366 Messrs. Milne and Gray's Experiments 



The measurements on the rate of dying out were mostly 

 made by means of rolling-sphere seismographs. 



The velocities obtained for direct and transverse vibrations 

 were 438 feet and 357 feet per second. This gives 1 # 23 as 

 the ratio of the propagation of these two sets of vibrations. 



The very low rate of propagation is due to the soft medium 

 through which the vibrations were propagated. The actual 

 results are probably very nearly correct, from the manner in 

 which they were performed, there being no personal equation 

 to take account of. Similar experiments were tried at a point 

 where the medium was rock. As might have been expected, 

 in this case we obtained hardly an appreciable motion even 

 at a short distance from the ball. 



Actual velocities, measured in America at the time of the 

 Hell-Gate explosions, varied between 3500 and 8000 feet per 

 second. The rock was in this case gneiss (?) 



As the base-lines which were used in America were long, 

 they must have afforded an excellent opportunity for obtaining 

 a good separation of the normal and transverse vibrations. If 

 observations were made on these, we have not met with any 

 quotation of them. 



Mallet obtained velocities, in feet per second, as follows: — 



Sand 824-915 



Granite (shattered) . 1306-425 

 „ (solid). . . 1664-574 



"We presume that these velocities are all for direct waves. 



(c) From records drawn by seismographs on smoked glass 

 plates, it appears that the motion during an earthquake is 

 generally irregular in its character: the extent of movement 

 increases gradually to a maximum, and then varies back- 

 wards and forwards through several minima and maxima 

 during one earthquake. This gradual increase from small to 

 large amplitudes at the beginning of earthquakes has consider- 

 able value in connexion with deductions from falling columns 

 <fcc, with regard to the side from which the earthquake came. 

 The amplitude of movement in earthquakes felt in Yedo is 

 seldom above a few millimetres, even when they are sufficient 

 to do considerable damage to buildings. Usually the maxi- 

 mum amplitude is even under one millimetre. The period 

 usually varies between ^ and ^ of a second, having in the 

 majority of recent earthquakes been nearer to the latter value. 

 In several cases the direction of motion has been found to change 

 during the same earthquake, thus giving evidence either of 

 reflection or of transverse motion. Most of the points here 

 referred to will be found illustrated in the adjoining diagram. 



