TION) into high-speed media, propagate in them, leave them in the same manner and reach the surface. (See EARTH, fig. 2). The depths and media reached depend on the distance between shot point and receiving points; the first impulses or "breaks" in a seismogram (see fig. 2) are caused by waves having travelled through the deepest high-speed layer within range. Reflection im pulses are due to the reflection of seismic waves on media of greater elasticity and arrive later than the first refraction im pulses. Their time of arrival depends on the average velocity be tween surface and reflecting bed and the distance between shot point and receiving points. (See Echoes and Echo Depth Sound ing, in SOUND.) In refraction methods, depth reached is a frac tion of distance between shot and receiving points; in reflection methods, a multiple thereof.
The most extensive application of refraction methods has been made in the location of Gulf Coast salt domes. In salt, the ve locity of (longitudinal) elastic waves is of the order of 4•7 to 5•2 kilometres per second; in the adjacent media, from 2 to 3.5 km. salt domes thus bring about an appreciable reduction in travel time compared with normal ground. Refraction methods are also used for depth determinations of shallow layers which generally possess great velocity contrasts. In mining and civil engineering applications, overburden velocities may range from several hundred to one thousand metres per second, while bed rock, if composed of igneous or metamorphic formations, may be 5-6000 m. Velocities in sedimentary rocks range from about I000 for the least consolidated to 400o metres for the more consolidated formations; limestone is of the order of m. If consolidated beds exist in sedimentary col umns, reflections are obtained from them. Velocities of less than that of sound in air are observed in unconsolidated near-surface formations (weathered layer). Uniform increase of compaction and velocity with depth produces curved rays (see fig. 2).
Seismic exploration equipment is elaborate (see fig. 2). For maximum utilization of energy, shot-holes are drilled (in deep refraction and reflection work) with truck-mounted rotaries. From 6 to I2 (and more, in series and series-multiple connection) receivers are employed which convert the mechanical ground vi brations into electrical energy in much the same manner as mi crophones or phonograph pickups. The electrical impulses are amplified and recorded by oscillograph-galvanometers on rapidly moving paper on which time lines are projected at intervals of hundredths of seconds. The instant of the shot is transferred elec trically to the record so that the time elapsed between the arrival of impulses and firing of the shot may be read with an accuracy of I/I000 of a second and be used in the construction of the time distance ("travel-time") curve. Amplifiers and camera are mounted in a recording truck. For shallow refraction applica tions, the equipment is less elaborate but follows the same gen eral arrangement. Reflection methods are applicable to depths between Soo and 2o,000ft.; refraction methods to depths of sev eral feet to 6,000 feet.
Interpretation of refraction records is based on the "travel time curve." Contrarily to the travel-time curves shown in fig. 3 of the article EARTHQUAKE for great crustal depths with uniform velocity increase, travel-time curves in seismic prospecting con sist mostly of straight portions with well defined breaks. For horizontal stratification, the slopes of these portions indicate true wave velocities in the layers traversed. Considering two layers, one with low velocity above, the other with greater ve locity below, the waves will travel through the upper layer but will also be refracted into the lower; the latter will follow the principle of Fermat which states that the waves follow such path as will permit them to arrive at their destination in the shortest possible time. This means that they enter the lower me dium by the "critical" refraction angle whose sine is equal to the velocity ratio. A break (or intercept) will occur in the travel-time curve when the direct waves through the upper medium arrive simultaneously with those through the lower medium. Setting up the conditions for equality of their two travel-times it is possible to express the "intercept" distance as function of depth and velocities in the two layers. This applies to any number of horizontal layers. For inclined beds, not the true but apparent velocities are indicated by the slopes of the travel-time curve and are functions of true velocity ratio and dip. By shooting pro files in opposite directions, two travel-time curves are obtained which by combination yield depth under the shot point, dip and true velocities. For great velocity contrasts it is permissible to substitute vertical paths for the classical refraction paths ; for uniform velocity increase, straight path interpretation is replaced by "curved ray" interpretation. Depth calculation in reflection shooting is much simpler than in refraction shooting. For hori zontal stratification or low dip it is sufficient to time reflection impulses at a mean distance, to apply a correction for the delay of the echo in the weathered layer (determined from separate refraction or "weathering" shots), to reduce the time thus ob tained to a regional datum and to calculate the depth of a reflect ing bed below it by assuming a straight wave path and an average velocity. The latter are obtained by shooting near a well in the area and measuring travel-times to a detector lowered to the depths corresponding to the reflecting beds. To determine dip, receivers may be set up in two opposite directions from the shot point whereby up-and-down-dip travel-times and up-and-down-dip time-gradients are obtained ; to obtain dip and strike two such profiles are laid out at right angles (or another given angle) to one another. Reflection shooting is at present the most accurate method of geophysical oil exploration ; depth determinations can be made with an accuracy of better than 2 per cent.
The self-potential method makes use of the fact that many (sulphide) ore bodies when in contact with solutions of a dif ferent chemical nature at their upper and lower ends, act as a battery and are therefore surrounded by an electrical field. Its distribution may be measured by surveying lines of equal po tential with non-polarizable electrodes and a milliammeter, or by measuring potential differences between successive profile stations with a potentiometer (see INSTRUMENTS, ELECTRICAL). The upper end of an ore body is indicated by a negative potential cen tre, usually of the order of several hundred millivolts. In all other electrical methods, electrical energy is supplied to the ground, either galvanically or inductively. Galvanic supply has the advantage that lower frequencies of greater depth penetration can be used, and that the potential distribution between the power electrodes can be investigated by search probes making direct contact with the ground. Potential methods have the fur ther advantage of being sensitive to small differences in conduc tivity. Direct current is usable but the general tendency in po tential studies is now toward A.C. of low frequency or to im pulses. In the equipotential-line method, power is supplied to two point- or line-electrodes ; the equipotential lines are traced by two search electrodes connected through an audio-amplifier with head-phones in the output ; one of the probes is held stationary and the other is moved until the sound disappears. Good con ductors beneath force the equipotential lines apart, poor con ductors attract them. Interpretation is largely qualitative and empirical. More quantitative results are obtained by measuring potential differences between successive stations on profiles, in reference to voltage and phase in the supply circuit. For this purpose an A.C. compensator is used (see fig. 8). A third pro cedure is to connect three ground stakes to a bridge network (ratiometer, fig. 7), giving the ratio of voltage differences be tween the adjacent ground intervals as well as their phase differ ence. This is known as the "potential-drop-ratio" method; it is applicable to the detection of horizontal variations in conductivity (location of faults, ore bodies) ; for horizontally stratified ground it gives depths to formation boundaries ; ratio peaks occur at centre-stake distances from the power electrode equal to about times their depth.
A variety of the potential method now in wide-spread applica tion is known as the "resistivity" method. Four electrodes are driven into the ground ; the external pair is supplied with current from dry batteries which is read on a milliammeter while the in ternal pair is connected to a potentiometer to determine the voltage difference between them (fig. 6) . To avoid polarization effects, a commutator is inserted in the circuit so that the current through the ground changes in direction every half cycle but re tains the same direction through the instrument. This is known as the Wenner-Gish-Rooney arrangement. The ratio of voltage and current, multiplied by a factor depending on electrode spac ing, gives the true resistivity for homogeneous ground, or the so called "apparent resistivity" for non-homogeneous ground. If the electrode arrangement is moved over the ground with constant spacing, the variation of resistivity in horizontal direction is ob tained; this is known as "resistivity mapping." A variant of re sistivity mapping is the "electrical coring" method applied in wells (see below). By leaving the centre of the arrangement con stant and expanding the spacing, the change of resistivity with spacing is equivalent to its change with depth; this is called "resistivity sounding." Interpretation of results obtained in both cases is mostly qualitative ; if not too many media are involved, depths may be calculated from the apparent resistivity by apply ing the theory of images. The potential distribution at the sur face is then obtained by reflecting both current source and sink on whatever vertical or horizontal formation boundaries are as sumed to exist. In the potential-drop-ratio method, there is a more direct relation between distance at which indications are ob tained and depth, than in the resistivity method (see above) ; the oretical calculations follow the same procedure as in the resistivity methods.
For audio-frequency fields between the primary electrodes, their distribution and distortion by subsurface conductors may also be determined by electromagnetic measurements. The re ception equipment in this case consists of a coil connected to a bridge network known as compensator, an amplifier, and a pair of head-phones (see fig. 8). This gives the field in terms of amplitude and phase of the primary current. Generally the hori zontal components parallel and at right angles to the supposed strike as well as the vertical component are determined. Above a linear current concentration in a subsurface conductor the horizontal component at right angles to the strike is a maximum and the vertical component is zero ; the distance of the minimum and maximum in the latter away from the conductor is equal to twice its depth. Instead of determining fields in terms of generator current, relative measurements may be made by "ratio meters," comparing the fields induced in two identical coils in respect to intensity and phase (see fig. 7). Reception devices in inductive methods are virtually the same as in electromagnetic methods, the primary field being supplied to the ground by in sulated loops. This will induce eddy currents to flow along the edges of subsurface ore bodies; the depths of these current con centrations may be determined approximately from horizontal and vertical intensity measurements as mentioned above. Induc tive methods have also been used in structural oil exploration; depths of conductive subsurface beds may be determined from measurements of the horizontal intensity ; anomalies in the latter may be assumed to be due to the images of the primary cable produced by reflection on conductive layers at twice their depth.
Radio or high-frequency measurements are not in extensive use in geophysical prospecting owing to their lack of depth pene tration and tendency to emphasize non-commercial conductors. An exception are devices based on the principle of the induction balance, also known as "treasure finders," intended for the loca tion of pipes and other buried metallic objects at shallow depths.
For each major geophysical method or its subdivisions, the tabulation of Table I indicates the principal field (oil, mining, or engineering) and the more important geologic problems to which a method is adapted. Applications in oil are (with the ex ception of electrical coring) of an indirect nature; the object is to locate structural traps likely to contain oil. Such traps are the tops of anticlines and domes which may be mapped by seismic reflection, gravimeter, torsion-balance, electrical potential and inductive methods and also by magnetic surveying if magnetic formations are interbedded, or if their cores contain igneous or metamorphic rocks. Other important traps are the flanks of salt domes ; the latter have been located chiefly by seismic refraction, torsion balances and gravimeters (Gulf Coast) ; for mapping formations on the flanks or above a dome, electrical inductive and seismic reflection methods are applied. Faults, constituting an other structural oil trap, may be mapped by seismic reflection, torsion-balance, electrical potential and inductive, and magnetic methods. Oil occurrences associated with igneous intrusions may likewise be located (indirectly) with magnetic methods. An im portant application of geophysical methods in oil exploration is electrical coring. This is a modification of the resistivity method (see above) ; electrodes are lowered into a (uncased) well and a record is obtained showing variations in resistivity of formations traversed ; it not only is less expensive than mechanical coring but reveals occurrence and depth of oil and water-bearing strata.
Applications in mining are of both an indirect and direct na ture. In the former, geologic structure giving a clue to the oc currence of ore is located; examples are : Magnetic surveys to determine the suboutcrop of magnetic formations bearing a known stratigraphic relation to deposits of gold (Witwaters rand), copper (Lake Superior), or other non-magnetic minerals; seismic refraction and gravity surveys of gold-placer channels (Australia) ; magnetic surveys tracing magnetic black sands as sociated with placer gold (Alaska, British Columbia) ; seismic and gravity surveys for locating salt domes whose cap rock may contain sulphur (Gulf Coast) . The second group of mining ap plications involves a direct location of the sought minerals. This is true particularly for all types of sulphides (chiefly copper, nickel, and lead, with the exception of sphalerite) which can be located with the self-potential, the potential, electromagnetic, in ductive and sometimes (pyrrhotite) the magnetic method, and the iron ores (chiefly magnetite) which are most readily located with magnetic methods. In non-metallic mining, applications are both direct and indirect (determination of structure) ; they are very diversified, each mineral or type of deposit presenting virtu ally its own problem. Suffice it to say that the methods most uni versally applicable in non-metallics are electrical potential for the direct, and electrical potential, seismic, and gravitational methods for the indirect applications.
In the field of engineering and engineering geology, geophysical prospecting has many diversified applications. The most impor tant is probably the investigation of foundation conditions. Sites for proposed dams and tunnels, railroads, highways and bridges may be examined most conveniently with the electrical potential and the seismic refraction methods. If the latter are supple mented by vibrator tests it is frequently possible to derive from both surveys such elastic properties as are required in engineering design. Vibrator measurements likewise furnish data determin ing the destructiveness of artificial vibrations and earthquakes for foundation and proposed structure. Considerable expense for haulage may often be saved by applying geophysical methods to the location of construction materials for dams, railroads, and highways, the preferred methods being electrical potential and seismic refraction. Water for irrigation and domestic use may be located geophysically; the most reliable methods are, at pres ent, the potential electric methods - The problem is not a simple one as water is not located directly but stratigraphically and may impart either a greater or a reduced conductivity to the impreg nated layers. Other applications of geophysics in engineering in clude corrosion surveys (by self-potential and resistivity meth ods) and the detection of leaks from water and gas pipes (acous tic and gas detection methods). Uses of geophysics in military engineering are closely related to those in civil engineering (loca tion of water; assistance in drainage problems; location of con struction materials; foundation tests for roads, fortifications, and shelters) and are supplemented by various acoustic, thermal, and electrical detection methods of enemy activities.
(h) Geophysical Education; Societies; Literature.—Geo physical exploration may be considered as a highly specialized field of engineering which requires not only a thorough knowledge of physics, mathematics, and geology, but extensive field experi ence. Owing to this combination of requirements of a somewhat conflicting character, the tendency of many engaged in this field has been to specialize either in the physical or geologic phase of geophysical exploration. Special courses in all or several phases of this subject are given at the Colorado School of Mines, the Massachusetts Institute of Technology, the California Institute of Technology, the Imperial College of Science and Technology, The Technische Hochschule in Charlottenburg, and the University of Strasbourg, the latter giving a degree of Ingenieur Geo physicien. The above list is not complete; several universities and technical schools offer, in addition, general courses in geo physical prospecting in their geological departments.
Interest in research pertaining to geophysical exploration has been largely divided up among physical and geological societies; as yet there are but few societies organized specially for this particular profession. The most important are : The Society of Exploration Geophysicists and the Deutsche Geophysikalische Gesellschaft; the Geophysical Committee of the American Insti tute of Mining and Metallurgical Engineers also belongs in this group. Articles on geophysical exploration are found in a great variety of scientific, technical and trade journals in many lan guages, notably English, German, Russian, and French. Journals devoted to geophysical exploration alone are few: the journal Geophysics issued by the Society of Exploration Geophysicists, the Zeitschrift der Deutschen Geophysikalischen Gesellschaft (which also contains articles on geophysical science), the Tech nical Publications of the American Institute of Mining and Metal lurgical Engineers, and Gerland's Beitraege zur angewandten Geophysik. The international literature on geophysical prospect ing has been reviewed semiannually, since 1928, in the geophysi cal section of the Annotated Bibliography of Economic Geology. German reviews are contained in the "Geophysikalische Berichte," appended to almost every number of the Zeitschrift der Deut schen Geophysikalischen Gesellschaft.
