Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
GEOPHYSICS EARTH'S CRUST DEFORMATION"S IN GEOSYNCLINES BY F. A. VE~ING MEINESZ (Communicated at the meeting of Dec. 17, 1949) The beIts of strong negative gravity anomalies found in the Indonesian, Caribbean and Japanese areas have led to the hypothesis 1) that in these areas the Earth's crust has in recent geological periods buckled downwards thus forming a large bulge of crustal matter at the lower boundary of the crust. This crustal bulge has pushed away the plastic subcrustal matter and, being of smaller density, has caused the deficiency of mass revealed by the gravity anomalies. This explanation of the anomalies has been more and more wirlely accepted although other views have also been advanced; it will form the hase of this paper. For the problems dealt with here we need not go into the question that the rigid crust probably consists of more than one layer of increasing density wh en we go downwards, and th at thus the down-huckling gives rise to more areas in the cru stal cross-section where lighter rocks have been substituted to den ser ones, each area contributing its share to the anomalies. In this paper the writer wants to attack the problem ho\\" this crustal deformation got into being and this involves the question how far this deformation had a plastic and how far an elastic character. For this purpose he shall often make use of the investigations of BYLAARD about this matter and his conclusions will for a great part coincide with BYLAARD'S results. He shall especially base his discussions on the Indonesian Archipelago because so much is known about this area and among BYLAARD'S papers he may refer particularly to those of 1935 and 1936 which likewise deal with this area. In the first place we have to choose between the supposition of a purely elastic beginning of the buckling of the crust and, at least partly, a plastic start of the deformation. The second assumption, defended by BYLAARD, has several advantages with regard to the first but it offers also difficulties; wh en the lateral compression in the crust has increased to a value th at plastic deformation sets in, the problem arises e.g. whether the crust will not bulge to both sides and continue to do so in such a way that isostatic balance will remain valid (fig. lb) instead of the development of a geosyncline Ieading to a buIging downwards as fig. la shows. As the gravity anomalies clearly indicate th at in the critical 1) F. A. VENING l\'[EINESZ, 1930, 1934, etc. 28 belt isostasy is strongly disturbed and that, therefore, the actual crustal deformation must be much nearer to la than to lb, it is only possible to assume a plastic behaviour of the cru st if we can prove that the deformation in that case will not lead to lb. ~ Fig. la. Fig. lb. A second problem we have to consider is whether in case we may assume deformation la the plastic behaviour of the crust can lead to a second crustal wave as shown by the left part of fig. 2 which apparently .::::::: ~ V ~ Fig. 2. is present in tbe western part of Indonesia viz. in the oilgeosynclines of Java and Sumatra. 2) Before starting an investigation of these problems we shall first examine the difficulties involved in the hypothesis of a purely elastic beginning of the phenomenon. In tbe first place we find in that case that the stress needed to bring about buckling of a rigid crust of a thickness of 30 km must be of the order of 40.000 to 50.000 kg/cm 23 ) and it seems hardly likely that the cru st could stand stresses of such an amount. It has, however, already been pointed out by SMOLUCHOWSKI, 4) who arrived at the same conclusion, that much smaller stresses would be sufficient if we could assume that the crust consists of severallayers. If we may admit that the friction between these layers is negligible, we find that the stress required for buckling would be lln th part of the above figure, n being the number of layers, jf we assume the layers to have equal thickness. For bringing down the buckling-stress to a more acceptable value of 3000 kgfcm 2 we would, therefore, have to assume the cru st to consist of some 15 layers and this seems a rather large number. It seems, moreover, questionable whether the friction between the layers would not prevent such a development. There is still another way out of this difficulty. GRIGGS 5) has suggested 2) Called by UMBGROVE idiogeosynclines, p.43 e.s., Umbgrove 1942, or 2nd Ed. 1947, p. 49 e.s. 3) F. A. VENING MElNESZ, 1934, p. 48 or 1948, p. 70. ') M. SMOLUCHOWSKI, 1909. 5) D. A. GRIGGS, 1939. 29 that the downbuiging of the erust might be eaused by the sinking part of a eonvection system descending below the belt. The drag exerted by the horizontal parts of this system on the lower boundary of the crust to both sides of the belt is on eaeh side direeted towards this belt. This may thus explain the strong eompression of the cru st in the belt. At the same time the effect of the sinking eurrent on the erust may lead to the cru stal down-buIging at stresses below the buekling-limit. This hypothesis is no doubt attraetive aud it would solve our problem. The writer, however, se es difficulties. In the fi.rst ' place it appears likely that su eh a eonveetion-eurrent eould not have great dimensions; as the sinking eurrent ean not have much Iarger cross-section than a few hundred kilometers and as the distance of the two beits of the Banda arc is not more than 600 km, the total horizontal dim en sion of the system could hardly exceed 300-500 km. As from other sources we must suppose the speed to be at least a few cm per year, it is unlikely th at sucb a current could continue for the whole period of an active geosyncline which persists probably over some fifty million years or more; in th at time it would make more than one complete revolution while the study of convection in the Earth leads to the conclusion that such currents never make more than a half turn. This is also GRIGG'S hypothesis as given in the paper mentionecl. It would of course be possible to suppose that a CUlTent was only present during the perioels of actual folding which occur from time to time during the history of the geosyncline and each of which cover a much shorter time. But in that case it would be surprising that· the current would repeat itself at the same place after a first one had occurred. This first current-system must already have been causing a certain concentration of crustal matter in the belt above the sinking current anel the higher percentage of radio-active minerals in th is matter must thus bring about a gradual rising of temperature in the belt which makes it unlikely that a sinking current below it would again co me into being. A second anel still more serious clifficulty with regard to GRIGG'S hypothesis of local convection-currents causing the down-buIging of the crust is based on the impression that the beIt in Indonesia as weIl as that in the Caribbean area each form a great unit extending over many thousands of kilometers and showing a systematic course and a systematic oecurring of the anomalies; this makes it unlikely that local eurrentaystems could have brought them about. Concluding the writer thinks that this mechanism of bringing forth the down-buIging of the cru st is difficult to admit and that thus the problem remains how su eh a phenomenon coulel take place if we maintain the supposition of elastic deformation of the crust at the start of it. A seeond difficulty is hrought ahout by the hypothesis, in recent times arrived at by the author, that the substratum itself is no viscous fluid but that it has a small strength which must be overcome before flow 30 ean set in. 6) It has likewise been advaneed by GRIGGS in his paper of 1939 and sin ce long by JEFFREYS. The writer shall not go here into the arguments leading to this hypothesis but he may point out that it obviously makes it diffieult to adhere to an elastieal down-buekling of the crust. For maintaining this irlea we should need an initial downward deformation of the cru st whieh would make it possible that the momentum thus eaused by the resultant of the horizontal stress in the erust would not only overeome the resistanee against bending of the erust but also the elastie resistanee offered by the substratum. It is, besides, likely that this last resistanee will involve a eonsiderably larger value of the buekling-stress needed in the erust and that thus the number of separate layers in tbe cru st whieh we had to assume for bringing this stress down to an admissible value would have to be still greater. A third problem offered by the hypothesis of a purely elastie beginning of the crustal deformation in the tectonic areas is to find an explanation for the eurved and often tortuous course the deformed beIts mostly show. It is hardly likely that this is eaused by an intrieate pattern of forces bringing about these deformations. If we ex amine e.g. the main anomaly belt in lndonesia we eome to the conclusion that it is probably eaused by a regular field of forees showing an unique or nearly unique rlirection ') given by the arrows on the map and so it seems out of the question to attribute the eurved shape of the belt to the forces. It is, however, possible to account for part of the irregular courses of the mobile beIts by the hypothesis of pre-existant zones of weakness of the Earth's crust as advaneed qy the writer in 1943. 8 ) The presenee of su eh zones must make it likely that the crustal deformation will in part follow them. The pattern of shearzones derived in the papers referred to has been based on the assumpt.ion of a polar shift in a very early stage of the crust's history; the directions derived for lndonesia have been inrlicated in the map. Jt is unlikely, however, that the presenee of such zones of weakness can explain all the features of the eomplicated belt-pattern of the orogenic areas; in lndonesia e.g. the part parallel to Java and several other parts have still to be aeeounted for. This is diffieult in case we assume the pattern to develop as an elastic deformation of the crust. As BYLAARD has remarked sueh a deformation must be expeeted to oeeur in beIts at about right angles to the force and although the eurvature of the Earth might somewhat affect it, it could not deviate as much as the existing beIts, as e.g. that south of Java, aetually do. As we shall afterwards examine more in detail a plastic deformation eould give a mueh better explanation of these direetions. It is well known that in a plate subjeeted to a single stress-direction the plastic deformation 8) 7) 8) F. A. VENING MElNESZ, 1948, p. 39. VENING MElNESZ, Gravity Expeditions at Sea, Vol. II, p. 122, Vol. IV, p. 30 etc. VENING MElNESZ, 1943 and 1947 (see e.g. 1947, p. 56). 31 after the elastic limit is exceeded will occur under an angle of about 55° with this direction and not at right angles to it; this direction is e.g. shown by the lines of HARTlI'IANN developing at this moment. BYLAARD and afterwards VAN ITERso::-r have given an explanation of th is angle and have also derived the formula for the angle occurring in the general case of a more complicated stress condition in the plate. BYLAARD has al ready shown that these directions of the plastic deformation may give a more satisfactory explanation of the pattern of the tectonic beIts. \Ve shall co me back to this point. Concluding we may say th at the assumption of an elastic deformation at the beginning of the phenomenon involves many difficulties which are avoided by the supposition th at from the start it has a plastic character. \Ve shall now examine the two problems raised by the plastic hypothesis, the question whether a downbuckling of the erust can be thus eXplained and whether it is possible that an adjoining wave develops. We shall begin by attaeking the first problem. Let us assume that in a belt over a breadth b the eompression exceeds the elastie limit of the ernst and that a plastic deformation has set in, leading to a thiekening of the erust, whieh towards the edges of the belt disappears as represented by fig. 3. We may probably assume th at the rate ofthickening is extremely T ~-- Fig. 3. slow, e.g. of the order of a few mm per year or less, and in that ease we may praetically neglect the viscous resistance of the sub stratum against the forming of the bulge at the lower boundary of the erust. 9) As a eonsequence of this we may expeet the heights ho and hl of the upper and lower bulges to have a ratio eorresponding to iso statie balanee and ") This may be derived from the speed of Ï!lOstatie readjustment whieh the writer found to be given by thc formula of p. 662 of his paper of 1937 w = 5 t:, L emfyear in whieh w = vertieal speed of the erust in emjyear t:, = vertieal deviation of crust's surfaee from the equilibrium position in km diameter in 1000 km of equilibrium cleviation The formula has been derived from the rate of post-glacial uplift of Scandinavia. L = 32 if we adopt the usual values for the densities of the crust and the substratum of 2.67 and 3.27, thîs gives (I) This asymmetry is important for our problem because it must involve a deviation from the regular distribution of stress in a vertical crosssection of the crust. As ho and !tI are no doubt small with regard to the breadth b of the bulge we shall probably not make a great error in neglecting the complications in the transition zones at the edges of the belt and in assuming that the normal stress-component in each crosssection is a linear function of the vertical coordinate z. As we also know th at the resultant of these components must coincide with the original axis of the crust, our assumption allows us to derive the distribution. If we denote the va lues of the normal stress a at the upper and lower boundaries by ao and al' their difference by a f:, and their mean value u:; alJ ~ 4, - ,I "4 I 1 Fig. 4. by a rn which is also the mean value of the stresses a oyer the whoie eross-section, we can derive (2a) and if we use (I) and neglect hl + ho in the denominator (2b) " f:, =20.70~. G,n T The differenee a f:, of the norm al stresses at the two erustal boundaries must lead to a difference in the shortening of the horizontal dimensions at both boundaries and, therefore, to a down-bending of the crnst. We see that thus a downward move ment of the ernst may be explained. We must, however, derive its value to find out whether it can compensate and even exceed the rising ho of the upper surface so that a geosyncline is formed . The downward move ment affects the formules (2a) and (2b). If we denQte the sinking of the crustal axis in an arbitrary eross-section by 33 zand in the middle of the belt by Ze, the value of al - U o (see fig. 4) is increased by 2 z, resp. 2 ze' and so instead of (2a) and (2b) we get (3a) and using again (1) and neglecting hl (3b) + ho in the denominator Ot:; = 12 (1.725h o +z). T Om For the middle cross-section we have to substitute Ze to z. As the dimensions of the phenomenon in the sense of the belt are large we may assume the rate of shortening of each partiele of the crust in the sense of the compression by the stress a (x direction) to be equal to the rate of thickening in the vertical sen se (z direction). As we suppose no stress to be working in vertical direction on the crust the usual stress-strain relation of viscous fluids gives here for the rate of both deformations ( 4) where u and ware x and z comp. of displacement and 17 the coefficient of viscosity. If we denote the length in x direction after deformation of the whole deformed area at the surface by bo and at the lower boundary by bI (undeformed both = b) we get h ~ (b1-b o)= -i~ b f a t:; dx= 6~~ f (1.725h o +.z) dx. o u Contenting us with an approximate solution we may - if Z is zero at both ends of the deformed part - assume th at the mean value of Z over this area is i Ze. It would, however, be possible that the adjoining non-plastic parts of the cru st also take part in the ben ding and in that case we should have to add a figure for the sinking of the ends of the plastic part. Supposing this to be proportional to Ze we get a mean value of zover the plastic part of Introducing this we obtain: ( 5) We shall now express bI - bo in ze and ho in the time t. 3 34 From fig. 5 we derive: and so bT e=-b bo 1- Fig. 5. and (6) Multiplying (5) by hJ8 T we get (7) azc 3 Om b ~ = 4 '1T2 2 [ 1.7 25 h0 + ("§ +,u )Ze] • The value of ho and of the total increase of the crust's thickness which we shall denote by h (h = 5.45 ho) can be found because we know that the compressive stress U m and, therefore, the rate of thickening of the crust since the beginning of the plastic deformation, i.e. since the time t = Ot has been constant and that this rate is given by formula (4). 'We thus obtain: (Sa) 35 and integrating: h= (8b) Om 21'] Tt. Introducing the value of ho = 0.1835 h in (7) we get the following first order differential equation in ze: aZe (9) ( 3) ~- 1+~,u 2 fl m b 2I']T2 Z e - 0 .475 (flm)2 b2 t _ 2~ T - O. We can simplify this equatian by choosing a new varia bIe instead of the time t, yiz. the increase of the crust's thickness h which is proportional to t. By means of (8) we find ( 10) This no longer contains explicitly the quantities (1", and 1]. The solution of the equation is simpIe. Taking into account that ze = 0 for t = h = 0 we obtain ( 11) = Z c 0.475 [_1_ T3 1 4" 1 iJ f' b2 + + (e(1+~I') ~.: h_ 1) - hJ. For finding the total amount of the sinking of the cru stal axis in the middle of the plastic belt we have to add the sinking at the edges of this belt which we assumed to be ,u Ze and so we have to multiply our formula by 1 ,u. Denoting the result by z' we get + ( 12) For ,u = 0, i.e. when the adjoining non-plastic parts of the cru st do not take part in the bending, the formula becomes _b: z'=O.475 [ T3 ;:;2- ( eT' ( 13) h ) J - 1 -h . We obtain the formula for the sinking z~ of the crust's surface in the middle of the belt by subtracting t.he rising ho = 0.1835 hand so we get ( 14) which for ,u = 0 simplifies to b' ( 15) , Zo = 0.475 T3 ( Ti' h b2 e - 1) - 0.658 h. b2 b2 By multiplying these equations by T3 we see th at T3 z~ is only functian of -~- h anel ,u, a conclusion which we could already have drawn 36 from the differential equation (10). So we find th at z~ and hare proportiona.I to the third power of Tand inversely proportional to the square of b. We may add that by computing numerical values for these quantities for different values of ft we find th at as a rough approximation both quantities are inversely proportional to 1 ft, at least inside the limits o < ft < 1, which will probably not be exceeded by the cases occurring in practice. Formulas (14) and (15) give the following values (Tabie I) for three values of ft , i.e. ft = 0, ft = tand ,u = I. The curve derived from these + values for b~~i as a function of ~~ is given by fig. 6; it is practically the same for the three cases but on different scales. For all three the vertical dimensions are four times the horizontal ones. The z~ axis is positive downwards. TABLE I p.=o b2 h p. = O.i'i b2 zo' T3 1'3- Q.327 0.618 0.654 0.981 - 0.031 0 0.009 0.149 + + b2 h b2 zo' - j'3 T3 0.212 0.401 0.4240.636 I' = 1.0 - 0.0207 0 0.0056 0.0997 + + b2 h b2 zo' T3 0.1572 0.2971 0.3144 0.4716 'f3 - 0.0154 0 0.0047 0.0765 + + The horizontal axis of fig . 6 gives h which is proportional to t ; it may, therefore, be considered as a time-axis. The first line of Table I shows the maximum negative values of z~ , i.e. the maximum rise of the Earth's ,surface in the belt. We see th at after this first rise, the effe et of the sinking of the crust'g axis begins more and more to dominate the effect of the rise caused by the plastic thickening of the ernst. At the values ·of h given by the second line of Table I the rise in the centre of the belt is already neutralized and, as fig. 6 shows, the sinking becomes .afterwards more and more marked. So we see that our attempt is successful and that a down-buckling of the cru st as a result of plastic deformation can not only weIl be explained but appears to be an unavoidable consequence of it. We shall now assume values for Tand b in order to derive the values of z~ for the maximum rise and for other parts of the curve. For T we shall assume the value of 30 km which follows from the great majority of the observed gravity values. It is more difficult to make an estimate of the breadth of the plastic belt b. It is in fact likely that it will be variabie, even for parts of the same tectonic belt. \Ve shall choose three values for our computations viz. 50 km, 100 km and 150 km; the last is probably too large. Table II gives a list of the maximum rise in the 37 centre of the belt for the different cases and of the sinking after a three times longer time-interval. TABLE II maximum rise at time tm b 50 km 100 " 150 " 1'=0 I' = 0.5 335 m 84 " 37 " 223 m 56 " 25 " I I' = 1.0 166 m 42 " 18 " sinking at time 3 tm 1'=0 1610 m 401 " 178 " I I' = 0.5 I I' = 1080 m 269 " 120 " 1.0 826 m 207 " 92 " We shall now make an attempt to find the time-intervals involved in the phenomenon. We can begin by deriving h by means of the same Or---------------~~--------- Fig. 6. assumptions about Tand b as used above. Next we have to apply formula (8b) for obtaining the corresponding times t. Here, however, a great deal of uncertainty sets in because of our lack of knowledge 38 about the stress U m and the viscosity 'YJ. The stress U m must slightly exceed the strength of the crust and so the writer thinks that (16) Um = 3000 kgJcm 2 may be areasonabie value. Of 'YJ we know practically nothing. It must no doubt considerably exceed the viscosity of the sub stratum which has been derived from the post-glacial uplift in Scandinavia; the writer found in this way a dynamic viscosity 'YJ of 3 X 1022 poises. 10) From the fact, moreover that never any evidence has been found of flowage of rocks in mountains in case of strong shearing-stress, an indication of an even higher lower limit of 'YJ may be derived; we thus find that it is likely to exceed at least 1023 poises. For our purpose we shall put (17) 'YJ = 1024 poises. This value is of course subject to considerable uncertainty and the same is, therefore, true for the following estimates. Before applying these assumptions to our main problem we shall use them for obtaining an idea of the rate of thickening of the crust. By means of formula (8a) we find ()h = 0.15 cm per year. ot ( 18) As far as the order of magnitude is concerned this result appears acceptable. As we have already mentioned on p. 31, a rate of a few mm per year allows us practically to neglect the viscous resistance of the sub stratum and this is necessary, for otherwise this resistance would play a part in preventing the free development of the down-bending and down-buIging of the crust. The above value of (jhJ(jt implies a· rate (jhoJ(jt of the rise at the surface of 0.03 cm per year. Returning to our main problem we shall apply formula Sb and use the values mentioned before of Tand b for changing from the quantity b2hJT3 to the corresponding time. We shall do so for computing the time-interval tm in which the maximum rise at the centre of the belt comes into being. We thus find the values of Table lIl. TABLE 111 Time tm of maximum rise in years b 50 km 100 " 150 " 10) . VENING MElNESZ, #=0 # = 0.5 # = 1.0 2.490.000 620.000 280.000 1.610.000 403.000 179.000 1.200.000 298.000 133.000 1937. 39 These re!mlts for the time elapsed since the beginning of the plastic deformation of the erust till the maximum rise oecurs, must be doubled to obtain the time-interval between this maximum rise and the sinking at the time 3 tm given in Table Il. Most of that sinking is oeeurring in the seeond half of this last interval, that is to say in an interval equalling that of the tabie. The order of magnitude of these resuIts seems acceptable. It is hardly necessary to add that the sinking wiII continue with increasing speed after t.he time 3 tn! till the whole phenomenon wiII change when it leads to the entire breaking together of the crust. Considering the long periods we have found 11) it appears that the value we assumed for the viscosity 1] of the crust could not be many times larger; a value of 1025 for in stance would give ten times greater time-intervals and such intervals seem highly improbable. Refore leaving our subject we may point out th at in the developing geosyncline sedimentation must occur and, because of its load and the isostatic readjustment, t.he rate of sinking must be considerably increased. In general we may estimate the sediments at four to five times greater thickness than the depth of the geosyncline if na sediment at ion or filling up by water had oecurred, When during the further development the crust breaks together and is pushed downwards, the buoyancy of the whole structure must become larger and larger and this must lead to the lifting up of the adjoining cru stal beIts. In the gravity field th is shows up by beIts of positive anomalies accompanying on both sides the belt of strong negative anomalies caused by the downbulge of the crust. This prove8 again the extremely high viscosity of the crust; if it were less the central belt would have been ab Ie to rise for readjusting its isostatic equilibrium and no negative anomalies could have been left. The time elapsed sin ce the downbuiging of the crust is some ] 5-·20 million years and so during this whole period no large isostatic readjustment has apparently taken place. The second problem to be attacked is whether for a crust in a plastic state we can explain th at a second syncline develops on the outside of one of the raised beits adjoining the main syncline; we have already mentioned that this seems to have occurred in the idio-geosynclines in Java and Sumatra, i.e. on the inside of the curved arc. As these last synclines do not appear to lead to an ordinary tectonic belt with great folding and overthrusting of the surface layers as we have discussed up to now (although there is evidence of some folding of these layers) it seems indicated to surmise another phenomenon. In particular it appears unlikely that the whole cru st has become plastic here and has started bya plastic thickening. 11) The last line of TableIII will probably not occur as it corresponds to a larger value of b than is likely. 40 In view of these facts it seems reasonable to suppose that the crust was near the limit of plastic flow before .the wavè developed ' and that during this development the part of the croSs-section whéPe the bending caused a decrease of the compressive stress remained in the elastic state while the part where the compression increased became in its entirety subject to plastic flow (fig. 7). We denote the limit of plastic flow by av and the stresses at surface and lower boundary by ao and al. 1 1 1 1 p=Ta--~) 1 V I ! r +---p Fig. 7. The lengthening per unit of length at the lower boundary has an elastic character and can, therefore, be written while the speed of the shortening at the surface IS given by and, in case ao remains constant, the total shortening by t 2 '1. 00 We have two conditions for a(p al and w. In the first place the crosssection must remain plane, and, therefore, if we neglect the thickening of the cross-section because of the plastic flow in the upper half ( 19) In the second place the total integral of the stresses over the crosssection added by the bending must be zero. Denoting the increase of the thickness T of the cru st because of the flow by 8, this condition gives and as 8= oot. 2'1 we find (20) t (T-w) 41 Eliminating (21) G. - GI from (19) and (20) we obtain = (T~- _)2 W 00 - 0 •• 00 2'7 Et + 00 E in which the second term of the right member is brought along by the plastic thickening of the crust. Taking the X axis horizontal and the Z axis vertical and denoting the deviation of the crustal axis from its original position as it is caused by the ben ding by z we find for the momentum G v Tz of the bending-stresses of the crustal cross-section (neglecting 8) ( 22) For the curvature of the crust's axis caused by the bending we obtain 1 è)2z (23) è)x 2 = W 0.-0, -E· Combining (22) and (23) we get the differential equation for the curve of the cru st 's a xis (24) which has the same shape as the equation for buckling of the crust in the elastic state provided w = t T, which in the elastic state is of course true. The half wave-Iength L of the curve is given by L2 (25) = 'll2 ~ = è)2z ;t2 3 w2 ~ 0v è)x 2 The distance w is provided by (21); we see that the wave-Iength L is proportional to it. Examining (21) we find th at for small values of t we obtain large values of wand L. These values become smaller for increasing time till for t infinite we should get the following limits T w=--- (26a) 1+ I/~ r 0. (26b) which for T = 30 km E = 1.000.000 kgfcm 2 G v = 3000 kgfcm 2 w= 1.58 km give L= 52 km We may conclude th at there must be a tendency towards a shortening of the wavelength with time but we can not assume a gradual shortening as mentioned here because such a shortening would imply a continual 42 change of Go and so our formula of p. 40for the shortening at the upper boundary of the crust of Gotf2 1] would no longer be valid; this wonld invalidate all the deductions. Formulas (21) and (25) lead to the further important conclusion that waves of this type can not originate without being started from the outside by an effect enforcing a cru stal deformation. This must continue for such a time-interval th at the first term of the right member of (21) gets down to an acceptable value. In order to obtain an idea about this question we may point out that for 1] = E and t = = 102~ poises l.000.000 kgfcm2 100.000 years we get 2'1 Et = 0.64. This wonld lead to reasonable values of wand L. Putting e.g. Go - G~ = 120 kgfcm2 we obtain 11: = 4.3 km and L= 140 km For smaller values of Go - G v the value of t could of course be smaller. Resuming we may conclude that this type of wave can only come into being by an outside disturbance working for a considerable time. This is of course the case for our problem. The lifting by the main belt's buoyancy of the adjoining crustal beIts must have a disturbing effect on the cru st outside these beIts and if this crust would fulfill the condition of being near the limit of plastic flow it seems probable that a downward wave of the type here described would develop. As the crust in this case is not entirely plastic it appears unlikely that such a wave would lead to a complete giving way and to a down-buIging of the cru st as occurs in the main tectonic belt and this is in good harmony with what is known about the idio-geosynclines of Java and Sumatra. As this type of wave can only come into being in the special condition mentioned, it is likewise clear that it does not occur everywhere near a tectonic belt. We may point out here the possibility that such waves will also originate when the compression G of the crust is not exactly equal to G v but slightly below it. Our formulas then become somewhat more complicated but the principle remains the same. We may safely assume that such waves will be more difficult to bring about for G differing more from G v• We may lastly remark that the tendency towards a shortening of the wave~length in a later stage is perhaps discernible in the idio-geosyncline 43 in Java, where in the area south of Surabaya it shows clearly a division in two synclines by a secondary ridge in the middle. A similar breaking up in smaller synclines appears to occur in the Sumatra idio-geosyncline. Resuming our investigation we may conclude that a plastic state of the crust may provide us with a good explanation of a cru stal downbuIging in the main tectonic belt and that it does not make it impossible to explain the originating of the idio-geosynclines of Java and Sumatra. In view also of the other advantages there appears, therefore, good reason to accept th is hypothesis. As we have al ready l1lentioned it enable!'i us to come to a better understanding of the complicated pattern of the tectonic beIts. RIJLAARD has already pointed this out in 1935 12 ) and shown this to be the case 13) for the whole tectonic area east and Houtheast of Asia including Indonesia. The writer shall deal here with the latter area for which his views slightly differ from RIJLAARb's treatment although in many regards he shall follow the same lines. Examining the map we see that the direction of the belt of negative anomalies west of Sumatra fo11ows one of the two directions of the world 's shear-pattern already l1lentioned on p. 30 and supposed to have been brought about hy a shift of the poles in an early stage of the crust's history. The same is approximately true for the belt fo11owing the eastcoast of the Philippine Islands. In both areas the gravity anomalies, the seismic movements and the topography are in harmony with the view that the main component of the relative movel1lent of the crustal blocks of Indonesia with regard to the outside crust is parallel to these parts of the belt and that these blocks onIy show a slight overriding overthe outside cru st combined with a small compression. It is not difficult to understand th at the tectonic belt in part fo11ows pre-existant zones of crustal weakness. The character of the relative movement of the crustal blocks points to a direction of this movel1lent in the sen se of the arrows, one for the western and one for the eastern half of the archipelago; the two directions nearly coincide and perhaps the sl1lall di vergen ce is not more than an error of interpretation. The part of the anomaly-belt south of Java differs from the crustal shear-pattern. Here thehypothesis of BIJLAARD, dealt with in our paper can give a good explanation. As it has already been mentioned the direction of the belt of plastic deformation for our case of a single stressdirection may be expected to make an angle a of 35° with the crosssection on which this stress works, i.e. an angle of 55° with the stressdirection itself. 14 ) If, besides this main principal stress el' there is a second smaller principal stress e2 working in a crustal cross-section at 12) BIJLAARD 13) BIJLAARD 14) BIJLAARD other papers. 1935. 1936, see e.g. map on p. 20. 1931, 1935 and 1936 and many other papers, lTERSON 1943 and 44 right angles to that of the stress el' the angle a has a different value for which BIJLAARD derived the formula (27) cos 2 a = ~ + 1.'1 1.'2 31.'1 - et which for ~2 = 0 gives a = 3(j9 (i.e. 55 9 with the stress el itself). As the map shows this value of a agrees weIl with the belt south of Java at least with the western part of it. Towards the east the belt curves slightly and a diminishes towards a value of about 20 9 • This ma)' perhaps be explained by the direction of flow which we may expect in the belt in the beginning of the phenomenon when no great downbuIging has yet occurred; according to BIJLAARD the flow in the belt, if free to do so, would make an angle of a = 35 9 with the stress; th is direction is indicated in the map by the dotted arrow. This flow-tendency must probably give rise to a secondary stress in the sen se of the belt graduaIly increasing towards the east. According to formula (27) a value of e2 = 0.4 el would bring down the value of a to 20 9 • The eastern half of the archipelago shows a much more complicated pattern of the tectonic phenomena. From north to south we have two beIts of strong negative anomalies of which the southern one is the prolongation of that south of Java. The northern one ends over east Celebes; it shows still stronger negative values of the gravity anomaly, viz - 204 mgal a.gainst about - 130 mgal in the southern belt. We must, therefore, assume that it means a greater shortening of the cru st ; it is, ho wever, more irregular in the sense of its axis. lts disappearance in Celebes must lead us to the supposition that at least in the northern part of the archipelago a fault-zone must exist dividing the eastern from the western Indonesian blo ck ; along this fault-zone astrong relative movement must have taken place. We shall not further elaborate this problem and only mention a study by the writer published in Gravity Expeditions at Sea, Vol. IV, p. 31 e.s., where he is led to suppose a major fault-zone in one of the two shear directions already mentioned, running over S.E. Celebes towards the NNW and ending at the eastern end of the Himalayas. The crustal shortening of the second Indonesian belt here discussed would thus represent the continuation of that of the Himalayan zone. The SSE movement of the eastern Indonesian block would have about the same direction as this fault-zone; it is given by the arrow of the map. Returning now to our subject we see on the map that the northern and southern beIts show similar directions making an angle in the rniddle and that in both cases these directions from both sides are enclosing angles of 55 9 with the sense of the force as given by the arrow. So here again the hypothesis of a plastic beginning of the crustal deformation may give a satisfactory explanation of the course of the two beIts. If we can accept this also for the southern belt or, in other words, 45 if we have no accidental coincidence here, we can draw the important conclusion th at the course of this belt would not primarily be determined by the shape of the Australian continent as has often been assumed. This view would be in harmony with the opinion given by many geologists th at the whole area of the Banda Sea has continental character and that this sea has fairly recently been formed by the sinking down of the crust. The further course of the southern tectonic belt to the north of the part here considered, i.e. from the Key Is towards the eastpoint of Ceram is given by the NNW direction of the crustal shear pattern and adjoining this part we get the Ceram-Buru stretch which may perhaps again be considered to make an angle of 55° with the arrow. Near Buru, where the negative anomalies become weak, it beg ins to deviate from th is direction. The northern belt of negative anomalies practically stops towards the north where the part considered ahove comes to an end. Resuming we may say that the intricate pattern of the tectonic beIts east of the supposed major shear-zone through S.-E. Celebes and the Mangkalihat PeninsuIa can for the greatest part be weIl explained by a combination of the crust's shear-zone pattern and the direction with regard to the force in which plastic deformation will be likely to occur. There is still one part of the southern belt not yet discussed, viz. that between W. Sumba and E. Timor. In the same way as it is the case for the part further to the east which we have already examined, the inner Banda arc is more regular here than the outer arc where we find the belt of negative anomalies anel the great crustal deformation. More than one supposition can be made to explain the curve of the latter belt in the Sumba-Timor area but the writer thinks that the data are not yet sufficient for coming to a satisfactory conclusion. Resuming our investigations we may say th at BIJLAARD'S hypothesis about the crust's deformation in the tectonic beIts having from the beginning a plastic character can explain many questions otherwise difficult to account for. Summary. In comparing the hypothesis that the great deformations of the Earth's cru st in the tectonic beIts begin by having an elastic or a plastic character the writer comes to the same conclusion as formerly BIJLAARD already arrived at th at the latter is more likely. By assuming a deformation of the plastic kind throughout the whole phenomenon a satisfactory explanation can be given of the downbuIging of the crust which the beIts of negative gravity anomalies lead us to suppose in these beIts. At the same time we can thus explain the intricate pattern of the beIts in Indonesia anel it appears likely that this is also possible in other orogenic areas. 46 REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. BIJLAARD, P. P., Weerstand van een verzwakte scheeve doorsnede van een getrokken plaat, berekend volgens HUBER·HENCKY. Ingenieur, 37 (1931). - - - , Beschouwingen over de knikzekerheid en de plastische vervormingen van de aardkorst in verband met de geologie van den O. Indischen Archipel. Summary in german. Ingenieur in Ned. Indie, 11 (1935). - - - , Théorie des déformations plastiques et locales par rapport aux anomalies négatives de la gravitation, aux fosses océaniennes, aux géosyncIinaux, au volcanisme, à l'orogénie et à la 'g éologie de l' océan pacifique occidental. Rapport congrès d'Edimbourg de I'Union Géodésique et Géophysique Intle, (1936). GRIGGS, D. A., Theory of Mountain buildi~g. Americ. Journ. o. Sc., 237 (1939). ITERsoN, F. K. TH. VAN, Bijdrage tot de plasticiteitstheorie. Short summaries in english, frenoh, german. Versl. Nederlandsche Akad. v. Wetensch., 52, (1943). SMOLUCHOWSKI, M., Anzeiger d. Akad. d. 'Viss. Krakau, Math. Naturw. Kl. 2 (1909). UMBGROVE, J. H. F., The Pulse of the Earth, Ie Ed. 1942, 2e Ed. (1947). VENING MEINESZ, F. A., Maritime Gravity Survey in the Netherl. East Indies; tentative interpretation of the results. Proc. Kon. Akad. v. Wetenseh., Amsterdam, 33, 6 (1930). VENING MElNESZ, F. A., UMBGROVE, J. H. F., KUENEN, PH. H.; Gravity Expeditions at Sea, 1923-1932, Vol. 11, Netherl. Geodetic. Comm., Delft (1934). VENING MEINESZ, F. A., The determination of the Earth's plasticity from the post-glacial uplift in Scandinavia. Proc. Kon. Akad. v. Wetensch. Amsterdam, 40 (1937). Spanningen in de aardkorst tengevolge v. poolverschuivingen, short, summaries in english, french, german. Versl. Ned. Akad. v. Wetensch. Amsterdam, 52 (1943). - - , Shear Pattern of the Earth's crust. Transaction" Americ. Geophys. Union, 1 (1947). - - - , Gravity Expeditions at Sea, 1923 - 1938, Vol IV. Netherlands Geodetic Comm., Delft (1948). F. A. VENING MEJNESZ: Earfh's crust dejorrnafions o · .. . B RE.GIONAL ISOSTATIC ANOMALlE.S . [:<.::::J +100 - +150 milligal. [2J + 50 - +100 0- + 50 50 - 0 -100- - 50 . . -150- -100 rnIlII - 200 - - 150 _ -200 0 ITIITill _ -250- .. . , " " ~[ARROWS GIVE THE DIRtCTlOIJS OF THf SUPPOSfO COMPRfSSIVf FORCES DIRECTIONS CRUsr's SHEAR PATTERN o R N E