1897.] on Becent Advances in Seismology. 327 



1. Bradyseismic or slow secular changes, resulting in tlie elevation 

 or depression of countries and mountain ranges. 



2. Secular flow or crush. Of this we have only indirect evidence. 



3. Annual or longer period changes in level. 



4. Earthquakes or accelerations in bradyseismic action or secular 

 flow. Volcanic earthquakes. Sea waves. Air waves. 



5. Unfelt earthquakes, common to all countries. 



6. Irregular changes in level completed in a few minutes, or in 

 many days. 



7. Diurnal waves. 



8. Tremors, or microseisms and pulsations. Possibly in part 

 atmospheric movements. 



The advances that have been made during recent years by 

 recording movements which may possibly have a bradyseismical 

 character are, as compared with the information derived from the 

 study of the other movements with which we have to deal, but few 

 in number. Both in Germany and in Japan, horizontal pendulums 

 have been carefully installed underground, and it has been found 

 that in both instances, as with the levels of Plantamour, although 

 there is an annual change in inclination which cannot be accounted 

 for by seasonal changes in temperature, there is for periods of several 

 years' duration a continuous tilting in one direction. 



A very curious observation made in Tokio, was, that very often 

 for several days before a local earthquake, a horizontal pendulum 

 would gradually wander towards the west. Although such a sequence 

 in phenomena may have been accidental, because it has been shown 

 by observation with seismographs founded on the solid rock that the 

 greatest and most frequent motion is in the direction of the 

 dip rather than parallel to the strike, indicating that the direction of 

 folding is a direction of pronounced yielding, whilst slow change in 

 level is apparently most pronounced in districts where mountain 

 growth is possibly yet in progress, we see in the Japan observations 

 an indication of the possibility that crises in bradyseismical motion 

 may be foretold. 



I learn from Col. J. Farquharson, H.E., Director of the Ordnance 

 Survey, that some years ago the question whether during recent years 

 there had been any changes in level in Britain was carefully tested 

 in Lancashire and Yorkshire, under the direction of Sir Charles 

 Wilson. The first levelling in these counties was carried out between 

 1843 and 1850, and the second between 1888 and 1894. Excepting 

 in the coal and salt districts, no material changes were found to have 

 taken place. It is, however, to be remembered that this re-levelling 

 was confined to lines of level along roads, and whether there have or 

 have not been any changes in the height of hills or mountains since 

 the first measurements were made we do not at present know. 



One method of measuring bradyseismical eftects within a period 

 of three or four years, and to determine how far such movements may 

 be connected with the occurrence of earthquakes, would be to estab- 



z 2 



