72 REPORT— 1902. 



all the elements which are required to calculate the heights of these waves. ^ 

 Now these heights are frequently as much as 1 or 2 feet, and apparently 

 represent accelerations j}^ of gravity. The magnitude of these quantities 

 is certainly sufficient to create a suspicion that the angular values assigned 

 to large waves has hitherto been exaggei'ated.^ 



3. The slight evidence of vertical displacements afforded by the 

 experiments described on p. 70. 



4. Dr. F. Omori's observation that the amplitude of seisraograms is 

 not dependent upon the sensibilities of the seismographs to tilting suggests 

 that the movements represented by large waves are horizontal rather than 

 undulatory. 



5. The smallness and paucity of records obtained from bifilar pen- 

 dulums. 



On the contrary, observations which support a surface undulation 

 hypothesis are the following : — 



1. Surface undulations exist in epifocal districts, and these by the 

 movement of water in ponds and lakes, the movements of the bubbles 

 of spirit levels, the apparent movement of stars in the fields of telescopes, 

 and by other phenomena, have been detected in districts' many hundreds 

 of miles beyond the epifocal area.'' 



2. The approximately constant velocity of propagation assigned to 

 large waves (see pp. 65 and 67). 



3. Observations which show that the magnitude of a seismograra 

 is dependent upon its sensibility to tilting, p. 70. This conclusion is 

 apparently contrary to that arrived at by ])r. Oniori. 



4. The indications of a vertical component of motion, which have been 

 recorded, p. 71. 



With these latter observations before us, it seems reasonable to 

 conclude that the large waves of earthquakes have an undulatory 

 character, but the tilting involved is not so great as generally supposed, 

 and in this sense the above quotation from ' Nature' requires correction. 



In the seismograms of a large earthquake we have the records of at 

 least two, and probably three, types of movement, and the manner in 

 which they are presented to us depends upon the character of the instrument 

 by which they were recorded. An ordiiaary long Tp^riod horizontal pendulum 

 shows the preliminary tremors, which are regarded as compressional waves, 

 which have passed through the earth to be recorded as ripples with a 

 small amplitude ; whilst the large waves, which are assumed to be very 

 flat undulations passing round the eartii in or beneath its ci'ust, are 

 shown as large displacements, which are magnified effects due to very 

 slight tilting. 



The same disturbance recorded by an apparatus, the natural period of 

 which is short, but which is provided with indices having a high multipli- 

 cation, gives records in which the preliminary tremors are large, whilst 

 the large waves are small, if not entirely absent.^ 



X. Melationship between Rockfolding, Seismic, and Volcanic Activities. 



Lyell remarks in his ' Principles of Geology,* vol. ii. p. 177 (12th 

 edition), that near the Bay of Naples there appears to be a connection 



' Brit. Assoc. Hep., 1898, p. 206. ^ Ihid., 1900, p. 83. 



» Ibid., 1898. p. 219 ; aliso 1900, p. 73. * Ibid., 189S, p. 263. 



