Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 31, 1898
This text is also available in PDF
(559 KB) Opens in new window
– 583 –

Art. LIX.—The Wanganui Earthquake of the 8th December 1897.

[Read before the Philosophical Institute of Canterbury, 2nd November, 1898.]

Plate LVI.

The earthquake of the 8th December, 1897, was felt generally in the districts around Cook Strait, from Opunake to Nelson, and beyond those districts as far north and east as Auckland, Gisborne, and Napier, and as far south as Timaru. The returns are sufficiently definite to determine the epicentrum and the velocity, and the circumstances afford a good opportunity of reviewing the data of the shocks in recent years that have proceeded from the, same origin.

The returns received from the telegraph-offices through the courtesy of the Telegraph Department were as follows:—

– 584 –

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Place. Time of Beginning of Shock, N.Z.M.T. Apparent Direction. Apparent Duration. Effects. Remarks. Intensity. Rossi-Forel Scale.
A(a)Opunake A.M. 2.41* E to W. and N. to S. 20 secs. Everything on the move; violent oscillation; people awakened; feelings of nausea; small ornaments, &c., thrown off shelves, Long and loud rambling; shock a succession of violent jerks; after shock loud roaring of surf on shore, and heavy gust of wind; peculiar hissing sound at conclusion of main shocks, followed by a slighter shock vi.–vii.
Wellington 2.41* (?) some secs. Several clocks stopped at 2.41 in Wellington Post-office, including the tower-clock. Very, unpleasant rumbling noise previous to shock [Prolonged tremor; violent disturbance, then vibrations; doors burst open at General Post Office; bells rung.—Press Association.] vi.–vii.
Napier 2.42* (?) 6–8 secs. General awakening of sleepers; stopping of clocks. No rumbling during or after shock vi.
Auckland 2.43* E. to W.(?) (see next column) First a slight tremor; then pause from 1 to 2 minutes; then prolonged shock, 20 to 30 seconds, Windows rattled, wooden walls creaked; not felt by everybody; shook beds; awoke some people, No rumbling heard [“Two distinct shocks.”—Auckland Star.] v.
Lincoln 2.43* From Mr. G. Gray and Mr. Coleridge Farr—a good time observation iv.

[Footnote] * Verified.

[Footnote] * Verified.

[Footnote] * Verified.

[Footnote] * Verified.

[Footnote] * Verified.

– 585 –

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Place. Time of Beginning of Shock, N.Z.M.T. Apparent Direction. Apparent Duration. Effects, Remarks. Intensity, Rossi-Forel Scale.
A(b) Gisborne A.M. 2.41 ½* 2.43 S.W. to N.E. 53 secs. Clock stopped; no alarm, No preceding ramble; at first series of shocks, strong, gradually dying away iv.-v.
Gisborne Christchurch 2.43 E. to W. 15–18 secs. Distinct prolonged shock iv.
B(a) Wakapuaka 2.42* (barely) S. to N. 30 secs. Loud rumbling, 12 seconds; first shock, 3 seconds; short interval; second shock, 5 seconds; rumble, 8 seconds, General awakening vi.
Wanganui 2.42* N. to S. and N. W. to S. E. 1 min. First part very heavy, 15 seconds N, to S.; then sharp upheavals N.W. to S., upsetting movable objects and chimneys; tremors gradually subsiding [Sharpest since 1855; 2,40, lasting 3 minutes, Damage in crockery shops; chimneys levelled in all directions; main water-supply stopped; fissures in ground sub-sidence of railway for several chains on reclaimed ground; house burnt down through upsetting of lamp.—Press Association.] viii. +
Picton 2.42* N.W. to S.E. ab't 1 min. Two office-clocks stopped, No damage vi.
Marton 2.42* N.W. to S.E. 20secs. Felt by every one, Oscillation of lamps; clocks stopped; some movable objects overthrown; plaster cracked in some buildings; chimneys fell three or four miles off, but none in Marton, One long shock, marked at beginning; gradually increasing, attended by rumbling, Liquids 1 ½ in. below rim overflowed vessels, Direction, N. of W. vii. +

[Footnote] * Verified.

[Footnote] * Verified.

[Footnote] * Verified.

[Footnote] † “It has recently been discovered that a piece of ground, 20 acres in extent, near East Tokomaru, twelve miles north of Wanganui, has been tore and rent in all directions by the earthquake which was experienced two months ago, Many of the openings are 10 ft, or 12 ft across, and strong fumes of sulphur are issuing therefrom.”-Press Association, 11th February, 1898.

[Footnote] * Verified.

[Footnote] * Verified.

– 586 –

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Place. Time of Beginning of shock N.Z.M.T. Apparent Direction. Apparent Duration. Effects. Remarks. Intensity. Rossi-Forel Scale.
B(p)Nelson A.M. 2.43* N.E. to S.W. 20 secs. Plaster shaken from ceiling; crockery shaken; sleepers awakened; clocks stopped. Rumble immediately before like approach of very strong gust of wind. Two shocks, 2 seconds between [The first shook for 1 minute, at 2,40. No damage.—Press Association.] vi.+
Blenheim 2.42 N.W. to S.E. nearly; 1 min, Houses creaked; windows shakes; crockery and glasses jingled; sleepers disturbed; town-clock rung. Shook sudden, rocking like jolting in rock; then long tremor and rumbling vi.
Greymouth 2.45 N.E. to S.W. 20 secs. Doors and windows rattled; sleepers awakened. Some say rumbling before and after—not heard by observer v.–vi.
Bull's 2.42 N.E. to S.W. 50 secs. One clock stopped; a little crockery broken. vi.–vii.

Press Association Reports.

Cambridge, 2.43: Woke every one;.no damage (vi,). Hamilton, 2.43: Many docks stopped; no damage (vi.). New Plymouth, 7: Of exceptional duration (v.–vi.). Hawera: Extraordinary severity; no damage (v.–vi.), Masterton; Prolonged and severe; most severe on east coast, where clocks stopped, and people alarmed (vi.), Pahiatua and Carterton, 7: Severe (v.–vi.). Woodville; Lasted 45 seconds; severest for years (vi.). Manaia and Patea: Severe; a few chimneys fell (vi.–vii.). Timaru, iii.–iv.

[Footnote] * Verified.

– 587 –

The earthquake was the severest felt in the colony since the Nelson earthquake of the 12th February, 1893; and the severe effects were more widely felt than in the latter. These two were more severe than any others since the memorable earthquake of the 23rd January, 1855.

Besides the intensity, which was rather over viii. On the Rossi-Forel scale, the prolonged character of the shocks is a noteworthy feature; there were evidently several maxima, and normal and transverse vibrations appear to have been so fax distinct in their times of arrival as to give the impression of two different shocks.

A(a) and B(a) are first-class observations, the times being verified, and other indications of consistency and the known experience of the observers confirming this idea of their value. All the verified times—i.e., A(a) and B(a)-together with Gisborne and Nelson, were taken at first to form eleven equations of observation. All were counted as of equal weight, for I do not yet see my way to give greater weight to the times from more distant places, as has been suggested. The greater numerical value of the coefficients naturally gives a steadying value, as it should, to the equations based on the data from them.

The normal equations were formed and treated as shown in my previous papers. It hardly seems necessary to give them here.

An examination of the residuals clearly showed the existence of two sets of observations, marked A and B, and led also to the rejection of the returns from Nelson and Gisborne, both of which seemed a little doubtful on other grounds.

The observations in A and B were then examined separately.

Set A.—To find the origin, time at the origin, and velocity Of propagation (x, y, z, t, v) there are only five sets of data, and we gain nothing by forming normal equations; indeed, in this case the graphic method of circles is as effective as the method of co-ordinates, and much more manageable. Both were used, in order to have a check on the result. They give the epicentrum R; time at the origin, 2 hours 40 minutes 11 seconds ± 5 seconds; velocity of propagation, 85–87 miles per minute (7,440 ft.–7,656 ft. per second, or 227,000 cm:—233,500 cm. per second.) The value of z, or depth of origin, is swamped by the high velocity; and we have no means bi finding it from the available observations. R is fifty-one miles from Wanganui and seventy miles from Wellington.

Set B.—There being only four places of observation, with the same time at each, the method of straight lines or that of circles can be used. Either gives S for the epicentrum, taking into consideration that the time at Wakapuaka is

– 588 –

given as 2.42 barely, and that the previous rumbling lasted for 12 seconds; so that it was possibly at least 6 seconds earlier. This seems to prevent one being influenced by the times at Nelson and Bull's, which would be satisfied more nearly by an epicentrum nearer R. S is sixty miles from Wanganui and sixty-six miles from Wellington.

The question remaining to be settled is the relation between the results obtained from the sets of observations A and B. They are certainly the observations of two different phases. Are they simply the observations of two different maxima of disturbance, or do those of one set belong to the normal vibrations and those of the other to the transverse vibrations? If the B times give the slower (probably transverse) vibrations, starting from the origin (R or S) at the same time as the quicker (normal) vibrations of A, then their velocity of propagation was thirty to thirty-three miles per minute (2,640 ft.–2,904ft., or 88,500 cm.-88,600 cm., per second). The difference in the velocities of propagation is very marked; but it must be remembered that the theoretical value of the transit velocity of transverse vibrations at the origin is less than that of the normal vibrations, and that the amount of loss of the former is greatly affected by the question whether the vibrations are along or across the planes of lamination, while the loss of velocity of the normal vibrations is scarcely affected by this at all.

The calculated velocity for normal waves in granite (the elasticity being that determined by Gray and Milne from their experiments on Japanese rocks) is, according to them, 395,000cm. per second; the velocity of the transverse waves 219,000 cm. per second, or eighty-two miles per minute nearly.* The latter depends on the rigidity modulus and the density only; and Gray and Milne's figure for the former is very low (128 × 106 grammes per square centimetre). Lord Kelvin's figures (157 × 106) gives, I find, about 258,000 cm. per second, or ninety-six miles per minute nearly.

The actual velocity of the waves (B) when observed is about one-third of the theoretical value for transverse waves, and this is not an improbable value; often as much as eleven-twelfths of the theoretical value is lost in the initial stages; after that the velocity remains tolerably uniform. (See Dutton, “Charleston Earthquake”; and Ibbetson, “Elastic Solids”).

The explanation, then, that the A observations are those of normal vibrations and B those of transverse vibrations is quite a plausible one, if supported by other considerations. No single velocity that can be assumed will agree with both A

[Footnote] * Quart. Journ. Geol. Soc., vol. 39, p. 139.

– 589 –

and B, even if the times are those of the maxima farthest apart in time, unless we assume a much greater duration of the earthquake than the recorded facts allow us to assume, or unless we suppose the possibility of errors of observation in no way warranted by the character of the returns, which bear internal marks of accuracy, and show in each class a remarkable agreement, Gisborne and Grey mouth being the only places where the error amounts to half a minute (and these were not used in the calculation). If the velocities are different the times cannot well be times relating to different-maxima of the same series of waves, for the velocity of propagation of the waves of any series will be constant at the origin.* The position of S is not affected by the question just discussed.

The velocity—-eighty-five miles per minute, or, say, 7,500 ft. per second (for the A places)—is the highest yet found for any New Zealand earthquake, the velocity in the Nelson, 1893—about fifty miles per minute, or 4,400 ft. per second—being the only one at all near it.

The maximum intensity measured by the acceleration of the earth particle at the surface at Wanganui was about 800 mm. to 900 mm. per second—say, one-twelfth of that due to gravity—according to Holden's mechanical equivalents of the Rossi-Forel scale.

The amplitude or the period we have no means of measuring, and consequently cannot find the wave-length.

I have drawn the probable isoseismal for the intensity vi. on the Rossi-Forel scale, and partly those for vii. and viii.

On the map I have marked the positions of the epicentra of all the earthquakes of this district for which the data were full and accurate enough for mathematical purposes. They are as follows (approximately):—

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

F. Feb. 20, 1890 Trans. A.A.A. S., 1891, “New Zealand Earthquakes.”
E. Dec. 4, 1891 Trans. N.Z. Inst., xxv., p. 362.
A. July 5,1891 Trans. N.Z. Inst., xxiv., p. 577.
P. Aug. 20, 1891
Q. May 18, 1893
R. or S. Dec. 8, 1897 Now first published.
T. May 16, 1898
V. July 8, 1898

A large number of other shocks agree generally, as far as

[Footnote] * Ibbetson, Elasticity Arts,” 268 and 276.

[Footnote] † Holden, American Journal of Science, June, 1888; Hogben, “Earthquake Intensity in Australasia” (Trans. A.A.A.S., 1893).

– 590 –

the indications given, with the same origin. There is no other shock since the present system of records was begun (i.e., 1890) the data of which are sufficient to determine the epicentrum exactly—in other words, the epicentra given include all those ascertainable at present.

A circle of seven miles radius would include all the points except A, and I think I may justly claim to have found, within the limits of error of our present observations, the region whence the great majority of the Cook Strait earthquakes proceed.

The new instruments about to be set up in the colony will, though not primarily intended for that purpose, no doubt give us information that will help us to determine other elements, especially the period of the vibrations, and, if set up in the proper plane, the amplitude of the vibrations. Hence we can calculate the intensity of shock and the wave-length, and may be led to reasonable speculation on the nature of the underlying rocks between the origin and the place of observation.