THE COLLIERY GUARDIAN AND JOURNAL OF THE COAL AND IRON TRADES. Vol. CXV. FRIDAY, JUNE 14, 1918. No. 2998. Terrestrial Magnetism in Relation to Mine Surveying.* By C. CHREE, Sc.D., LL.D , Superintendent of Kew Observatory. (1) In looking at the advance of scientific knowledge from an historical point of view, it will be found that what has led to the most fundamental discoveries has been intellectual curiosity, not the direct pursuit of any immediate practical end. Thus there is reason to think that the pursuit of knowledge for its own sake is the real royal road to progress. On the other hand, the exchange of ideas between the practical and the theoretical man may be for the benefit of both. There are a considerable number of observatories— some forty or fifty, in fact—which take magnetic observations, but most fulfil a variety of purposes. Some are primarily astronomical, whilst others are meteorological. The majority are somewhat poorly equipped, and do not possess a staff suitable for the prosecution of research. In this country the work which has been of the greatest practical importance, namely, the magnetic survey of the British Isles for the epoch 1891, was the enterprise of two men-—Sir A. Rucker and Sir E. Thorpe—neither of whom, so far as the writer is aware, was ever attached to a magnetic observatory. The ordinary practical unit of magnetic force, ly, is simply an abbreviation for 0’00001 C.G.S. mag- netic unit. The mean values of the horizontal (H) and vertical (V) components of magnetic force at Kew Observatory for 1916, the latest year as yet fully worked up, were respectively 18457y and 43395 y. Thus in either case ly represents but a small fraction of the force. It is usual to publish force components to the nearest ly, that being about as near as it is possible to measure the changes shown by the ordinary magnetogram; but it cannot, the writer thinks, be claimed that the absolute value of even H—the most easily measured component—is known accurately to ly. It is, in fact, very doubtful whether any exist- ing magnetometer can be relied on to give an in- variable standard from one year to the next. Differences exist between the values of H obtained , with different instruments, and there is as yet no satisfactory means of determining where the fault lies. (2) The magnetic element as to which there is least difference of opinion is, fortunately for engineers, the declination (D). If any systematic difference exceed- ing 1' or 2' presents itself between the values of D given by two magnetometers, it probably repre- sents the presence of some magnetic material, or the existence of a plate of glass the faces of which are not plane parallel. Elimination of any error arising from non-coincidence of the magnetic and optical axes is secured by actually rotating the magnet through 180 degs. in its stirrup, as in the Cooke pattern, or by providing the suspension with two shanks and suspending the magnet first by the one and then by the other, as in the Dover and Elliott patterns. The same degree of concordance cannot reasonably be expected from the pivoted magnets of ordinary theodolites. Very carefully constructed compasses, such as are used for some purposes by the Carnegie Institution of Washington, U.S.A., give, it is true, remarkably consistent results in skilled hands—at least when new. The degree of accuracy obtainable in observatory work, and the consistency between different instru- ments, may be illustrated by the results of a com- parison made at Kew Observatory in 1915 between a magnetometer belonging to the Carnegie Institution of Washington and the standard instrument of the Observatory. The differences were as follow : — Difference: —0'’5 —0''2 —0'T 0'’0 +0'T +0'’2 +0''3 +0'*4 Number of occurrences :1 1 155 2 2 1 The mean difference is + OZ’O6, and the standard derivation—that is, square root of mean squares of differences—is OZ’2O. The result is consistent with the view that under favourable conditions, namely, with magnetometers indoors, on substantial piers, in a good light, at a magnetically quiet time, the observational errors to be expected in the average observation are of the order 0z-2, and errors exceeding O'-5 should be rare. To obtain accuracy of this order, the suspen- sion, whether silk or metal, should be as fine as is consistent with safety, and the torsion must be re- moved as far as possible before the observation by means of the plummet and torsion head. An observatory has a fixed mark, the azimuth of which is accurately known, so that results are free from what is a prin- cipal source of error in field observations, namely, uncertainty as to the geographical meridian. In the field, especially when the observer’s fingers are cold, inverting the magnet may involve risk to a fine sus- pension. If proper care is exercised, prospecting work in the field, such as would be required to ascer- tain the characteristics of a disturbed locality, could obviously be carried out without inverting the magnet * From paper read before the Institution of ~ Mining Engineers. every time. Occasionally, of course, a complete obser- vation should be taken with scale erect and inverted, to provide against some possible change. (3) One great obstacle to the use of the declination magnet for prospecting work is the necessity of knowing the geographical meridian, or the exact bearing of some visible object. If one has a good chronometer, the error of which is exactly known, one can indeed by means of a sun’s observation, after the method followed by Rucker and Thorpe, deter- mine the geographical meridian to within about 1'; but in this country, especially in winter, the sun may be invisible on many successive days, and astro- nomical observations of the pole star—the most satis- factory method of all—may be equally impossible. On the other hand, it may be taken as practically an invariable rule that disturbance in D is accompanied by disturbance in H and in I (inclination or dip). Observations of I and H, if more laborious than those of D, are not dependent on the visibility of any heavenly body, and so can be carried out at any time. Thus the writer cannot but think that it might be well if engineers showed more interest in these elements than they have usually done. Dip observa- tions may be made either with a dip circle or a dip Mid* CX Noon la% Mi£’. Midi 6h Noon -5 Summer Winter 19)3 Disturbed Day Quiet Day; Summer •35* ■io’ ■IS’ HO* HO HO* M5' •io’ .io' -IS1 •io •25’ Summer I II I I I Winter Quiet Days Summ.e Oh J 4 e « 10 Naon 14 If Kew Observatory •Kf "♦k Oh 2 + 6 S IO tiooa H 16 1« 20 22 2+h Antarctic Diurnal Inequality of Declination. inductor. The latter instrument has the higher accu- racy, but it involves the use of a sensitive gal- vanometer, and so is 'more suitable for observatory than field work, although the Carnegie Institution has of late years made considerable use of it in the field and even at sea. With good English dip circles—at least, when the needles are new—differences ffom the Kew standard seldom exceed 2Z. A complete dip observation with two needles, in- cluding reversal of the magnetisation, two readings being taken of each end of the needle in each posi- tion, occupies about 45 minutes. This supposes the magnetic meridian already determined. At an ob- servatory occasional redetermination of the meridian suffices, as accuracy to 30' is good enough. But in the field it has to be done on each occasion, and the operation takes some ten minutes. If one were in- vestigating a moderately disturbed area, it would be unnecessary to reverse the magnetism on each occa- sion, and observations with a single needle might suffice—-at least, for preliminary work. The necessity for reversing the magnetisation in a complete ob- servation is due to the fact that the centre of gravity of the needle departs more or less from the axis of support, and there is thus a gravitational component tending to increase or reduce the dip, according as the centre of gravity lies on the same side of the axis as the dipping end of the needle, or on the opposite side. So long as the position of the centre of gravity and the strength of the needle are invariable, the difference between the readings obtained at a given place with pole A and pole B dipping remains constant, and if half this difference be applied as a correction to the result obtained with either pole dipping, the true dip is obtained. (4) The determination of H is more complicated. It involves two main operations, viz., taking the time of vibration of the collimator magnet, which supplies a value for mH, where m is the moment of the magnet, and deflecting an auxiliary magnet, which gives a value for m/H. Combining the two results, H is determined and also in. The first difficulty is that the magnetic moment falls as the temperature rises : thus a correction is needed to reduce the vibration and deflection results to a standard temperature. To this end the temperature of the magnet must be accu- rately known during both operations—not an easy matter in field work when the temperature of the air is changing rapidly. Temperature uncertainties cause errors of fluctuating sign, and for observatory work are much less important than several others. One of these is that through temporary induction the magnet is stronger during the vibration experiment, when its inclination to the magnetic meridian is always small, than during the deflection experiment, when its inclination to the meridian is large. An error in the induction coefficient—a quantity the deter- mination of which presents difficulties—causes an error in H in one definite direction. A similar remark applies to the moment of inertia of the magnet. What is perhaps the most troublesome uncertainty of all remains to be mentioned. The force exerted by the collimator magnet during the deflection experi- ment on the auxiliary magnet, of moment m', is given by the formula— 2 m/ni'r‘3 (1 Pr ’2 -|- Qr~4) where r is the distance between the centres of the magnets, and P and Q the so-called “ distribution constants.” It used to be assumed that the term involving Q is negligible when observations are taken as usual at 30 cm. and 40 cm, and that Iq-Pr-2 may be replaced by 1/(1 —Pv2). Unfortunately, neither assumption is usually justified. We cannot, in short, z hope to obtain correct absolute values of H unless observations are taken at three distances instead of two so as to eliminate G as well as P, or else increase the deflection distances to such an extent that the Q term becomes really negligible. The latter alternative might be adopted if the strength of the collimating magnet could be very largely increased without in- creasing its size, but at present we seem driven to the former alternative. While a complete observation of H is laborious, occupying at least an hour, more often an hour and a half, it is possible with much less expenditure of time to obtain results of high accuracy in exploring a disturbed area, where the occupation of a number of stations in a limited time is expedient. This arises from the fact that if a collimator magnet has been properly seasoned, and is carefully handled, the magnetic moment changes very slowly. (5) It may seem that undue space has been allotted to ways of simplifying the use of magnetic instru- ments for investigating the existence or general characteristics of local disturbances. The considerable number of enquiries, however, which the writer has received for values of magnetic declination from people who seem to think their approximate geo- graphical position is all one requires to know in order to give a definite answer, has impressed him with the belief that the probabilities of a sensible departure from the value suggested by the smooth isogonals in ordinary magnetic charts are in general imperfectly appreciated. Engineers would be well advised to con- sult, for example, maps Nos. 5 and 9 of Rucker and Thorpe’s survey,* showing unsmoothed isogonal lines and declination disturbances. They should also re- member that with stations so widely separated as Rucker and Thorpe’s, anomalies confined to small areas have little chance of detection. The following two examples of what may occur in limited areas must suffice. They are taken from observations at sea in water from 80 to 100 ft. deep, and so are not purely superficial effects, such as may be observed within a few feet of a comparatively small mass of basaltic rock*. A minute survey by H.M.S. “ Penguin,” near Port Walcott, off the north-west Australian coast, over an area 3 miles long by 1| miles at its widest, dis- cussed by Captain E. W. Creak, f R.N., F.R.S., gave values of D varying from 26 degs. W. to 56 degs. E., a range of 82 degs. Again, at East Loch Roag, Lewis, Hebrides, a survey by H.M.S. “Research,” discussed by Admiral A. M. Field,! R. N., F.R.S., showed within a small area a range of 22 degs. in D. Anomalies apparently much larger than even that at Port Walcott have been described in Russia. (6) At a fixed observatory the absolute observations are used to standardise the curves. At a first order station, such as Eskdalemuir, it is usual to measure all the curves at one-hour intervals, the mean of the twenty-four hourly measurements being accepted as the mean for the day. The mean for the month represents the mean derived from the included days, and the mean for the year the mean * Roy. Soc., Phil. Trans. A., 1896, vol. clxxxviii. t Roy. Soc., Phil. Trans. A., 1895, vol. clxxxvii., p. 345. + Roy. Soc’, Proc. A., 1905, vol. Ixxvi., p. 181.