Art. LI.—Volcanic Action regarded as due to the Retardation of the Earth's Rotation.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 8, 1875, Unnumbered Page

Art. LI.—Volcanic Action regarded as due to the Retardation of the Earth's Rotation. By John Carruthers, M. Inst. C.E., Engineer in Chief to the New Zealand Government. [Read before the Wellington Philosophical Society, 21st August, 1875.] Volcanic Action is so impressive in all its manifestations that it is difficult to realise (what, nevertheless, is true) that the mechanical energy required to produce it is much less than that displayed by the more uniformly acting powers of nature. If, for instance, the whole energy of the tides were converted into volcanic action, it would, in a short time, cover the face of the globe with mountains higher than the Alps or Andes, and render the earth as mountainous as the moon, where, if the hypothesis I now bring forward is correct, much actual motion has been converted into volcanic action, which with us, has not yet been so converted, but still remains in the form of actual motion of rotation. The power which has raised and still maintains our hills and continents above the sea is, I believe, derived from the retardation of the earth's rotation. It is quite certain that the earth does revolve less quickly than it used to do, owing to the friction of the tides against the bottom of the ocean, and it has been calculated that on this account the day is longer by one second than it was about a hundred and seventy thousand years ago. Motion cannot be destroyed, and, therefore, that which the earth has lost has taken some other form. The greater part of it has passed insensibly away as heat, having, after first slightly warming the earth, been radiated into space. Even the small part, which by my hypothesis, becomes volcanic action, would also pass insensibly away if it were not accumulated, as a weak stream of electricity is accumulated in a Leyden jar until sufficient intensity had been obtained to make its effects sensible. The strength and rigidity of the earth's crust act the part of this volcanic Leyden jar. There is, we know, a tide in the ocean due to the attraction of the sun and moon on the parts of the earth nearest to them, being greater than on the parts more remote. There is also a tide in the solid crust of the earth from the same cause, although the rigidity of the latter makes it one of strains rather than of movements. Still, it is contrary to all our knowledge of matter to suppose that there is absolutely no movement, for matter

when subjected to strain, more or less, and there must, therefore, be an actual daily tide in the land as well as in the ocean. If the earth had no more rigidity than water, there would be no rise and fall of the sea on the shore, for the land would rise as fast as the water. Our tides are the difference between the land and water tides; they prove and measure the rigidity of the crust. We know from them that the earth is rigid, and that it does not yield to strains, tending to change its form in the way a molten or viscid mass would do. We will now examine the effect of this land tide in order to ascertain the kind and relative amounts of the strains to which it exposes the crust. The tide due to the sun alone is about two feet; that due to the moon alone is about five feet, or seven feet altogether. The major axis of the equatorial tidal ellipse is, therefore, at spring tides, fourteen feet longer than the minor axis, or at least it tends to become so. The distance measured on the surface through the pole is twenty-one (21) feet longer from one end to the other of the major axis of the tidal ellipse, than from one end to the other of the minor axis. By the earth's rotation the end of the major axis passes away from under the sun and moon, and the end of the minor axis takes its place. The major then becomes the minor axis, and must tend to shorten fourteen feet, while the minor tends to lengthen the same amount. Part of the crust tends to stretch and part to compress twenty-one feet. The polar axis of the earth undergoes no alteration in length from these changes, but if the Sun and Moon in their motion in the heavens separate and come into quadratures, the tide instead of being seven feet is reduced to three feet. No change would take place in the strains above described, but the polar semi axis would be lengthened two feet, thus introducing new and very important tensile strains. We thus see that the crust of the earth is subjected to a racking movement which exposes every part of it to a tensile and compressive strain alternately, and that the polar axis is continually lengthening or shortening or at least tending to do so. The crust does not break under these strains, but if another action were added, which increased indefinitely the tensile strains caused by the tides, it will readily be perceived that fracture must take place sooner or later. Such an action does exist in the retardation of the earth's rotation, by which the strains due to the elongation of the polar axis are indefinitely increased. The earth owes its spheroidal form to its rotary motion; the compression of the poles depends on the velocity of rotation and varies as its square. If, therefore, anything reduces this velocity, the polar axis must

tend to lengthen. As before stated, the tides have this effect, the velocity is lessening, the polar axis is therefore lengthening, thus subjecting the crust to a continually increasing tensile strain under which it eventually yields. That it does yield is proved by the known fact that the earth has now the form due to its present rate of rotation. This cannot be a mere coincidence, but must indicate that the crust has yielded to the strains which tend to lengthen the polar axis, and we may take for granted that it will continue to do so. These are tensile strains and the fracture must therefore be sudden and complete. If they had been compressive, no actual fracture would take place, for the molecules would simply arrange themselves closer together and change of form would follow without fracture. With tensile strains this is impossible for, with them, the molecules are forced further and and further apart becoming every day less able than before to resist the strain, until at last the distance between them becomes greater than that over which the attraction of cohesion can act, when sudden fracture takes place. As soon as the crust yielded, the fluid interior to the earth, no longer held back by it, would at once take the form of equilibrium; that is, it would rush towards the poles, leaving the equatorial part of the crust unsupported; the latter would be unable to bear for a moment its own weight, or even a ten thousandth part of it; it would break at once at the weakest parts. The primary lines of fracture would tend to be north and south, but as a curved surface like the earth's crust cannot fit itself to a different curve, fractures transverse to the first would be necessary to allow even such an approximation as would enable the strength of the material to bridge over the minor inequalities. There will, therefore, be east and west as well as north and south fractures. Each line of fracture would become a line of volcanic action. Fig. 1 Let Fig. 1 represent a section of the earth at the equator after fracture. The central part is of less diameter than it was before, part of it having gone towards the poles. The crust is therefore too large to fit on to it, and cannot fall in so as to be supported by the central part without shortening by compression. We will assume that all the crust except the two masses at A has compressed and fallen in. It is apparent that before they can do so also, they must crush together lengthwise, as they have to occupy a smaller space than before. Wherever matter is compressed heat is evolved, and this will happen at A. It is to the heat thus evolved that I attribute the elevation of a mountain

range at A; we must therefore try and find out its amount. If we knew how much the earth masses had to compress before they could fall in, and the resistance offered to the compression by the elasticity of the material, we could calculate accurately the amount of heat generated; these we will now endeavour to ascertain. The change which takes place in the shape of the earth in about a hundred and fifty thousand years is equal to lessening its equatorial diameter one foot, and at the same time lengthening its polar diameter two feet; a reduction of one foot in the diameter at the equator would reduce the circumference three feet. The whole compression of the crust in Fig. 2 would therefore be three feet, for this amount of change of shape. If we assume the masses at A to be each one thousand miles long, or 1-24 of the whole circumference, each of them would have to compress ⅛ of a foot, which is therefore the amount of the compression at A. In order to ascertain the resistance offered to this compression we must find what resistance is offered to the compression of ordinary stone under different conditions, and for this purpose we will take for our standard a pillar one mile long, and for our unit of heat that quantity which would raise the temperature of the pillar one degree Fahrenheit. The area of the pillar is of no moment, as we want to know the length which would be warmed one degree by a given compression of the whole area. Stone expands about .000005 of its length when heated one degree; our pillar would therefore expand 1-40th of a foot when heated to the same extent. If instead of being heated the pillar were compressed 1-40th of a foot, an amount of heat would be generated exactly equal to that which would have expanded it 1-40th of a foot. Our unit of heat therefore represents also a unit of energy sufficient to compress stone 1-40th of a foot under ordinary conditions. The conditions may be, however, so altered that the same compression would require any given number of times more energy to effect it. For instance, if we placed the pillar in a great screw press and lowered the upper plate until it just touched the top of the pillar; if then we applied one degree of heat the pillar would not expand, being prevented by the upper plate; the same would follow at the second and the hundredth application of heat. If at last the screw were turned, and the pillar compressed one unit, the heat developed would be, not one degree but a hundred degrees; the elasticity of the stone had been increased a hundred-fold by the hundred units of heat applied to it, it therefore required a hundred times more force to effect the compression, and this is measured by a hundred times more heat.

If instead of warming the pillar we compressed it by applying equal amounts of mechanical energy the result would be the same, The first compression would warm the pillar one degree, it would then have double the elasticity, and the second amount of energy would effect only half the amount of compression, which would, however, be made against double the resistance and would therefore develope the same quantity of heat as the first, the pillar would therefore be warmed another degree, or two degrees in all, and so on with each application of energy. This assumes that all the heat generated is retained within the pillar. In experiments with short pillars this can never be the case, for the molecular motion is carried through the supports to the earth, but when the pillar is the earth itself, the heat can only be carried away by the slow process of radiation. The earth is, in fact, exactly in the position of our supposed pillar. Force causing compression has been applied to it by gravity, and is still being applied, heat must therefore be generated. When part of this heat is radiated into space and lost to the earth, the material loses part of its elasticity; gravity, which before was unable to compress it further, is then able to do so, heat is again evolved by the compression, and the elasticity of the material nearly maintained, the loss being measured by the radiation, less the heat generated by the further compression If our pillar were compressed until it were only half a mile long instead of a mile, and at the same time the temperature were maintained at 100,000 degrees (to which it would have been raised by the compression); as soon as the constraint was removed the pillar would expand to its full length of one mile, as every degree would expand it 1-40th of a foot, and 1-40 × 100,000 is equal to 2500 feet, or half a mile nearly. In the same way, if a portion of the interior of the earth which had been compressed until its specific gravity was doubled, were brought to the surface, it would expand until it had its original bulk, if it brought with it 100,000 degrees of heat; if, however, its heat were less, that is, if any part of it had been lost by radiation, it would not expand to the full bulk, and it would, therefore, have less elasticity than would be due to the compression it had undergone. In the case of the earth, some of its heat has certainly been lost by radiation. We cannot tell how much, but, for the sake of illustration, and to show that on any supposition which can reasonably be made there is sufficient left to account for volcanic action, we will assume that it has lost by radiation not more than 49 parts out of every 50 of its elasticity. On this supposition, the elasticity would be that corresponding to 2000 instead of 100,000 degrees of heat at that depth where the density is double

that on the surface, say, at a depth of about 800 miles, and it may be taken, for the purpose of illustration, to decrease uniformly towards the surface. If we assume the thickness of the crust to be 400 miles, its average elasticity would be that due to its average depth of 200 miles, or 500 degrees. Adopting this elasticity we can calculate by a simple proportion the heat which will be generated when the earth masses at A, Fig. 1, fall in. If a compression of 1-40th of a foot generates one unit of heat, how much will be generated with a compression of one-eighth of a foot when the resistance is increased 500 times ? It would be 2500 of our units, each of which would warm a mile of stone one degree Fahrenheit. As the earth masses have only a quarter of a foot to fall, the whole action would occupy about a seventh of a second. The molecules require a sensible time to transmit motion from one to another, the rate of travel being probably not more than a mile a second; the whole of the molecular motion would therefore be confined to a thin vein extending for only one-seventh of a mile on each side of the line of fracture, but reaching from the surface to the inner side of the solid crust. From this line of action the energy would spread along the crust, not by the slow process of the conduction of heat, but it would be transmitted at the rate of a mile a second. The difference between the conduction and transmission of heat may be illustrated by a modification of a time-honoured experiment. If the first of a row of suspended balls be warmer than the last, it will be a long time before any of the excess of heat reaches the latter by conduction, but if the first be lifted and allowed to fall, heat is generated, transmitted through the other balls, becomes potential energy, and may be converted into heat by allowing the last ball to fall again. It is by this rapid means that the energy collected along the line of fracture is spread; it will find the readiest means of getting into equilibrium, which would be by expanding the matter forming the crust. If we assume that four-fifths of it take this form within 500 miles of the line of fracture, we should then have 2,000 units of heat (being four-fifths of the 2,500 above mentioned) spread over five hundred miles. This would raise the average temperature of the whole by four degrees; the second earth mass at A, would also have its temperature raised in the same manner, giving together a mass 1,000 miles long and 500 miles thick. The expansion of such a mass when warmed four degrees would give a range of parabolic-shaped hills 1,000 miles wide at the base and 360 feet high. As the earth masses were assumed to be 2000 miles long by one mile wide, or 1-10,000th part of the earth's area, the whole volcanic energy represented by a decrease of three feet in the earth's compression, or in

150,000 years, would be 10,000 miles of such hills, which is much more than denudation would be able to wash down to the sea level in the same time. As soon as the expansion had taken place a great part of it would be lost at once, for the heat was only just sufficient to cause the expansion, and having done its work it ceases to exist. Molecular attracton and gravity would at once begin to pull down the elevation again, heat would be again generated, and would prevent further compression until the heat was lost by radiation. The first movement of depression would be almost instantaneous, indeed, the hill would not, unless the material were perfectly elastic, reach the full height above given; it would rise, however, to a height greater than that it could maintain, the excess of elevation being a function of the elasticity of the stone; it would then sink, rapidly at first, and afterwards very slowly. The normal condition of the earth is thus one of subsidence, and if it were not for volcanic action the solid land would soon sink beneath the sea level. Let the figure represent a section through the poles of the earth, the dark line being the form before and the fine line that after fracture. The equatorial belt having fallen in on to its supports, as above described, it would seem at first sight that there would be no compression in the polar sections, as the line E p E is shorter than E P E, and extension rather than compression would be looked for. When, however, the crust yielded to the tensile strain, the point p would be transferred bodily to P, and the curve E p E having a longer radius of curvature than E P E, there would be, in proportion to the area, quite as much compression before it could fit to its new place as there was in the equatorial belt. Fig. 2 We have hitherto regarded the sun and moon as being in the plane of the equator. Their declination would not have much influence on the amount of the strains to which the crust is subjected, but it would largely alter their direction, the pole of the tidal ellipsoid would travel nearly 30 degrees from the pole of the earth, instead of being coincident with it. This would greatly lessen the tendency of volcanic fractures to run in the cardinal points of the compass. When the earth changes its form the parallels of latitude passing through E E do not change their diameter, and there should be less volcanic action due to north and south lines of fracture at least, at these parallels, than elsewhere, the sun and moon's declination, however, by making these points

travel 20 to 30 degrees north and south, would quite obliterate this distinction. Lines of volcanic action would tend to be of long continuance, as the same causes which had at first decided the position of the line of fracture would be apt to make the next fracture occur in the same place; it is more than likely, however, with the enormous pressures which obtain in these movements of the earth, accompanied as they are by great heat, that rock when fractured would be welded together again, and it would not follow that a line of former fracture was weaker than other parts of the crust; because it had been once fractured. We will now return to the earth masses at A, Fig. 1. They have just fallen in on to the support below, and the crust on each side of the fracture is in a state of violent molecular agitation. The heat generated by the compression of the film nearest the surface, at once expands the matter of the crust, and is, by the expansion, imprisoned; it can afterwards escape only by conduction or radiation. From each point of the fracture, heat is rushing in every direction, some of it along the earth horizontally, but some vertically towards the surface. By the time equilibrium had been established the surface near the line of fracture would be raised by continual additions of heat from below to a high temperature, but not sufficient, except in rare cases, to cause volcanic eruptions. Where, as in the Andes, the elevation is very rapid, we may expect volcanoes on the summit of a range, but generally they are due to quite a local cause, and occur, not on the summit, but somewhat to one side of the main line of elevation. Where a bend occurs in the line of fracture, and still more, where two lines of elevation intersect each other, a cause of volcanic eruption arises, to which I believe is due by far the greater part of the active volcanoes in the world. If, for instance, the line of fracture be semi-circular, the points along the circumference would be elevated more than the points near the centre. The surface crust of rigid and inelastic rock would be left unsupported in the middle, and as it is far too weak to bear its own weight, it would at once crush together and fall in. If the bend be on a scale of considerable magnitude, the sheet of rock may have to crush together several feet before it can fall in. Every foot of compression, at the surface, developes enough heat to raise the temperature of one mile of rock by 40 degrees, as a compression of 1-40th of a foot would raise the temperature one degree; at a depth of 33 miles, owing to the increased elasticity, each foot of compression would raise the temperature of the same mass of rock by 3,300 degrees.

It is not an improbable supposition that, where the rocks had been warped to a convex form by previous elevations, a compression of three feet may often occur in this manner in a sheet of rock 33 miles thick. We should have then a vein extending right across the bend, and one mile wide, of which the temperature was increased by 120 degrees at the surface, and by 10,000 degrees at a depth of 33 miles. At one-quarter of the latter temperature the most refractory rocks would be fused; at the former, water would boil if the original temperature were 92 degrees. If once activity were established by an explosion of steam at the surface, the pressure on the heated rock below would be removed, and the eruption would extend to the very bottom. There is always water mixed mechanically with rock, and this would be raised to a temperature of 10,000 degrees. The tension of steam at a twentieth part of this temperature is immense, and if the pressure which retained it were suddenly removed an explosion would follow as violent as has ever been observed. The temperature of the lower strata is great enough to decompose the rocks and water subjected to it, and to account for the chemical action observed in eruptions. What occurs on a small scale at every bend in the line of fracture occurs on a great scale on the flanks of all lines of elevation. The inelastic surface rocks are supported after elevation has taken place only at the line of fracture, where the elevation is the greatest, and at the line where the elevation ceases, as shown in Fig. 3. Fig. 3 For a moment they bear their own weight and then crush down till they can rest fully on the mass below. Earthquakes of great violence would be felt, and under very favouring conditions a line of flanking volcanoes, parallel to the main range, would burst forth. In all cases of surface movement, where the heat developed has been insufficient to cause an eruption, the rocks are heated and thus placed in a position more likely than before to cause an eruption when a new movement occurs. The rain water also percolates to them, getting heated and returning to the surface as geysers and boiling springs. It will appear from what has been said that volcanoes are generally not deep seated. The mass of liquid lava which fills a crater rests on a cold and solid bottom instead of being in direct communication with the fluid centre of the earth. In the same way if shafts were sunk into the strata disturbed by the superficial action, the temperature would increase for some distance, but would afterwards decrease as we reached the part which had been affected only by the primary deep seated movement, and which had been increased in temperature only a few degrees.

The theory that the earth increases in temperature about one degree for every fifty feet as we descend is founded on observations made principally in mines in western Europe, that is, along a line of volcanic action, the present activity of which in England was shown by the earthquake of Lisbon affecting the waters of Loch Ness. Superficial action must here still be taking place and maintaining the warmth of the upper strata. This theory cannot be at all considered as proved by observation. It requires to be tested by sinking mines in various parts of the world, especcially in areas where subsidence has been long continued. We have spoken of the solid crust and liquid interior of the earth more for convenience of illustration than anything else. There is no sufficient reason, however, to suppose that any part of the earth is, or ever has been, fluid, but it is as well established as any other scientific knowledge founded only on inference, that as we descend, the elasticity increases, and for the purposes of our hypothesis, nothing more is required. On the assumption we have made above, the temperature increases only about 2,000 degrees in 800 miles. If this is near the truth the earth must be solid to the very centre, and some explanation is required of the slight power of resistance to tensile strain which we know it to possess. The explanation lies in the increasing elasticity as we approach the centre. Where stone is under compression, and is only kept by force from expanding, it has no tensile strength whatever to resist a strain tending slightly to lengthen it, or rather its strength is a negative quantity. The inner part of the earth is in this position, so that the whole strain is thrown on a thin outer layer which has to bear not only its own tendency to change its shape, but also that of the much larger inner part. It is, therefore, not surprising that it yields to very little more strain than that caused by the tides. That it does not yield to the tides alone is probably largely due to the rapidity of the earth's rotation, which carries away the part under the greatest strain so quickly from under the moon that the molecular motion has not full time to bring up all its forces against it. We must now show that the elasticity we have assumed for the interior of the earth is sufficient to prevent further condensation except as the interior heat is radiated away. The weight of a pillar of stone one square inch in area and one mile long, would be about 5000 lbs. At a depth of 800 miles the specific gravity is doubled, but the earth has there lost half of its attracting power, so the effective compressing force of equal bulks of stone would be the same in both positions. The total weight therefore which each square inch

would at this depth have to carry would be 800 × 5000 = 4,000,000 lbs. As the elasticity at the same depth is assumed to be two thousand times greater than at the surface, this would be equivalent to loading stone at the surface with a weight of 2000 lbs. per square inch, which it is well able to bear, especially if, as is probable, the resisting power of stone to compression is much increased by its increased density at the lower depth. There is, therefore, on this account nothing unreasonable in the assumption we have made, that the elasticity at a depth of 800 miles is not more than 2000 times greater than at the surface. Indeed, if the rocks at this depth are not greatly different from those we know at the surface, our assumption cannot be very far wrong. If the elasticity were much less than we have assumed it to be, the rock would be unable to bear the weight of the overlying strata, and would compress until sufficient heat had been evolved by the compression to give it the elasticity required. If on the other hand, the elasticity were much greater than our supposition, the rock would expand, lifting the weight above it until its elasticity had been sufficiently reduced by the loss of heat due to the expansion. It now remains to compare the hypothesis with the results of observation. If it will not stand the test it must of course fall to the ground, but I hope to show that my deductions agree well with the records of volcanic action in various parts of the world. It would follow from what I have been endeavouring to prove, that lines of elevation would tend to be continuous and of considerable length. This is so well known to be the case that I need not take up your time by giving illustrations. Secondly, such lines would tend to run in the cardinal points of the compass. A glance at the map will show that this occurs so frequently as to constitute a rule, although there are many exceptions—New Zealand for instance is a notable exception. The western slope of north and south mountains is generally much steeper than the eastern. This may, I think, be accounted for as follows: If a ball be conceived to be rolling along a plane surface without friction and to receive a blow in the opposite direction to that in which it is rolling, it will, if the blow be of a particular energy, be brought to rest; that is the motion of the ball, as well as the whole energy of the blow, has disappeared as energy, and has been converted into heat. If, on the other hand, a blow of equal energy had been given in the direction in which the ball was rolling, the velocity of the ball would be increased, that is, the ball would derive some energy of motion from the blow, and only a part of the latter would be converted into heat. In the same way the earth masses at A, Fig. 1, receive a blow when

they clash together. The molecular motion travels to the same distance in both; but in the western mass, the blow was administered in the opposite direction to that in which the mass was moving, and the greater part of its energy is converted into heat, close to the line of fracture, leaving only a small part to be carried on and distributed over the rest of the distance to which the action extends. The expansion depends on the heat evolved, so that it would nearly all take place close to where the blow was administered. With the eastern mass the heat would be carried on more uniformly, giving a more gentle slope to the hills. By the hypothesis, earthquakes and all the other sensible effects of elevation are quite superficial, extending only a few miles in depth. The great primary movement would be quite unfelt if the outer part of the crust were flexible. It is only the surface crushings and fractures which are felt. This agrees with the observations of Mr. Mallet, who, from calculations founded on the direction in which the shock was felt by observers at some distance apart, has deduced the depth at which the jar was given in different earthquakes. Quoting from memory, and at second hand, I believe he states this to be not greater than 30 miles. The elevation of a mountain chain does not add more matter to that part of the earth where it occurs; the matter which was there before is expanded, and thus has less specific gravity than the rest of the earth's crust. In the case of a great range like the Himalayas this loss of specific gravity might become sensible, and it has been found that this really is the case. The earth under the Himalayas has less specific gravity than elsewhere. My hypothesis requires that there should be a tendency towards volcanic action along the flanks of lines of elevation, at a considerable distance from the latter. To show that this agrees with observation, I will quote from Scrope's work on Volcanoes, to which I am indebted for most of the facts recorded below:— “In the body of this work it was stated that a more or less distinct parallelism, or coincidence, is traceable between the leading mountain chains of the two hemispheres and the linear bands of volcanic vents, active or extinct, insular or continental, by which they are traversed.” And again, after describing the great east and west line of elevation, extending from the north-west of Spain to China, he says:— “Now, it is the fact that these several mountain chains are bordered at moderate distances, on one or both sides, by linear bands of rocks of volcanic formation, along which therefore eruptions have, at some time or other, taken place.”

In New Zealand we see a line of extinct volcanoes, at the Waiau, Dunedin, Timaru, Lyttelton, and the Kaikoras, flanking the main range of the Southern Alps. In 1835, at the moment when the coast of Chili was shaken by a violent earthquake, a submarine volcano burst out near Juan Fernandez, fully three hundred miles away from the centre of elevation. That volcanic eruptions are most violent when different lines of elevation intersect each other, is proved wherever such intersections occur. In New Zealand we have one such line passing along the South Island and thence to East Cape in the North Island, reappearing again probably at Samoa and the Sandwich Islands. This line is intersected between Tongariro and White Island by another line which forms the north part of the North Island; a line of active volcanoes and hot springs marking the intersection. That Auckland does not belong to the same volcanic area as the rest of New Zealand is, I think, proved by the fact that only two out of the 234 shocks of earthquake which have been recorded have been felt there. One of them was felt nowhere else, the other was felt as a severe shock in every other town and in Auckland only as a very slight one. This fact is very remarkable, as Auckland is situated in the centre of a perfect nest of volcanic cones, and is, of all the places in the Colony, the one where manifestations of present activity would be looked for. The most active volcanic area in the world is the group of islands around Borneo. Here three great lines of elevation intersect each other, one forming the Malay Peninsula, another passing through the Phillipines to Japan, and the third forming Sumatra, Java, and New Guinea, and after passing through New Caledonia terminating in the Province of Auckland, unless it again reappears in the Chatham Islands. Another area scarcely less active than the above is formed by the intersection with the Andes of the line which forms the islands of Jamaica, San Dominga, and Puerto Rico. On the theory that the lava poured out by a volcano is derived from the fluid centre of the earth, it has puzzled speculators to account for many peculiar occurrences which are capable of easy solution by the hypothesis which I have brought before you. In March, 1861, the town of Mendoza was destroyed and 10,000 of its inhabitants killed by an earthquake, which was not felt on the other side of the range, only a few miles distant. If earthquakes are due to the sudden explosion of gas from the greatest depths of the earth, how can their effects be so local? If, however, they are due to the sudden fracture of a comparatively thin sheet of rock, there is no difficulty in understanding

how their effects may be felt over only a few square miles. At the moment of this earthquake the volcano, at the foot of which Mendoza was situated, burst into eruption. On the other hand, in February, 1797, when Riobamba was destroyed by an earthquake, the volcanoes of Cotopaxi and Tunguragua, at the foot of which the town was situated, were not in the least influenced, while Pasto, at a distance of 120 miles, “suddenly ceased to throw up its habitual column of water.” When the great earthquake visited Chili in 1835, its three great volcanoes burst into activity, and continued eruptive for months, while during the earthquake at Valdivia, in 1822, the neighbouring volcanoes were active for a few minutes only. In 1772 a Javanese volcano, Papandayung, burst into activity. At the same moment two other volcanoes, distant respectively 184 and 352 miles, also became active, while many intervening volcanic cones remained undisturbed. When, in 1843, Mauna Loa, one of the Sandwich Island volcanoes, was in violent eruption, the neighbouring permanently active crater of Kilauea, only fifteen miles distant, remained in its normal state. To account for these and other facts the central fire theory has to be modified by most extraordinary suppositions, all, of course, resting on no foundation whatever. First astronomers showed that the original idea was incorrect that the solid crust of the earth is very thin; it was shown that a less thickness than 400 miles would not accord with the observed precession of the equinox. Mathematicians also showed that the earth would not cool in the manner assumed, that is, only at the surface. It was then found that even 400 miles were too little for the thickness of the crust, and that 800 miles were necessary. This made it very difficult to believe that two pipes 800 miles long leading down to the fluid centre, could exist within fifteen miles of each other, and still be independent, as was necessary to account for the fact that Kilauea was not influenced by the eruption of its neighbour, Mauna Loa. A new disposition of the fluid matter was then devised, by which it was supposed to be placed as a fluid shell between a solid outer crust and a solid centre. The fluid shell was then supposed to be divided by walls of lava into separate channels, each crater having a connection with its own channel; Mauna Loa and Kilauea, for instance, were not connected with the same channel. Papandayung, and the other two volcanoes which broke out at the same time with it, were supposed to be connected with the same channel, while the intervening craters were connected with different ones.

Where two neighbouring craters are sometimes in eruption together, and sometimes not, the separating walls of the different channels are supposed to be melted down, or formed by solidification, as might be required to meet the case. These explanations are however, nothing better than accounting for the earth's stability by placing it on an elephant's back. We are not told on what the elephant stands, nor what causes the subdividing channels. The explanation, indeed, requires explaining more than the subject matter. Even if any amount of central fluid be granted, and channels subdividing it, the walls of which are formed or melted to meet every case as it may suit the convenience of the volcanologist, it is difficult to see how it explains volcanic action. The fluid matter either expanded, or it did not, after it received its excess of heat beyond that of the neighbouring solid masses. If it expanded, it would have no more tendency to rise to the top of a volcanic mountain, or even to rise a foot above its former level, than the Niagara River has to change its direction and run up the falls; if it did not expand, its excess of heat would be transmitted to the rest of the earth's mass at the rate of a mile a second, and it would not remain an hour in the fluid state, unless indeed it had sufficient heat to reduce the whole earth's mass to a fluid state along with it. On our hypothesis, however, each volcano is separate and distinct—they are all due to an elevation extending over great distances, causing local action wherever the local circumstances are favourable. They may, therefore, burst out together or not; may remain long in eruption, or only a second or two, without in the least invalidating the hypothesis. It is difficult to conceive how a channel of subterranean communication can exist over such great distances as are sometimes embraced in a volcanic throe. In 1755, for instance, when Lisbon was nearly destroyed by an earthquake, Kötlugja, in Iceland, burst into violent eruption while the lakes of Scotland were at the same time thrown into oscillation. It would, however, militate greatly against our hypothesis if such long lines did not exist. There remains one other test to which we can subject this hypothesis, but I am sorry to say I have not been able to collect sufficient information to apply it. As stated above, the primary cause of all movement of elevation is the lengthening of the polar diameter of the earth. I have assumed that the elongation necessary to cause fracture is so small that the daily tides would greatly influence the time and manner in which it would take place, which it would do even if the crust were strong enough to resist a considerable elongation of the polar axis.

When the strain caused by the tendency to elongation became as great as the strength of the crust would bear, the crust would begin to yield, and as point after point of it gave way, lofty ranges of mountains would arise in every part of the world. The sea would be confined to deep and narrow valleys; on every range great glaciers would form; icebergs would be floated from the polar seas in immense numbers, and the earth would pass through a glacial period. When the axis has reached its full length, and the crust was in equilibrium, elevation would cease—subsidence, which is never ceasing, would still be at work; the mountains would sink into the sea; glaciers would disappear, and a period of mild and insular climate would follow, such as seems to have characterised the deposition of the Coal Measures. If we take 10,000,000 years as the length of this cycle, it would indicate 130 feet as the amount of elongation of the polar axis, the tendency to which would cause fracture of the crust. The elongation due to the tides is two feet, so that they should influence to an appreciable extent the times when eruptions would take place. * Instead of a short period of 10,000,000 years, the cycle may, perhaps, embrace the whole time covered by the geological record, and as far as negative evidence is of value the teaching of geology seems favourable to such a supposition. In the oldest formations marine fossils only, have hitherto been found, a few land fossils afterwards appear; they gradually increase in numbers, as compared with the others, until the land and lacustrine flora and fauna assume eventually an importance equal that of the marine. In the earlier formations the types of life which predominate in our Museums are such as prefer an insular climate; types suited to a rigorous continental climate gradually supplant these, and appear to have reached and passed the time of their maximum importance. All this may be due to the incompleteness of the record; but it is exactly what would follow from the supposition under discussion. Many former geological cycles may have existed, each cut off from its predecessor, and the last of them cut off from the present cycle by a long period, during which the whole globe was covered by a nearly shoreless ocean. In each cycle there would have been continents and mountains as at present, due to the elongation of the polar axis, and at the close of each, when the crust had yielded completely to the strains to which it was subjected, a long period of subsidence would follow, during which the differences between the actual length of the polar axis and the length which it tended to assume, was too small to cause fracture of the crust. Gradually the land would disappear, and with it the whole of the terrestrial flora and fauna, except, perhaps, a few genera, which might be preserved on islands. When by the continued retardation of the earth's rotation, the theoretical length of the axis became greatly different from the actual, fracture of the crust would again take place, and the land would reappear. All fossils of the former cycles would be destroyed by old age. Those of the intermediate oceanic period and of the earlier stages of the new cycle, if they lasted long enough for us to see, would, as the oldest fossils really are, be little more than indications of organic matter. The genera which had not been destroyed during the oceanic period would, however, spread over the newly formed land, and our earliest terrestrial fossils would be those of plants and animals with fully specilised organs, so that we should have no record of the evolution of the higher forms of life from the lowest.

The greatest influence of the sun and moon occurs on the days of half moon, but fracture would most likely take place somewhat later. The weeks following the days of half moon may be considered those on which the majority of great eruptions would take place. I have only been able to ascertain the exact days on which 27 of these have been recorded—the earliest being the eruption of Monte Nuova, which occurred on the 29th September, 1538, and the latest that which took place in Iceland on the 11th June, 1873. Calculating the moon's age for the day on which each eruption occurred, I find that 61 per cent. are favourable to the hypothesis, and 39 per cent. unfavourable. There have been 234 days since 1868, inclusive, on which shocks of earthquake have been felt in New Zealand—nearly all very slight. I cannot say the result is very favourable, as the number of those which occurred during the weeks following, exceed those during the weeks preceding half moon by only 3 per cent. The total number of shocks contained in this analysis is far too small to be of any importance. As far as it goes, it is not unfavourable; indeed, the proportion of great eruptions which occurred during these weeks is very much greater than would be looked for. It would, however, require a complete analysis of many thousands of shocks before it could be stated with any confidence that the result was favourable or the reverse. It is, of course, the purest speculation to endeavour to estimate the per centage of influence which the tides would have; but they must have some influence if my hypothesis is correct. I now leave my hypothesis in the hands of the members of the Society for their criticism. I believe it explains most of the facts which have been recorded by observers better than any of the theories generally received; at least it does away with a great deal of unscientific world making in which volcanologists are too fond of indulging. The only essential suppositions I have made which are not supported by experiment or the clearest inference, are that the earth owes its interior heat to condensation, and has lost by radiation not more than 49 parts out of every 50 of the heat acquired by the compression it has undergone, since its average specific gravity was equal to that of the surface rocks. These suppositions are not at variance with the Condensation Theory, by which Helmholz has calculated that the solar system has lost 453 parts out of every 454 it had when condensation began. * I also assume that the inner parts of the Earth are not very dissimilar from the outer part which we are acquainted with. The theory of a great inner heat requires that we must suppose the centre of the Earth to be formed of molten gold, or of some other substance of at least equal specific gravity. Such startling theories can only be accepted when proved by the clearest evidence.

Before concluding, I will allude to two theories of mountain formation which have been lately brought prominently before the Society—the “Contraction Theory” of the Rev. Mr. Fisher, and the “Denudation Theory” as explained by Captain Hutton. The former explains the existence of mountains by the contraction of the earth's mass, due to the radiation into space of its heat. The upper part of the crust does not partake of this loss of heat, and consequently does not contract; it becomes, therefore, too large for the space it has to occupy, and in some manner, not clearly explained, the excess of matter is supposed to be arranged into mountains. Captain Hutton's theory is, that the matter brought down by denudation spreads over the bottom of the sea, where it attains great thickness; the lower parts then become warmed, in accordance with the theory that, as we descend into the earth, the temperature increases. Expansion takes place, and the superincumbent mass is lifted in a domical shape above the sea level. To both these theories there is what appears to me to be an unanswerable objection. They both assume that rock will not contract when subjected to a compressive strain slowly applied, while we know that in fact it will do so to almost any extent. In both cases the pressure is far within the ordinary elastic limits of the material, and the only change that would take place would be that the molecules would be pressed closer together. Long before the pressure became excessive, the molecules would have time to arrange themselves in their new form, and would be prepared to compress still further. We should have from both theories a denser rock but no mountains.