February. 1911. 



KNOWLEDGE. 



55 



square niik-s. In the four succeeding years, the num- 

 bers c)bser\'ed were nineteen, thirteen, seventeen and 

 eighteen, and they show no sign at present of abating 

 in frequenc\' or strength. During the ten years 

 1900-1909 the number felt was eighty-three. They 

 are interesting from their probable connection with 

 the great fault \\'hich skirts the southern side of the 

 Ochil Hills, which, notwithstanding its advanced age, 

 still seems to be growing. The small movements, 

 which give rise to the earthquakes, appear to be 

 limited to the portion of the fault between Airthrey 

 (near Bridge of Allan) and Tillicoultr\-, and to take 

 place at a very small depth below the surface. 



Observations in Mixes. 



Isolated observations have been made in mines in 

 man\- cases, both in this and other countries, the 

 general result being that the shock is more feebh' 

 felt in mines than upon the surface of the ground. 

 Several recent I^ritish earthquakes have occurred 

 within or near mining districts, and these, especialh' 

 the Derby earthquake of 1903 and the Swansea 

 earthquake of 1906. have thrown further light on 

 an interesting subject. The\' confirm the pre\ious 

 impression that the shock is more manifest on the 

 surface than underground : indeed, the disturbed 

 area in mines is only a small fraction of that upon 

 the surface. On the surface, again, the sound in 

 strong earthquakes is heard over an area much 

 smaller than that within which the shock is felt. 

 In mines the sound is observed as far as, or farther 

 than, the shock. Some interesting observations were 

 made on the sound in mines in the two earthquakes 

 referred to. Thev show that the sound appears to 

 travel through the rock overhead rather than through 

 that below, but this is onh' the case at distances of 

 more than five miles from the centre. Again, in 

 both earthquakes, there is some, though not decisive, 

 evidence showing that the intensity of the shock 

 increases with the depth of the workings below the 

 surface. This point is one on which it would be 

 worth \\hile to make careful in\-estigations in future 

 earthquakes. 



Origin of British Earthqu.vkes. 



The most interesting result to which the study of 

 British earthquakes has led is the proof of their 

 connection with the slow growth of faults. The 

 longer axes of the isoseismal lines are parallel, or 

 \-ery nearly so, to the main faults of the central 

 district, or, if the faults be unmapped, to the prin- 

 cipal lines of folding. The centres of the earthquakes 

 lie on the side towards which the fault-surface is 

 inclined, and at a short distance (generally a mile or 

 two) from the fault-line. When manv after-shocks 

 occur, as with the Inverness earthquake of 1901, 

 their centres lie within a narrow band parallel to the 

 fault, and again on the side towards which the fault- 

 surface slopes. It is possible, indeed, to trace the 

 migrations of the seismic foci from side to side along 

 the fault, and gradual!}', as the series come to an 

 end, towards the surface of the earth. 



The occurrence of true earthquakes in this 

 country i- therefore decisive e\-idence that the growth 

 of Britisl . faults, ancient as man\- of them are, is not 

 yet at an end. This is not the c;-.se, of course, with 

 all our faults. The great majority, are, perhaps, at 

 rest, possibly h:-\e altogether ceaseci to grow. But 

 some, it may be a very small minor ;.', have kept 

 themselves bj' constant, if slight, exertions free to 

 move. That the absciute displacements a. 5 of small 

 account is evident from the fact that the;; are so 

 seldom destructive to property. But the displaceu-.ent 

 to which any one earthquake is due extends ov.er a 

 considerable area. In the Carlisle earthquake of 

 1901 and the Swansea earthquake of 1906 the total 

 length of the focus exceeded twent\- miles. 



The length of the focus of an earthquake can onh 

 be determined roughh' : in an\- particular case the 

 error ma\- amount to one, or several miles. But 

 there is a curious relation between the a\erage 

 length of focus in the different classes of earthquakes. 

 In strong earthquakes this average length is twelve 

 and a quarter miles, and in moderate earthquakes 

 thirteen miles. Slight earth(juakes may be di\'ided 

 into two well-marked groups. In one the focus is 

 nine miles or more in length, and the average length 

 tw elve miles : in the other the focus is six miles or 

 less in length, and the average length four miles or 

 less, b'urther evidence on this point is afforded by 

 the nature of the sound. Dividing the types of 

 comparison into two classes, according as they are of 

 long or short duration, we find that the percentage 

 of references to types of long duration is eighty-four 

 for strong earthquakes, eighty-one for moderate 

 earthquakes, eighty-four for slight earth(]uakes with 

 a long focus, and sixt\'-three for those with a short 

 focus. Thus, except for the slightest of all earth- 

 quakes, the average length of focus is practically the 

 same in all three classes. Their difference of 

 intensit\' depends only on the amount of displacement 

 witlnn the focus. 



There must be some reason for this close correspon- 

 dence in average length of focus. What the reason 

 is cannot at present he definitely settled, though 

 it can be surmised. The late M. Marcel Bertrand 

 published a map of France on which are depicted 

 the courses of the anticlines and synclines of the great 

 folds which traverse the rocks of that country. 

 The average distance between successive anticlines 

 or crests, measured along several lines at right angles 

 to them, varies from nine to twelve miles. For this 

 countrv no similar map. so far as I am aware, has 

 been prepared. I>ut it is at least a probable 

 supposition that the a\-erage distance between the 

 crests of the great crust-folds does not differ widely 

 in the two countries. If this be the case, then it 

 would seem that the length of a crust-fold, or 

 distance from crest to crest, ma\- govern the length 

 of the displacements along the faults which intersect 

 them. In other words, the growth of the faults to 

 which our earthquakes owe their origin is at present 

 chieflv due to the gro%vth of the transverse crust-folds. 



