*5'^'" per sec. at 

 half the distance to the center of the earth. The absence of 

 good records from distances beyond 126° prevents a knowledge 

 of the velocities at greater depths. Within the limits regarding 

 which information is given, Reid remarks that there is no indi- 

 cation of a sudden change in the velocity of either wave such 

 as we should expect if there were any sudden changes in the 

 nature of the earth's interior. 



In discussing the surface waves the most important point is to 

 determine the time when they begin. This is chosen by Reid 

 where the irregular movements due to the second preliminary 

 tremors give place to regular waves with a long period (30 to 50 

 seconds). The velocity of transmission is uniform along the 

 great arcs of the earth and the seismograms of the San Francisco 

 earthquake give 88 minutes as the time to reach the antipodes. 

 No variations in velocity along different paths are to be detected 

 greater than the errors of observation. Thus the velocity under 

 the Pacific ocean to Japan is the same as under North America 

 and the Atlantic ocean to Europe. 



On comparing these results with those of Milne in 1902 and 

 Oldham in 1900 it is seen that the preliminary tremors are for 

 the greater part of the curve about 2 minutes earlier, apparently 

 from lack of accuracy in the earlier observations. Determina- 

 tions of velocities given by different great earthquakes are shown 



