53§ 



NA TURE 



[April 9, 1903 



was taken in Fair Haven, the great bay at the north- 

 west angle of the main island, but it may be in Magda- 

 lena Bay. Incidentally, I may also mention that the 

 geographical nomenclature employed is very inaccurate, 

 Thus the name Mount Hedgehog, which" belongs to 

 Hornsundstind, is given to a hill on the east coast, and 

 other names are likewise misapplied. Mr. Arnold Pike 

 is called Mr. Pikes. 



The Swedes measured their base at Treurenberg Bay, 

 the Russians theirs near Whales Point (Fig. 2). For 

 this purpose they used the Jaderine apparatus, in which 

 .1 u ire consisting of Guillaume metal (a steel and nickel 

 alloy), about 25 metres long and 1-7 mm. thick, is sup- 

 ported at a fixed tension on a series of tripods, used in 

 pairs successively. By this means the base was measured 

 in four days, each measurement being repeated four 

 times with' two different wires. The limit of error is 

 slated to be not more than 1 in 400,000. 



At the beginning of the season of 1899 the Russians 

 went up to Horn Sound, and began establishing their 

 winter station close to a spot where Garwood and I 

 spent a week in 1897, so that it was not, as they 

 imagined, " a spot where for more than two centuries 

 no human being has lived." Here, in fact, throughout 

 the eighteenth and part of the nineteenth centuries the 

 Russians themselves had a trappers' winter establish- 

 ment. While the houses were building, the observers 

 went for a trip to the north, but the weather was very 

 bad. Then they went round to Wybe Jans Water 

 (which they call Storfiord) to commence the observation 

 of their ten triangles, one of which had a side 130 kilo- 

 metres long. They found the sea free of ice — an un- 

 usual condition to the eastward — and were able to land 

 anywhere with ease. They were astonished by the rela- 

 tively rich vegetation on Anderson Island. Not until 

 August 6 could they actually begin observations from 

 the signal point at Cape Lee, where they spent twenty 

 davs and could only work on three. They had to 

 abandon the place before their work was done. The 

 wintering partv settled in whilst the others returned 

 home. The winterers next spring made overland ex- 

 peditions to Mount Keilhau, and began work there. In 

 June, 1900, the other observers returned from Europe. 

 It was several weeks later before the Keilhau observa- 

 tions were complete. Meanwhile, others were explor- 

 ing the interior of the ice-sheet from Klaas Billen Bay, 

 to find a junction signal-point for the Swedes and 

 Russians. They succeeded after fortv-five days, and 

 built a pyramid on Mount Tchernycheff , a point first dis- 

 covered by me in 1S97. At Whales Head the observa- 

 tions were very protracted, and ice cut the observers 

 off, so that it was long before they could get away. An 

 expedition went overland to relieve them from Low 

 Sound (wrongly called Van Mijen Bav). This was 

 about all that was accomplished that season. 



In 1901 the weather was much more favourable. 

 The Russian base was measured near Whales Point. 

 The remaining stations were occupied as far as Thumb 

 Point, and the work completed. A final visit was paid 

 to the abandoned winter station, and the expedition re- 

 turned home in safetv and content. 



Martin- Conway. 



SEISMOMETRY AND GEITE. 



/"OBSERVATIONS on earthquakes which have trans- 

 ^S mitted vibrations to all points upon the surface 

 of our globe apparently lead to conclusions respecting 

 the physical nature of its interior. The following- notes 

 indicate the character of these conclusions, and at the 

 same time suggest directions in which these may be 

 harmonised with astronomical and other requirements. 

 Within a radius of io° or 20 of a centrum, the velo- 



NO. 1 745, VOL. 67] 



city of transmission of the larger earthquake waves 

 varies between i"8 and a little more than 3 km. per 

 second, such variations being usually attributed to the 

 nature of the medium through which the waves have 

 passed. Beyond these limits, and up to 165 — that is, 

 to near the antipodes of an origin — speeds which are 

 practically constant prevail. 



The large waves have a velocity which, if regarded 

 as " arcual," is constant at about 3 km. per second, 

 whilst the preliminary tremors, if it is assumed that 

 thev travel along paths approximating to chords, 

 quickly attain a velocity exceeding 9 km. per second. 



The constant velocity for the large waves and the 

 high velocity for their precursors preclude the idea 

 that either of them were transmitted through the 

 heterogeneous quasi-elastic crust. 



If the large waves are regarded as the outcroppings 

 of mass waves, then as pointed out by Dr. C. G. Knott 

 the law which would govern their transmission so that 

 their apparent arcual velocity should be constant would 

 be " most complicated and improbable." Considering 

 this uniformity of speed in conjunction with observa- 

 tions which indicate that as they pass beneath country- 

 after country they give rise to tilting phenomena on 

 the surface, and that the amounts of tilting recorded 

 at different stations in areas like Great Britain are, 

 at least for the smaller disturbances, practically equal, 

 the conclusion arrived at is, that the large waves of 

 earthquakes are transmitted through a comparatively 

 homogeneous medium beneath the crust, which, as 

 they pass, is forced to rise and fall like a raft upon 

 an ocean swell. 



If the preliminary tremors followed the same path 

 as the large waves, then their velocity would not be 

 constant, but would vary from 3 km. per second in 

 the vicinity of their origin to 15 km. per second as 

 they approached the antipodes. On the contrary, if it 

 is assumed that the paths approximate to chords, then 

 for chords of io°, 20 , 30 , 40 , 50 , 6o°, 8o°, 90 and 

 150 the corresponding average velocities in kms. per 

 second are from 3 to about 5, 7-3, 8*1, 8-5, 8-5, 8 - 8, g'o, 

 9-3 and 9-3 — these being minimum rather than maxi- 

 mum values. 



The lower of these velocities, all of which are aver- 

 age values deduced from observations dating back to 

 1889, may be due to the fact that they refer to the 

 shorter chords, a considerable portion of which lie 

 within and near what is assumed to be the crust of 

 the earth. 



But even accepting as appears to be necessary an 

 increase in average velocity along- paths as they are 

 taken nearer and nearer to the centre of the earth, the 

 above figures show that this increase is not very great. 

 The inference is that not only has the world a high 

 rigidity, but also that its interior is probably fairly 

 uniform so far as those properties are concerned which 

 determine the rate at which it transmits vibrations. 

 Possibly, therefore, it may have a density throughout 

 its nucleus which is nearly uniform. L'nless we as- 

 sume that as we descend in the earth elasticity and 

 density increase in about the same ratio, to which 

 hypothesis there are objections, it seems likely that the 

 nucleus of the earth has a density that is more nearly 

 uniform than is generally assumed. Prof. Wiechert 

 has shown that such a nucleus made of iron, density 

 8'2, and four-fifths of the earth's radius, covered by 

 a shell of density 3'2, satisfies the astronomer. Such 

 a world, however, does not comply with what appear to 

 be the requirements of seismology. Iron or steel do 

 not transmit vibrations at the observed rates, whilst 

 chordal velocities within the assumed shell would 

 1 ic ely approach those observed along chords which are 

 largely within the core. If a homogeneous nucleus 



