54 REPORTS ON THE STATE OF SCIENCE. 



the diurnal tilting to which it may be subjected, is unsuitable for certain 

 investigations. This, however, was pointed out by Sir George Darwin 

 in his Eeport to the British Association in 1882. The discussion suggests 

 precautions in the determination of the nadir at an observatory on the 

 sea-coast, and probably the deepest mine in central Britain is still 

 unsuitable as a place in which to measure the effects of lunar gravitation. 



The deflections accompanying tidal loads observed at Bidston indi- 

 cate a relationship between the yielding of areas represented by rocks and 

 other materials and loads which are fairly well measureable. 



These deflections which accompany a 10-foot tide amount at Bidston 

 to approximately 0""2. This yielding may be truly elastic, or it may 

 possibly be partly due to the sagging of a surface like that of a raft under 

 the influence of load. This latter idea falls in line with seismological 

 observations, which show day after day that the large waves of earth- 

 quakes, whether passing beneath the alluvial plains of Siberia or beneath 

 the crystalline rocks of North America, do so at a uniform speed. 

 Seismology suggests that we live on a congealed surface, which, whether 

 it is thick or thin, light or dense, apparently responds in a uniform 

 manner to undulations which pass beneath it. 



VII. Megaseismic Activity and Rest. 



From historical records it has been shown that there are reasons for 

 supposing that when there has been marked seismic activity in one 

 portion of the world there has been a corresponding activity in some 

 other part (see this.Beport, Section VIII., and also British Association 

 Eeport, 1909, pp. 56-58). Although the records on which this con- 

 clusion is based only refer to disturbances which have affected land areas 

 and seaboards, it suggests that periods of marked seismic activity are 

 governed by general conditions. We now possess a second register, 

 collected by stations which have co-operated with British Association 

 stations during the last eleven years, which refer to reliefs in seismic 

 strain in all portions of the globe. These I have divided into two classes. 

 First, those which have only been recorded in a single hemisphere ; and, 

 second, those which have been recorded in the whole world. To the 

 latter, which crossed an equator, I have given an intensity twice that of 

 those which only disturbed instruments in a hemisphere. Earthquakes 

 which have only been recorded throughout a single continent, no matter 

 how much damage they may have caused, have been omitted. When 

 these two classes are taken en bloc and arranged chronologically, it is 

 at once seen that they have occurred in groups, and to each of these 

 groups a value can be given dependent upon the number of shocks it 

 contains and their relative intensities. From centre to centre of each 

 group there are intervals, which usually vary between 10 and 30 days. 

 An interval of 20 days is common, but it rarely reaches 40 days. On 

 the accompanying diagram (see fig. 2) I have plotted the average inten- 

 nities or values of groups which have been followed by 8, 9, 10, to 34 

 days of rest. For example, groups with an average intensity of 4'5 were 

 followed by 10 days of rest, whilst groups with intensities of 5"4 have 



