If these predictions based on past performance are correct, there will be some impressive seismic events during the IGY. In anticipation of them a number of amateurs have built seismographs to observe the show from the sidelines. One of the best equipped is A. E. Banks, the proprietor of a stationery store in Santa Barbara, Calif. Banks has spent most of his free time during the past two years learning how to build simple versions of conventional seismographs and developing some not found in the textbooks.

Because two or three of Banks's instruments are always on the job, he has not missed a single earthquake of consequence during the past year. The most interesting sections of his record of the big Aleutian quake last March, which pounded beaches in Hawaii with five-foot waves, appear at the left. The complete record of the quake covers four feet of record paper.

The first section of the record shows the faint and ever-present vibrations known as microseisms, which seem to arise from variations in atmospheric pressure, the pounding of surf and other minor sources of seismic energy. The next section shows the beginning of thc quake, which is announced by the first and fleetest of three characteristic types of earthquake waves: the primary or "P" wave. The P wave vibrates in the direction of its travel like an opening and closing coil spring. It travels through the earth at an average speed of about five miles per second. The third section shows the arrival of the secondary or "S" wave. This wave travels at an average velocity of about three miles per second; it is often called the "shake" wave because it vibrates at right angles to the direction of its travel like an oscillating violin string. The trace made by S waves on the seismogram is not so jagged as that of P waves; moreover, the height or amplitude of the two are rarely equal. Hence P and S waves can usually be distinguished from each other without difficulty.

The fourth section of Banks's seismogram shows the final major wave of the Aleutian quake, the long or "L" wave. The L wave always makes a big, dramatic trace and, unlike P and S waves, travels largely through the earth's outer layers. P and S waves are accordingly most useful in investigations of the structure of the earth at great depths, and L waves are most useful in studies of its outer layers. The velocity of L waves is influenced by the character of the terrain but averages about 2 1/2 miles per second.

The seismograph senses earth vibrations by taking advantage of Newton's first law, which states in effect that a body freely suspended and at rest in space will remain at rest until something gives it a push. If an experimenter could float in still air, his position would not be affected by an earthquake, however violent; if a sheet of paper were fastened to the ground, he could make a seismogram simply by reaching down and allowing a pencil to move across it during the quake. Because gravity rules out such ideal suspensions, the designers of seismographs settle for an elastic suspension such as a pendulum or a weight on a spring. A pendulum may be considered an elastic suspension because it behaves as though a spring were stretched between the pendulum bob and the center of the earth.

Galileo was the first to observe two interesting properties of such suspensions. First, they oscillate or swing at a characteristic rate. For example, when an iron ball suspended from a coil spring is pulled down and then released, it bobs Up and down at a rate which depends on the size of the ball and stiffness of the spring. Second, when a pendulum is set in motion by a push it continues to oscillate for a time, the duration of which depends on how soon the energy imparted by the push is dissipated. This is why pendulums have historically been used to measure time. But if this property is prized by clockmakers, it is the bane of those who make seismographs.

The ideal seismograph would respond equally to waves of all frequencies. The seismograph designer, however, must make do with a device which favors some frequencies and discriminates against others. Consider a simple seismometer consisting of a heavy iron ball suspended from a spring attached to a framework which rests on the ground. Let the ball and spring be chosen so that when the ball is disturbed, it bobs up and down once a second. Assume the ball is at rest when an earthquake strikes. If the quake shakes the frame faster than one oscillation a second, the ball will remain at rest while the spring expands and contracts. When the frequency of the shake corresponds to that of the ball and spring, the ball absorbs energy from each oscillation. It starts to bob up and down; the distance of its motion increases with each added oscillation. When the frame oscillates less than once a second, both ball and spring move up and down with it. Earth vibrations are detected by observing the motion of a pendulum bob with respect to its supporting frame. Such motion is accordingly apparent when the frequency of the frame is higher than that of the pendulum. It is almost totally absent when the frequency of the frame is lower.

The most desirable seismograph would thus be one with a natural period of oscillation longer than any earthquake wave. It would require a pendulum capable of swinging only once in several minutes; ideally the pendulum should oscillate once in several hours. Such leisurely pendulums are either difficult to make or impractically large. Moreover, the most interesting earth-: quake waves have frequencies in the range of about three oscillations per second to one per minute. Seismological observatories usually record only the lower end of this range, and employ separate instruments tuned to oscillate at two, eight and 15 seconds. The two-second seismometer detects P waves and -discriminates against microseisms, which lie in the two- to nine-second range. The eight-second instrument responds strongly to both P waves and microseisms. The 15-second instrument records these plus S and L waves. An instrument of this period is difflcult to adjust; thus many amateur seismologists who have only one instrument compromise on a period of 10 seconds.

The second property of pendulums observed by Galileo–their tendency to continue swinging once they are set in motion–is controlled with friction or electrical resistance. This increases the rate at which unwanted energy is dissipated and so counteracts the pendulum's tendency to oscillate excessively at its natural period. It is often accomplished by immersing a vane attached to the bottom of the pendulum bob in a bath of oil. The. same result can be achieved by making the vane, or even the bob, out of copper and suspending it between the poles of a strong magnet. Eddy currents induced in the copper are transformed into heat by the resistance of the metal. Ideally the bob should be damped just enough for the pendulum to come to rest promptly. If the pendulum is damped too much, it will be sluggish and require more than one swing to reach equilibrium. If it is damped too Iittle, it will swing several times before coming to rest. The best compromise, called "critical" damping, lies between 5these extremes and is determined experimentally for each instrument.

Unless they are in the immediate vicinity, earthquakes move the supporting frame of the pendulum only slightly–a thousandth of an inch or less. For the pendulum to make a useful record this motion must be amplified. The earliest seismographs were equipped with a system of amplifying levers, some of whicll turned on jeweled pivots. The short end of the first lever was coupled to the pendulum bob, the frame of the instrument providing the fulcrum. The long end of this lever drove anotller lever fitted with a stylus. The stylus made a trace on a cylinder of smoked paper rotated by clockwork. The mechanism required a heavy pendulum bob so that enough energy could be transmitted from the frame to the stylus without setting the bob in motion. The modern seismograph usually employs a combination of mechanical, optical and electrical amplifying elements; they are arranged in such a way that the seismic energy merely controls the movement of an electrically powered recording pen.

In essence, then, a seismograph consists of an elastically suspended weight which acts as the sensing element, a device to amplify the motion of the weight with respect to its supporting framework, and n pen or stylus to record the amplified motion. A fully equipped seismographic station employs at least nine sensing elements. Three sense vertical motions and are tuned to periods of about two, eight and 15 seconds. The second set of three, similarly tuned, responds to horizontal oscillations in the north-south direction; the third set, to oscillations in the east-west direction.

Banks's example testifies that learning how to design, build and adjust seismographs can become a fascinating avocation. He writes: "I felt my first earthquake in 1914 while sitting in an Iowa farmhouse. Although that experience touched off my interest in seismology, it was not until September of 19S5 that I attempted to build an instrument. I had had no previous experience and so had to start at the bottom and go as far as I could with the information I could dig, out of books. Most of them were addressed to the specialist. I soon found that even all amateur seismologist should have majored in physics, geophysics, mechanical engineering and meteorology. Later I discovered that he must also be able to dig a ditch and mix a batch of concrete. A facility with hand tools and electrical circuits also comes in handy. But if the amateur is willing to settle for standards somewhat lower than those of the professional seismologist, the construction and operation of a seismograph turns out to be surprisingly simple.

"A so-called 'horizontal component' seismometer, one which detects lateral oscillations, can be made in a couple of hours if one has the materials on hand– a few scraps of wood, some screws, a bolt, a bit of strap iron and a small piece of sheet metal. The device essentially consists of a small triangular wooden frame hung like a garden gate from a larger frame [see Figure 2]. The upper pivot is made from two pieces of strap iron. As shown in the illustration, one piece is screwed to the hypotenuse of the frame, filed to a point and bent over like a fishhook. The point of the hook fits into an indentation made with a centerpunch in the second piece, which is fastened to the frame. The bottom pivot consists of a lag screw filed to a point. It fits into an indentation in the end of a stud bolt. The base of the frame is fitted with three leveling screws. These are adjusted so the beam swings to equilibrium in the center of the frame. The stud bolt provides a further means of adjusting the inclination of the beam and thus of altering its natural period.

"The period can also be-adjusted over a wide range by placing a weight such as a brick on the beam. The brick selves as a pendulum bob, the period of oscillation increasing both with the weight of the bob and its distance from the pivots. The period is determined empirically by giving the pendulum a push and counting the number of seconds required for the beam to reach the limit of its swing and return.

"Finally the pendulum is fitted with a damping vane and housed to shield it from air currents. A piece of nonmagnetic sheet metal two inches square is soldered in the slot of a wood screw and screwed into the bottom of the beam. The vane is immersed in a bath of light machine-oil. The depth of immersion is adjusted until the beam makes about two full swings before coming to rest.

"A more elaborate version of essentially the same device is the Wood-Anderson torsional seismometer [see Figure 3]. Here the wooden frame is replaced by a heavy metal base and a 14-inch length of half-inch pipe. A 50-gauge wire (diameter .001 inch) is stretched between the bracket at the top of the pipe and the base. The long edge of a rectangular piece of sheet copper–about 1/2 inch wide, an inch long and 1/16 inch thick–is soldered or cemented to the middle of the wire. This serves as the pendulum. The pendulum is damped by a magnet which may be moved to adjust the degree of damping The wire passes through oil-filled grooves in the ends of two studs supported by the pipe. These studs suppress the vibration of the wire. If the instrument is exposed to large variations in temperature, a coil spring must be inserted between the wire and its attachment to the base to compensate for differences in the expansion of the wire and its pipe support. Because an ordinary coil spring will twist the wire, it is necessary to use a double spring, one half of which is coiled in one direction and the other half in the reverse direction. Lateral movements of the earth shake the attached side of the vane with respect to its free edge. The wire, which corresponds to a hinge, twists slightly as the vane swings; this will account for the name 'torsional' seismometer.

"Seismometers to sense vertical vibrations usually consist of a pendulum hinged for vertical movement and supported in horizontal equilibrium by a set of springs. One of many possible arrangements is shown in the accompanying drawing [Figure 4]. A mainspring supports the beam of the pendulum by holding it against a knife-edge. A second knife-edge is mounted on a short beam at right angles to the main beam. This knife-edge bears against an anvil linked to a second spring which opposes the force of the mainspring, depending upon the position of the pendulum and hence of the alignment of the two knife-edges. Adjustment of the second spring alters the natural period of the pendulum. Here the pendulum bob is a coil of insulated copper wire which moves between the poles of a war-surplus magnetron magnet. Current induced in the coil is dissipated by a resistance connected across the leads and so damps the motion by an amount which varies with the resistance.

"My first instrument was of the garden-gate type. The beam was a 27-inch piece of one-by-four-inch wood; bricks stacked on the beam acted as the pendulum bob. I found that the sensitivity of the instrument varies both with the angle at which the beam hangs and with the amount of damping. If the angle is too steep, the swing is imperceptible. If it is not steep enough, the beam will not return to equilibrium in the center of the supporting frame. The critical adjustment lies between these two extremes. To find it and at the same time find the proper weight and position of a bob which will oscillate at the desired rate requires some tinkering. This does not take long once you get the hang of it.

"The beam must of course be coupled to an amplifier. My first one consisted of an arrangement of levers. It worked after a fashion but was so full of friction that it would have taken a 100-pound pendulum to drive it effectively. So I built the electronic system which was described by E. W. Kammer in 'The Amateur Scientist [SCIENTIFIC AMERICAN, June, 1953].

"In this arrangement I employ a coil of insulated copper wire at the free end of the beam, which swings between the poles of a magnet. The coil serves as the pendulum bob. Flexible leads connect it to Kammer's remarkable device, which can amplify frequencies from three cycles per second down to one cycle in 20 seconds. The amplifier drives an electrical pen recorder.

"The system has a number of advantages: the amplifier and recorder can be located away from the seismometer; their sensitivity may be controlled as easily as adjusting the volume of a radio; the assembly contains no delicate parts; one amplifier can be switched to several seismometers. Its disadvantage is that it responds to the velocity of the oscillations instead of their amplitudes; that is, the output of the system depends on the speed at which the magnet moves with respect to the coil. Regular seismographs show how much the earth moves, not how fast. Thus the recordings made with my seismograph differed from those in the textbooks, and I had difficulty in learning how to read them.

"I decided to make an amplifier that would indicate displacement rather than velocity. Most seismological observatories use an optical displacement amplifier, in which a pivoted mirror, either coupled to the pendulum through a lever or, in the case of the torsional instrument, mounted directly on the bob, reflects a beam of sharply focused light from a fixed source to a moving sheet of photographic paper. This arrangement is easy to set up and works nicely, but you pay for its simplicity with headaches. The instrument must be housed in a darkroom; you never know if it is working properly until the paper is developed. You also miss the thrill of watching a quake as it is being recorded. Finally, the large quantities of photographic paper required are expensive.

"There are a number of circuits which will translate the output of a velocity seismometer into its displacement equivalent, but all were beyond either my talent or my budget. I had just about concluded that the problem was hopeless when I came across the answer in a local radio store: a photovoltaic cell, the so-called 'sun battery' used in exposure meters. Why not substitute one of these for the photosensitive paper and drive the amplifier with its output? The principle has been used for years to amplify the deflection of galvanometers. Actually two photovoltaic cells are needed. They are arranged side by side so that the spot of light reflected from the mirror drifts from one to the other. The difference between the output of one cell and that of the other indicates the amount of deflection. The cells I use are made by the International Rectifier Corporation in El Segundo, Calif. They are called Type B-2-M and sell for $2.50 each. The system works with any convenient light source. A 35-millimeter slide projector works well. You can also make a source out of the reflector and socket assembly of a flashlight, a cardboard mailing tube and a spectacle lens [Figure 5].

"Although a mirror silvered on the back surface will do one silvered on the front surface will yield better results. The mirror can be mounted on two pivots which fit into indentations in a strip of metal bent into a U. The indentations are made with a center punch before the strip is bent. The lever linking the mirror to the pendulum can be a short length of magnetized piano wire. The wire is magnetically attached to and moved by a steel brad in the free end of the seismometer beam.

"The output of the sun batteries is directly connected to a Varian G-10 graphic recorder. This instrument includes a self-contained, high-gain amplifier with which the sun batteries can easily drive the recording pen to its full deflection. The recorder is manufactured in two units: the amplifier and pen motor comprise one and the chart drive the other. The units may be purchased separately.

"Although I have both units, I have replaced the chart drive with a homemade drum [Figure 6 ]. A seismograph eats up a lot of costly chart paper when it is driven fast enough to show the details of microseisms 24 hours a day. The drum is covered with a piece of common butcher paper 18 inches wide and 38 inches long. It turns two revolutions per hour. The pen is moved across the drum by a screw feed at the rate of 1/4 inch per revolution. Thus if the paper moves slightly more than an inch per minute, a 1,500-foot roll of it will keep the drum in continuous operation for three years! Butcher paper retails at $5 a roll.

"I used hand tools to build the whole assembly except for the drum, which must be true within 1/32 inch or the pen will skip. A local machine shop turned out the drum for me at a cost of $13.50 including all materials. The lead screw is a length of pre-threaded rod carried by most hardware dealers as stock for stud bolts. The threads are engaged by a pair of rollers from a Meccano set. The rollers support the plywood carriage 62 on which the pen unit rides. I now have two drum assemblies and an Esterline-Angus in operation.

"Contrary to the impression this description may create, the construction of a seismograph is neither difficult nor time-consuming. A few days ago I started to build a seismometer at noon; by five o'clock I had used it to bag a quake, the epicenter of which was 5,400 miles away. Incidentally, my recordings are timed by m electric clock checked against station WWV. The second hand of the clock closes a switch once a minute, discharging a two-microfarad condenser through the pen motor and recording a short pip on the seismogram.

"The casual observer who wanders into a seismograph station may not be impressed by the opportunities for experiment which these slow-paced instruments provide. Firsthand experience will soon correct that impression. Building, testing and attempting to improve seismographs has kept me away from my television set for many a happy month; more projects are lined up now than when I started. I am still trying to debug an instrument of the strain type which I put together over a year ago. This instrument, incidentally,is built on two pillars 60 feet apart. Between the pillars is a long piece of half-inch pipe. One end of the pipe is fixed to the first pillar; the other end is fitted with a coil which moves between the poles of a magnet fixed to the second pillar. When a quake alternately compresses and expands the crust of the earth between the two pillars, the coil develops a voltage [see Figure 7]. Unfortunately the setup is presently sensitive to wind and the rumble of nearby traffic."

Bibliography

SEISMICITY OF THE EARTH AND ASSOCIATED PHENOMENA. B. Gutenberg and C. F. Richter. Princeton University Press, 1949.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.