Page images

jor gravity. What would happen to it as it part we cannot hesitate in rejecting it on account of the wards the earth's centre?

known porosity of all matter, and its observed yielding ties and pressures in the outer layers of the and condensation under great forces. by Laplace's law, are as follows:

On account of the difficulty in handling liquids, especially

when at high temperatures, they have not been so careDepth below

Pressure in fully investigated in the laboratory as solids; but there the surface Density


remains scarcely any doubt that under planetary pressure km. 2'55

I '000

they would all yield like sponges. 31.85 2608


In indicating his interest in the paper on planetary 6370 2 667


pressures (Astronomische Nachrichten, No. 3992), one of 95-55 2725


the most eminent British mathematical physicists has 127-40 2785


pointed out that to his mind the present writer has under191*10 2'904


estimated the probability that the earth has a metallic 254.80 3'025


nucleus. I have since pointed out in a letter to the editor 318.50 3'144 89,400

of NATURE (April 13, p. 559) that pressure, and not metallic 382'20 3'265 109,860

constitution, is the true physical cause of the earth's 445 90 3386 130,130

rigidity; for under such pressure any kind of matter would 529-60

3 508

assume a hardness greater than that of steel ; and as the
593 30

material is above the critical temperature of every sub

... 175,470 65700 ... 3751 ... 198,760

stance it is really gaseous, and would expand with in

credible violence if the pressure could only be relieved. The above table shows that before the mass has de

In the Astronomische Nachrichten, No. 3992, I have scended 31-83 km. (1/200th of the radius) the pressure about

shown that in any mass of considerable size, so condensed it would have become more than 8000 atmospheres, which

that the pressure amounts to millions of atmospheres, would force the molten fluid deep into the heated rock.

circulation at great depth becomes practically impossible, on the rising temperature at that depth would also rapidly

account of the friction due to the increasing pressure as dissolve the mass, and before the solid has sunk through | we descend within the mass. The pressure and friction another equal space in the viscid liquid, and thus reached which prevent circulation also prevent separation of the

depth of a k it seems almost certain that it would / elements according to their densities. While it may le completely dissolved.

not be possible to say that there is not an increasing It must be borne in mind that the solid is not much | amount of metals, such as iron, towards the centre of the denser than the liquid ; and as the liquid is highly viscous

earth, it is, I think, clear that there is no distinctively the mass would sink slowly, while the increase of tempera

iron nucleus ; for the existence of such a nucleus would fure and pressure would conspire together in the most

| imply that the earth's mass had unimpeded circulation powerful manner to dissolve the mass and reduce it to the

when in a fuid state, all of which is to the last degree sume temperature and density as the enclosing liquid, which

improbable. would be forced into it on all sides by a pressure vastly

When the earth was less condensed it was at lower greater than any known in our laboratories.

temperature, and the elements may not have been fused ; Even if we make the violent assumption that the sink

and as condensation advanced, and the temperature rose, ing mass is a kilometre, or several kilometres, thick, it

the friction due to pressure operated with increasing inis difficult to see how it could continue its downward

tensity to destroy circulation, which would thus be recourse, undissolved by temperature and pressure, below a

stricted to the subsidence of compact masses decidedly depth approximating one-tenth of the radius, or 637 kilo

denser than the surrounding fluid. As the fluid was metres. The sinking would be quite slow, owing to stiff

necessarily at high temperature, a compact mass would ness of the fluid, and could hardly be accomplished to

soon be dissolved. and further circulation of its elements this depth inside of several days, or more probably weeks.

practically cease. Moreover, before the mass reached a depth of 260 kilo

It seems, therefore, very difficult to escape the conmetres, or less than one-twentieth of the radius, the density

clusion that the earth's interior is a magma of all the of the molten fluid would become 20 per cent. greater than

elements, the increasing density towards the centre being ir was at the surface, owing to pressure ; and when the

due primarily to pressure. If any separation of the metals wolid mass was no denser than the surrounding fluid it

from the rocks took place, it could only be near the survould cease to sink. Or, if it had acquired a small

face where the pressure is slight; but because the rocks velocity downward in the fall from the surface against the

predominate at the surface, we must not conclude the same visrous resistance of the fluid, which is enormously in material does not exist abundantly in the great central creased by the eddy arising from the condition of con

nucleus of the globe. tinuity, it might go down a little lower until the motion

The difference in the point of view here adopted and was overcome by the buoyancy of the denser Auid below. that held by the older school of physicists is based primarily Accordingly, so far as one can see, solidified crust in sink upon the effects of pressure. While there is a certain dising could by no possibility go lower than one-tenth of the appointment in negative results, they are sometimes useful radius, which would hardly accomplish the building up

| in leading us to new conceptions, and perhaps we may of a solid nucleus.

hope that further study of these difficult questions will In considering the effects of pressure in forcing molten produce results admitting of general acceptance. It should fluid into the sinking solid, we have not assumed that the

be added that the pressures for the interior of the earth, density would thereby be increased; for at the great

calculated in the Astronomische Nachrichten, No. 3902, temperature of the fluid it is obvious that the solid into would not be very greatly modified by any other admissible which the hot liquid entered would be dissolved, and heat | law of density. from the fluid would be conducted rapidly through the The researches of Radau and Darwin (cf. Monthly solid mass. Thus no cause seems to be overlooked which

| Notices, Roy. Astron. Soc., December, 1899, pp. 122-3) could invalidate our conclusion.

have shown that, so far as the mathematical conditions It rests primarily upon the enormous pressures known are concerned, the law of density within the earth might to prist at great depths in the earth, and their undeniable depart considerably from that of Laplace. But on physical effert in forcing the molten fluid into any possible solid I grounds, including the incontestably steady rise of pressure hody, so as to prevent it attaining any considerable depth towards the earth's centre, whatever be the exact law of without dissolving: and upon the assumption that even density, and especially the observed vielding and conmolten rock under such forces would take approximately densation of all matter under such forces, I hold that the the density given by Laplace's law, which hardly admits

tially that of

ture of reasonable doubt.

from it in the actual arrangement of the matter of the In considering these questions heretofore, the hypothesis | globe is likely to be extremely small and unimportant. of incompressibility for the molten Auid has been tacitly

T. J. J. SEE implied or assumed. Whether such an hypothesis is justi U.S. Naval Observatory, Mare Island, California, fied will appear differently to different minds, but for our

March 31.





and alignments in 1901, but other pressing calls on my

time then caused me to break off the inquiry. Quite VI.-ON THE SOLAR OBSERVATIONS MADE IN BRITISH

recently it occurred to me that a complete study of STONE CIRCLES.

the Stenness circles might throw light on the question IN my last notes I referred to the star observations of an earlier Stonehenge, so I have gone over the

which might be made by means of stone circles. old papers, plotting the results on the Ordnance map. I now pass to solar observations.

Now that the inquiry is as complete as I can make I have already pointed out that much time has been it without spending some time in Orkney with a lost in the investigation of our stone circles, for the theodolite, I may say that in my opinion Mr. Spence's reason that in many cases the exact relations of the contention in his pamphlet on Maeshowe is confirmed, monuments to the chief points of the horizon, and although many of the alignments to which he refers in therefore to the place of sunrise at different times of support of it prove to be very different from those he the year, have not been considered; and when they supposed and drew on the map which accompanies were, the observations were made only with reference

his paper. to the magnetic north, which is different at different The alignments on which he chiefly depended were places, and besides is always varying; few indeed two, one running from the stone circle past the have tried to get at the real astronomical conditions of entrance of Maeshowe to the place of sunrise at Hallothe problem.

ween (November 1), another from the same circle by The first, I think, was Mr. Jonathan Otley, who in the Barnhouse standing stone to the mid-winter sun1849 showed the “ orientation of the Keswick circle rise at the solstice. " according to the solar meridian," giving true solar I give a copy of the Ordnance map showing the bearings throughout the year.

true orientation of these and of other sight lines !

[graphic][merged small][merged small]

I wrote a good deal in NATURE ? on sun and star : have made out. From this it will be seen that temples in 1891, and Mr. Lewis the next year ex- , observations of the sun were provided for on the days pressed the opinion that the British stone monuments, in question, and that the circles and outstanding or some of them, were sun and star temples.

stones were undoubtedly set up to guide astronomical Mr. Magnus Spence, of Deerness, in Orkney, pub- observations relating to the different times of the year. lished a pamphlet, Standing Stones and Maeshowe Of course, as I have shown elsewhere, such astroof Stenness, ,"3 in 1894; it is a reprint of an article nomical observations were always associated with in the Scottish Review, October, 1893. Mr. Cursiter, religious celebrations of one kind or another, as the F.S.A., of Kirkwall, in a letter to me dated March 15, astronomer and the priest were one. 1894, a letter suggested by my Dawn of Astro- I shall not refer to all the sight lines indicated, but nomy,” which appeared in that year, and in which deal only with those, bearing upon the Stonehenge the articles which had been published in NATURE in question, which I have without local knowledge been 1891 had been expanded, directed my attention to the able to test and justify. pamphlet; the observations had no pretension to scientific accuracy, and some of the alignments are But first we must consider the astronomical differwrongly stated, but a possible solar connection was ences between the rising of a star and of the sun, by pointed out.

which we mean that small part of the sun's limb first I began the consideration of the Stenness circles visible.

It is too frequently imagined that for determining 1 Continued from vol. lxxi. p. 538. 2 See especially NATURE, July 2, 1891, p. 201.

the exact place of sunrise or sunset in connection with 3 Gardner : Paisley and London.

these ancient monuments we have to deal with the

sun's centre, as we should do with the sun half risen. shown that the half-way time between an equinox and As a matter of fact, we must consider that part of a solstice is when the sun's centre has a declination the sun's limb which first makes its appearance above approximately 16° N. or S. In Orkney, with the the horizon; the first glimpse of the upper limb of latitude of 59°, assuming a sea horizon, the amplitude the sun is in question, say, when the visible limb is of sunrise or sunset is 32° 21', the corresponding 2' high.

azimuth being 57° 39'.

[graphic][subsumed][subsumed][ocr errors][ocr errors]
[ocr errors]


[ocr errors]
[ocr errors]

FIG. 13.-The Azimuhs of the Sunrise (upper limb) at the Summer Solstice. The Values given in the table have been plotted, and the effect of the

height of hills on the azimuth is stown. To make this quite clear I give a table which has Now the most interesting and best defined line been computed by Mr. Rolston, of the Solar Physics with this azimuth on the Ordnance map is the one Observatory, showing the true azimuth with hills up stretching S.E. from the centre of the Stenness circle to to high for lat. 599 N., the latitude of Stenness, to the Barnstone, with an azimuth of 57° 15'. The and 31°, nearly the latitude of Stonehenge, of the line contains between the two points I have named sun's upper limb for the solstitial vear.

another stone, the Watchstone, 18. feet high, in the SOLAR AZIMUTHS

Lat. 59


Rising N. of E. or Rising N. of E. or

Setting N. of W. Setting N.of W. 1. Sun's centre ; uncorrected ..

39 16

50 40 2. Upper limb; corrected for semi-diameter and refraction ..

... 37 1

49 20 and hill to high

... 38 34 ...

50 16

51 12 41 30

52 4

Rising S. of E. or Rising S of E. or WINTER SOLSTICE.

Setting S. of W. Setting S. of w. 1. Sun's centre ; uncorrected ...

39 16

50 40 2. Upper limb; corrected for semi-diameter and refraction.

41 24

52 0 and hill to high

39 54
38 23

508 36 54

49 14 The first important thing we learn from the table precise alignment; and from the statements made is that although at any solstice the azimuths of and measures given it is to be inferred that a still the rising and setting of the sun's centre are the same, more famous and perforated stone, the “Stone of the azimuths of the upper limb at the summer and Odin," demolished seventy years since, was also in winter solstices differ in a high northern latitude by the same line within the extremities named. some 50. The difference arises, of course, from the If we may accept this we learn something about fact that the limb is some 16' from the sun's centre, so perforated stones, and can understand most of the that considering the sun's centre as a star with fixed folk lore associated with them, and few have more declination, at rising the limb appears before the connected with them than the one at Stenness. I centre, and at setting it lags behind it.

suggest that the perforation, which was in this case It will also be seen that at sunrise hills increase the 5 feet from the ground, was used by the astronomerazimuth from N., and refraction reduces it; while at priest to view the sunrise in November over the Barnsetting, hills reduce the azimuth from S. and refrac house stone in one direction, and the sunset in May tion increases it.

over the circle in the other. Not only does calculation prove the worship of There is another echo of this fundamental line; the May and June years, but I think the facts now that joining the Ring of Bookan and the Stones of before us really go to show that in Orkney the May Via has the same azimuth and doubtless served the year was the first established, and that the solstitial | same purpose for the May year. (June) year came afterwards, and this was the chief | But this line, giving us the May sunset and question I had in view,

| November sunrise, not the December solstitial sunI will begin with the May vear. I have already rise as Mr. Spence shows it, is not the only orienta

tion connected with the May year at the stones of We have the November sunset marked by a standStenness. The November sunset is provided for by ing stone on the other side of the Loch of Stenness, a sight-line from the circle to a stone across the Loch Az. 53° 30'. of Stenness with an azimuth of S. 53° 30' W.

June rising, Az. true 39o. The top of Hindera fiold, To apply the table to the solstitial risings and set- more than 500 feet high, the highest peak, triangulatings at Stenness, and the sight-lines which I have tion station. plotted on the map, it will be seen that the table shows December rising, tumulus (Az. 41°) on Ward Hill. us that the lines marked

December setting, tumulus Onston 36° 30'.
N. 39° 30' E.

S. 41° 0' E.
N. 41° 16' E.
S. 36° 30' W.

General Remarks. are solstitial lines; to get exact agreement with the It is not a little remarkable that the winter solstice table the heights of the hills must be found and rising and setting seem to have been provided for at allowed for. I have roughly determined this height the Stenness circle by alignment on the centres of from the 1-inch map in the case of the Barnstone- two tumuli across the Loch, one the Onston tumulus Maeshowe alignment.

to the S.W. (Az. 36° 30'), the other tumulus being on On the N.E. horizon are the Burrien Hills, four miles Ward Hill to the S.E., Az. 41° (rough measurement). away, 600 feet high at the sunrise place, gradually It looks also very much as if the Maeshow tumulus

[graphic][subsumed][ocr errors][subsumed][subsumed][merged small][subsumed][ocr errors][subsumed][subsumed][merged small]

ascending to the E., vertical angle = 1° 36' 30". The was an after structure to use the Barnstone for the near alignment is on and over the centre of Maeshowe. summer solstice rising ; then these two other tumuli

, Colonel Johnston, the Director-general of the Ordnance to deal with the winter solstice at Stenness circle, Survey, has informed me that the true azimuth of may have been added at the same time. All these this bearing is N. 41° 16' E., and in all probability provided for a new cult. it represents the place of sunrise as seen from the There are also tumuli near the line (which cannot Barnstone when Maeshowe was erected. What is be exactly determined because the heights of the hills most required in Orkney now is that some one with are unknown) of the summer solstice setting; none a good 6-inch theodolite should observe the sun's' was required for the sunrise at this date, as the line place of rising and the angular height of the hills passes over the highest point of Hindera fiold, a at the next summer solstice in order to determine the natural tumulus more than 500 feet high, and on that date of the erection of Maeshowe. Mr. Spence and account a triangulation station. others made an attempt to determine this value with Another argument in favour of the tumuli being a sextant in 1899, but not from the Barnstone.

additions to the original design is that the place of The Ordnance maps give no indication of stones, the November setting from the Stenness circle is &c., by which the direction of the midsummer setting marked, not by a tumulus, but by a standing stone. or the midwinter rising and setting might have been As the stone near Deepdale and the tumulus at indicated from either the Maeshowe or the Barnstone. Onston are only about 1200 yards apart, the sugges

To sum up the solar alignments from the circle. tion may be made that in later times tumuli in some

We have the May sunrise marked by the top of cases replaced stones as collimation marks. Burrien Hill, from 600 to 700 feet high, Az. 59° 30'.



beds, and are regarded as the first evidence of a

neighbouring highland on which glaciers gathered. M R. ROGERS has produced a handbook to the The Devonian Bokkeveld beds follow, and the still

geology of Cape Colony which is sure to re higher and famous Dwyka conglomerate is, as all main a standard treatise. New observations will be geologists know, of Permo-Carboniferous age. It is recorded in future editions, as the work of his survey somewhat fascinating to conceive the growth of is carried on; but results made public as recently as glacial conditions through at least two long geo1904 are included in the present volume. The book | logical periods, until the flood of ice at last spread appears with especial appropriateness, now that the southward from the Transvaal territories, and scored visit of the British Association to South Africa has and rounded all the preceding rock-masses down to been officially organised; and the included geological the region of the Cape itself. map, on the scale of about one inch to ninety miles, The Dwyka beds, a facies of the Kimberley-Ecca gives an admirable impression of the country. In it series, and long regarded as volcanic tuffs, are here we see the huge Karroo synclinal, occupying almost all the colony, and lying between the pre-Devonian masses that crop out upon the north and the closely folded rocks of the Cape system along the south; while Mr. Rogers's introduction connects the scenic features with the geological structure in a manner that attracts us at the outset.

It is unfortunate that the names chosen for the colonial systems of rocks are not readily represented by adjectives. Hence such ungrammatical expressions as “ preCape” and “pre-Karroo " have been received indelibly into literature. Even the International Congress may hesitate to speak of an “ étage bokkeveldien,” though we have, to be sure, “purbeckien” and “bathonien" in Europe. This use of local names is, of course, greatly to be commended, in view of the scarcity of fossils in the great majority of the series.

The invasion of the old Malmesbury beds in the west of the colony by granite is concisely described on p. 38; and it is interesting to note how gneissic structures have arisen in the granite, as in so many other instances, without “evidence of a great amount of crushing or rearrangement of its component minerals after it solidified.” The foliation-planes in the gneissoid granite are parallel with the strike and cleavage of the adjacent sedimentary rocks, and the whole structure seems one of subterranean flow. The granulites of the Darling area will clearly bear comparison with those that have been so much discussed in Saxony. The intercalation of orthoclase crystals from the granite in lenticular areas between laminæ of slate (p. 43) reminds us, again, of the composite rocks of Donegal.

Mr. Rogers gives an interesting account of the stages in the passage from the well known blue crocidolite to the more siliceous FIG.1.-Overfolded quartzites of the Table Mountain Series, Meiring's Poort, repreyellow “ griqualandite" in the slates of sentative of the great upheaval, which probably took place in early Jurassic the Griquatown series. The slates them

times. From Rogers's “Geology oi Cape Colony." selves are converted into jasper-rocks where the most altered amphibole occurs; and the very adequately described, with several effective illuscrests and troughs of the folds have afforded hollows trations. The glacial series at Vereeniging is associin which the fibres of amphibole have crystallised across ated with beds containing the Glossopteris fiora, and from one surface to another.

also Sigillaria and other northern forms; and Mr. The Cape system, including the Table Mountain Rogers points out that the cold cannot have been series at its base, has been greatly contorted and responsible for preventing a more frequent mingling overfolded in the south; but the southern edge of of these two well marked floras. The most novel the Karro beds is also involved (p. 407), and the portion of the account of the reptiliferous Beaufort great east-and-west ridges of the continental margin beds of the “Karroo system” is the strong hint date from somewhere about Jurassic times. Flattened (p. 198) that they should be regarded as Permian and striated pebbles occur in the Table Mountain rather than Triassic. This view, based on Ama

litzky's work in Russia, would lead to a reconsider1 "Ao lacrodaction to the Geology of Cape Colony." By A. W. Rogers, M.A., P.G.S, Director of the Geological Survey of Cape Colony. Pp.

ation of the Elgin Sandstone also, and to the acceptTiu - 463 (London: Longmans, Green and Co., 1905.) Price os. net. ance of a development of reptilian life in Permian

« PreviousContinue »