"1st. Of the time of transit, or 'wave period', at a given point of the earth's surface, of an earthquake shock, or earth wave, noting same to a small decimal of a second of time.

2nd. Of the vertical element of dimension, or altitude of the earth wave, at the moment of its transit, whether the wave be a positive or a negative one.

3rd. Of the horizontal element of dimension, or amplitude of the wave, at the same moment. 4th. Of the direction, as to azimuth, of the wave transit"[1].

In addition, the speed of the seismic waves were measured with several seismometers, that were installed at sufficient distance. In his first implementation of such an observation unit distributed over a wide area, MALLET measured the transit speed of earthquake waves in different media, as he described in his second report in 1850.

MALLET also claims to have invented the first completely self-recording seismometer and he provides us with the following brief description:

"It consists essentially of five fluid pendula, - glass tubes, partially filled with mercury, four for horizontal, and one for vertical elements of the shock. The displacement of the mercurial columns breaks contact, in an otherwise closed galvanic circuit, which, acting upon some simple contrivances, cause a pencil to trace a line upon ruled paper, whose length is proportionate to the time that contact remains broken, or to the amplitude and altitude of the earth-wave. The ruled paper, placed upon a cylinder, is maintained in motion by a clock; the position of the commencement of the pencil line traced on the moving paper, therefore, gives the moment in time, of the arrival of the wave, or initial instant of shock. The displacement of the mercurial columns is dependent upon inertia, and on the relative mass of mercury in the adjacent limbs of each bent tube."[2]

Figure 64: MALLET's self- recording seismometer

But MALLET also recognizes the huge disadvantage of all fluid seismoscopes: the original movement of the fluid caused by the earthquake shock is disturbed so much by transversal vibrations and reflected wave movements following the first shock of the compressional wave, that the information about the direction of the transit of the horizontal component of the earth wave is very unsafe. The same is true for is self- recording mercury instrument:

"Tubes partially filled with mercury give almost unobjectionable indications as to direction of transit. Their evils are too great delicacy or sensitiveness, for the observation of that class of earthquakes of mean power, which are the most important to be studied, and by which they are completely deranged occasionally, while they are continually being disturbed in such a seismic region by small tremulous movements that are unimportant to notice. As respects their indications of velocity and dimensions of the wave, they are liable to the objections already noticed as applicable to all pendula".[3]

MALLET had similar objections against the 'sismographo elletro-magnetico' that Luigi PALMIERI constructed and used in 1867 in the volcano observatory on the Vesuvius. PALMIERI's instrument was much more complicated: it could measure the vertical movement of the ground due to a heavy mass mounted to a spiral spring, measure the horizontal movement by U-form tubes filled with mercury where iron particles float on the surface, thus recording any movement. This device also had a watch and an alarm mechanism working on the basis of electromagnetic contacts. But also in this case, the sensitivity of the mercury device was too high and any strong vibration or violent jerks could discontinue the electromagnetic contacts and thus trigger the recording procedure.

Figure 65: Sismographo elletro-magnetico by PALMIERI

Thus MALLET comes to the early - and important - conclusion that earthquake instruments, that are based on the principle of the displacement of fluids, are only preferable if the seismic shocks are of a small dimension. In case of intensive shocks the heavy, fixed pendula areto be preferred.

Twelve years after his first treatise on the dynamics of earthquakes, when he recognized the difficulties in constructing self-recording seismometers in case they are supposed to measure all parameters of the earth wave on the surface, he proposed a ball-seismometer which still was constructed according to the old principle of unstable masses, but in addition had electrical devices for measurement and recording supposed to satisfy all requirements of self-recording seismometers.

There were two versions of this instrument. The first version of the ball seismometer was much more complicated, measuring both on the horizontal as well as on vertical levels so that both types of earthquake waves could be monitored. Fig. 9 shows this construction:

Figure 66: MALLET's ball-seismometer

Iron channels (ii) rising from the middle, are mounted on an iron plate (a,a) and two heavy iron balls are lying on the channels being in contact with a four-edged block and thus keeping an electric circuit closed. The iron plate including this device is mounted on an iron rod (b) that is located on a spiral spring (e) in a heavy iron frame (c,c).

As soon as the compressional wave is arriving at the instrument at an angle of the emergence indicated by the arrow a, the iron plate (a, a) that is lying on the spiral spring, is accelareted by the vertical component of the earthquake wave and the electric contact is discontinued. The duration of the interruption of the electric current is recorded on a strip and indicating the intensity of the current. The horizontal component of the wave, indicated by the north-south arrow, can be measured by using two of such instruments: one oriented north-south, the other east-west.

The second and simplier version of the 'projection seismograph' corresponds to MALLET's original idea of measuring the direction and intensity of the earthquake wave by bodies that were displaced, similar to his investigation of the damage to buildings that occurred during the Neapolitan earthquake.

Figure 67: MALLET's projection seismometer

The instrument consists of a column, with two balls on top, that keep an electric current closed by their contact to the device. When a vibration moves the balls, the current is interrupted while they fall down. In this case, the duration of the interruption of the current is recorded according to the distance of the ball thrown away. Obviously it is impossible to achieve a precise measurement of the different types of ground movements with such a method, in analogy to the old Chinese ball seismoscope. In addition it is obvious that a low speed and a long duration and a high speed and a brief duration would yield the same result, depending on the balls and the bearings, so that a real measurement of the intensity was actually impossible.[4]

The disadvantages of all such instruments built on this principle of unstable masses were so huge that the freely swinging pendula were actually preferred, as they had been in use for a long time in Italy and southern Europe, as we have seen already. But they also showed big disadvantages too, because a perfect stable mass that is independent of the movements of the ground can never be reached, since a pendulum always needs a suspension that is always linked to the ground somehow.

Thus the old vertical pendula did not show any clear relationships to the ground movements and these movements got mixed with their self-induced vibrations. The only way to approximate the theoretical ideal of an absolutely stationary mass is to fix a very heavy weight to a very long wire.

Thus further developments of the vertical pendula focussed on using heavier weights and longer pendula. The big and heavy pendula were typical of the Italian seismographic instruments. The weight of the pendulum used by CANCANI in Rocca di Papa near Rome had 100 kg and was 7 m long. A new version from 1895 weighed 200 kg and was mounted a 15 m long steel wire. The huge instrument used in Catania weighed 300 kg and was 26 m long. The biggest instrument was the microseismograph built in different versions by G. VICENTINI in Padua. It consisted of a mass of 50, 100 or 400 kg that were mounted to wires of 1.50, 3.36, 4.50, 6.00 or 10.50 meters length. The pendulum of the instrument used in 1896 weighed 408,65 kg consisting of 13 huge plumb disks instead of a ball. The overall length of the pendulum was 10,68 m.

Figure 68: Microseismograph by G. VICENTINI in Padua

This was not yet the end of this development. The director of the seismographic station in Rocca di Papa, G. AGAMENNONE constructed several sensitive seismometers with vertical pendula with heavy masses in the form of a whole cheese, i.e. cylinders of low height but with large diameters and weighing 200, 500 and finally 2 000 kg. Despite these dimension, the disadvantage of the vertical pendula, the natural self-induced vibration of the not really stable masses, could only be compensated by absorbers in order to reduce the resonance and an appropriate mechanical device for amplifying the movements of the stationary mass due to the horizontal ground movements.[5]

Figure 69: AGAMENNONE's vertical pendulum

This problem of compensating self-resonance was primarily dealt with by British researchers in Japan. Under the guidance of John MILNE , James EWING and Thomas GRAY at the Imperial University in Tokyo, constructed a series of instruments that compensated the self-resonance of the pendulum by appropriate devices without having to increase the length or mass of the pendulum. In this way, compact instruments were developed that were used all over the world.

In addition, the problem of an appropriate recording method had to be solved. The simpliest and most intuitive way of recording, that was used in the old vertical pendula and to a certain extent, used by the British researchers in Japan[6], that consisted of a pin that was mounted under the pendulum mass recording movements on a fixed surface, did not yield any clear relationships to the actual ground movements. The crossing lines indicate the diversity of movements but they do not represent the sequence of each phase of movement.

Figure 70: Earthquake registration on a fixed plate

Thus, it was an excellent idea by the director of the Central Institute of Meteorology in Vienna, Karl KREIL to use recording strips - or seismograms - moved by a clockwork instead of a fixed recording surface. MALLET was the first to recognize the value of this invention and in his report of 1858 he provided a detailed translation into English of the relevant parts of the commission reports of the Imperial Academy of Sciences, where KREIL reported on his registration device:

"A good seismometer is a desideratum still to be devoutly wished for. It should not only show the commencement of the stronger, but also of the weaker shocks, as well as their duration, direction and strength, - a task which is too great for a self-registering apparatus. Therefore, every idea towards the improvement of such instruments must be welcome, and on this account I venture to bring forward the following design. Let d e be a rod of wood or metal suspended at a, which at d is fastened to the elastic spring c, like a pendulum of a clock, and therefore can swing in the plane of this spring in a vertical direction. Let a b be a second spring upon the first vertical one, which permits the bar of the pendulum, de, to swing in the plane of the spring c, i.e. at right angles to the former vertical plane. The bar d e and the weight fastened to it can therefore swing in every direction, without its being permitted to turn on its own axis of vertical length, and as if there were but a thread or thin wire at b. The cylinder f g h i contains clockwork, which obliges it to turn round upon the bar of the pendulum (as its perpendicular axis fixed with reference to rotation) once in 24 hours. It is covered with paper or other material, which can be marked on without great pressure. It contains on the lower edge the numbers of the hours which can move behind an index m, fastened to the plate k l, which is fixed to the axis of the pendulum. Upon a neighbouring pin, o p, is an elastic and thin arm of brass, o n, which carries a pencil at n, which, by means of a screw can be pressed against the cylinder and removed from it. It is in firm contact with this, and marks upon it an uninterrupted line so long as the pendulum remains at rest; if, however, this begins to swing, in consequence of the whole system being shaken, this line will be broken, and strokes produced which will have a horizontal direction if the pendulum swings in the plane of n o, but will be perpendicular and crossways if swinging in the plane perpendicular to n o. The force and length of this stroke will give an approximation to the strength of the shocks. The middle of the stroke, or, if they are vertical, the end of the uninterrupted line, gives the time of the commencement of the shock. The strength and direction of the shocks may also be approximated if the (as respects rotation) fixed plate h i k l have an annular recess, filled with quicksilver until its surface reaches the holes s s s, made in the cylindrical sides. At the first motion of the pendulum, the quicksilver will be shed out through these holes into a dish divided into the same number of compartments as there are holes, like those already in use in many existing instruments of this kind (Cacciatores)".[7]

"Ein guter Erdbebenmesser gehört noch immer unter die frommen Wünsche; er soll nicht nur den Eintritt der stärkeren Stösse, sondern auch jenen der schwächeren, die Zeit, die Richtung und Stärke des Stosses angeben, eine Aufgabe, die für einen selbstzeichnenden Apparat zu gross ist. Es muss daher jede Idee zur Verbesserung solcher Vorrichtungen willkommen sein und aus diesem Grunde erlaube ich mir den beifolgenden Entwurf vozulegen.

Es sei d e eine in a aufgehängte Stange von Holz oder Metall, welche bei d an der elastischen Feder c nach Art eines Uhrpendel befestigt ist, daher in einer auf die Ebene dieser Feder senkrechten Richtung schwingen kann; a b sei eine zweite auf die erste senkrecht stehende Feder, welche der Pendelstange in der Ebene der Feder c zu schwingen erlaubt. Die Stange d e und die an ihr befestigte Last kann daher in jede Richtung schwingen, ohne dass ihr gestattet wäre sich um ihre eigene Längenaxe zu drehen, wie es z.B. bei einem Faden oder dünnem Drahte der Fall wäre. Der Cylinder f g hi umfaßt ein Uhrwerk, das ihn nöthigt, sich in 24 Stunden einmal um seine senkrechte Axe zu drehen. Er ist mit Papier überzogen oder einem anderen Stoffe, auf welchem ohne großen Druck geschrieben werden kann. Er enthält am unteren Rande die Stundenzahlen, welche hinter einem am Teller h i k l befestigten Zeiger m vorüber gleiten. An einem nebenstehenden Pflocke o p ist ein elastischer und dünner Arm von Messing o n angebracht, der in n den Bleistift trägt, der mittelst eines Schraubengewindes gegen den Cylinder geschoben und davon entfernt werden kann. Er ist in stäter Berührung mit demselben und zeichnet darauf eine ununterbrochene Linie, so lange das Pendel in Ruhe bleibt. Fängt aber dasselbe in Folge einer Erschütterung zu schwingen an, so wird die Linie unterbrochen und es entstehen Striche, welche eine horizontale Lage haben werden, wenn das Pendel in der mit n o parallelen Richtung schwingt, eine senkrechte, wenn die Schwingung auf die Richtung n o senkrecht vor sich geht. Die Stärke und Ausdehnung dieser Striche wird eine Schätzung der Stärke des Stosses erlauben. Die Mitte der Striche oder, wenn sie vertical sind, der Endpunkt der ununterbrochenen Linie gibt die Zeit des Eintrittes an.

Stärke und Richtung des Stosses können auch abgeschätzt werden, wenn der Teller h i k l eine ringförmige Vertiefung hat, die mit Quecksilber zu füllen ist, bis es die an der Seite angebrachten Löcher s, s, s ... erreicht. Bei eingetretener Bewegung wird das Quecksilber durch diese Löcher ausgegossen und in einer, in eben so viele Fächer als Löcher sind, getheilten Schale aufgefangen, wie dies bei schon bestehenden Vorrichtungen dieser Art gebräuchlich ist."[8]

Figure 71: Earthquake measurement with moving recording poper

The fact that KREIL's recording method could represent the whole temporal process of an earthquake was widely accepted so late, is only due to the fact that originally an earthquake was considered to consist of a single impuls, provoked either by a subterranean explosion (MALLET) or the subsidence of cavities (VOLGER). After the introduction of instruments it became evident that each earthquake consists of several phases of stronger and weaker ground movements. Different methods were used to record the complicated movements.

A direct record without any decomposition on a fixed record surface will become unintelligible very soon, as has been shown. The movable record surfaces on the other hand have the disadvantage of carrying the recording indices with them und thus irritating the mass of the seismometer. In this case the absence of friction was even more important than for static records. This is more complicated, the faster the registration surface is moved. High speeds are necessary, though, in order to detect the finest details in the disturbances. The most primitive recording instrument, that could only be used for static records, used a pencil on a fine strip or a brush on paper. This method was not used any more for the movable records, where fine steel or glass pins were used on smoked paper or glass. The best method was the photographic one, being used by MILNE and others already in the 19th century for the so-called horizontal pendula, and leading to a new era in instrumental earthquake observation.