April 9 1903] 



NA TURE 



539 



not less than 19/20 of the earth's radius sufficiently 

 dense and rigid to comply with astronomical tests can 

 be defined, the same might also approximate to the 

 conditions assumed not only by seismologists, but also 

 by physicists. The shell covering such a nucleus would 

 be about 200 miles in thickness. The physical char- 

 acters of this shell would in all probability change 

 rapidly from those of the crust of the world to those 

 of its nucleus, corresponding to the observed rapid 

 changes in chordal velocities. At a comparatively 

 shallow depth, say 40 miles, high temperatures would 

 result in fusion, and inasmuch as ice, iron, copper and 

 other substances at or near their melting point float 

 on their own solutions, fusion is a state that would 

 partly be promoted by high pressure. At greater 

 temperatures, whatever the pressure might be, fluids 

 would become gaseous, and the gases would be dense, 

 but slightly compressible and viscous. In certain re- 

 spects, therefore, they would resemble a solid. This 

 is the view of Arrhenius, who assumes a core of gaseous 

 iron the dimension of which is that assumed by Wie- 

 chert. 



One reason for selecting iron or gaseous iron in an 

 equally dense state is that a nucleus of such material 

 of the specified size will account for the weight of the 

 world as a whole. What, however, is sought for is 

 a bodv probably a mixture of the commoner elements 

 in a state approaching that of closest crystalline atomic 

 packing, which has a radius 19/20 that of the earth, 

 a specific gravity less than that of iron, but greater 

 than 5.5, which keeps fairly homogeneous, and 

 can transmit compressional vibrations half as fast again 

 as steel. This material may be called geite, a term 

 as much required as magma and crust, by which geite 

 is enveloped, and geoid, which refers to the form these 

 materials collectively exhibit. 



Whether solid or gaseous, geite may possibly find its 

 chemical equivalent in certain meteorites, and there- 

 fore largely consists of iron alloyed with a small pro- 

 portion of nickel and other elements. If we assume 

 that the shell covering this mixture has a thickness 

 1/20 of the earth's radius, and an average density of 

 2 -7 — the density of the world being taken at 5^5 — it 

 follows that the density of the geite core is S'96, or 

 approximately 6. The elastic modulus for a core of 

 this density which conveys vibrations with a speed of 

 at least 9-5 km. per second is 451 x io 10 C.G.S., or 

 roughly speaking, a little more than twice the Young's 

 modulus for Bessemer steel (207 x io 10 C.G.S.). 



With improvements in seismometrical arrangements, 

 it seems Likely that speeds somewhat higher than those 

 here given will be recorded. Within the core itself, a 

 velocity of 9"5 km. per second must be exceeded. For 

 the moment let this be increased to 10 km. per second 

 whilst within the crust let the average speed be 3 km. 

 per second. With such assumptions, if the covering 

 shell is about 40 miles in thickness, the calculated times 

 to traverse chords corresponding to axes of 20, 30, 40, 

 50, 60, 80, 90 and 150 degrees would be 6*1, 7'5, 8 - 7, 

 io'2, ir6, i4'5, i5'7 and 21 minutes. The observed 

 times for these paths are 5, 6'5, 8-5, io"5, 12, 15, 16 

 and 22 minutes. These approximations between calcula- 

 tions and observations suggest that the region of rapid 

 change between crust and geite commences where 

 melting- temperatures probably prevail. 



In venturing these speculations on a geitic core which 

 will satisfy seismometrical and other tests, the fact 

 must not be overlooked that, as earthquake measure- 

 ments are yet in an embryonic state, figures which have 

 been given relating to the same, although they re- 

 present the work of many years, are subject to modifica- 

 tion. Amongst the various earth cores which are in 

 harmony with the requirements of astronomv and 



NO. 1745, VOL. 67] 



geodesy, there is at least one which is homogeneous. 

 If the radius of this can be increased 1/7 and it can 

 have the properties of geite, it will also accord 

 with seismometrical observations. 



Other speculations respecting the arrangement and 

 character of materials beneath the earth's crust are 

 based upon the fact that at certain observatories mag- 

 netic needles are disturbed by the large waves of earth- 

 quakes. These perturbations do not appear to be ex- 

 plained by the assumption that the magnetometers have 

 been tilted. An alternative is to assume that they are 

 due to changes in magnetic intensity possibly brought 

 about, as Capt. E. W. Creak, F.R.S., points out, by 

 changes of stress in a near magnetic medium. If this 

 is the case at those stations where needles are caused to 

 rotate, magnetic intensity and gravity should have 

 abnormal values. This appears to be true for Batavia, 

 near to which there are many volcanoes, indicating the 

 proximity of dense magnetic materials, and for Bombay, 

 where there is basalt, and at no great distance a hidden 

 chain of heavy matter revealed by gravitational obser- 

 vations. At Kew and Greenwich and other stations 

 where needles are not disturbed, magnetic intensity 

 and gravity are not abnormal. Generally speaking, 

 where horizontal force is comparatively low, the differ- 

 ence between the value of g as observed and as ex- 

 pected is also low, and to a certain extent the con- 

 trary holds good. On these points, however, until 

 more material has been collected, it is impossible to 

 speak definitely. 



What seismometrical observations then lead us to 

 suspect is that beneath the light crust of the earth, 

 which we know to be thinner in some places than in 

 others, there is a magnetic medium of density greater 

 than the crust, which, as we descend in depth, may 

 rapidly pass into a fairly homogeneous nucleus of 

 geite, the dimensions, physical and chemical characters 

 of which have been suggested. J. Milne. 



THE SOUTHERN CROSS ANTARCTIC 

 EXPEDITION. 



THE magnetic observations made in this expedition 1 

 have been reduced and prepared for printing by 

 Dr. Chree, F.R.S., and M. Bernacchi, and the meteor- 

 ological by Commander Hepworth, C.B., and Mr. 

 Curtis, of the Meteorological Office, under the direction 

 of Dr. W. N. Shaw, F.R.S., secretary of the Meteor- 

 ological Council, and the results have been published 

 by the Royal Society. In this expedition, fitted out by 

 Sir George Newnes, the magnetic observations were 

 made in about equal proportions by M. Bernacchi and 

 Lieut. Colbeck, R.N.R., other observers also giving 

 their assistance in the meteorological work. 



The magnetic observations consist of determinations 

 of declination, horizontal force, and inclination, made 

 at Cape Adare, in latitude 71 18' south, and longitude 

 170 9' east, with some detached observations of inclin- 

 ation at other places. At Cape Adare observations of 

 declination were made on a number of days in the 

 months of April, May, October, November and De- 

 cember, 1899, giving 'a mean easterly declination of 

 55 49'. Corresponding observations for horizontal 

 force give a mean value (C.G.S. units) of 0-04143, and 

 observations for inclination a mean value of 86° 34'. 

 Observations for the diurnal variation of declination 

 were made on three days, in April and May, 1899, and 

 January, 1900, respectively, giving on the whole a 

 diurnal movement of some 2 , that on the April day 



1 Magnetic and Meteorological Observation 1 ; made by the Southern 

 Cross Antarctic Expedition, 1S98-1000, under the direction of M. Borch- 

 grevink, Commander of the Expedition. 



