330 



NATURE 



{Jan. 31, J 



fold in which it fell, and its weight disturbed the equilibrium of 

 the balance, which by making contact gave notice of the shock. 

 Many of these are at present in use in Italy. A recent improve- 

 ment is to suppress the balance, the contact is made by the 

 slug being caught between the supporting style and surrounding 

 umbrella-shaped grooves. I'he author of the paper experimented 

 on earthquake alarms for vertical shocks, and found that, with 

 three wire spirals, two supporting different weights and one 

 without, the last gave less repeated oscillatory movements 

 after the shock, returning more quickly to rest ; and he therefore 

 suggests this as an improvement on the old spring and bob 

 seismoscope. 



The next point discussed is whether it is preferable to stop or 

 start a clock at the moment of an earthquake, and is in favour of 

 starting one, as the few minutes between the shock and the 

 arrival of the observer could afford little time-error, whereas 

 there is no means of correcting the time-error of a clock 

 stopped by the shock. 



Following an observation of Prof. J. Milne on the facility with 

 which light objects, such as pens, pencils, &c., when propped up 

 nearly vertically, fall by the slightest movement, the author has 

 constructed a seismoscope. A long thin liar, standing in a con- 

 cavity, is supported by a small prop that can be regulated so as to 

 put the bar almost in a vertical position, which is surrounded by 

 an isolated ring of brass. On the occurrence of a shock, the bar 

 falls and completes a circuit with the brass ring so as to ring a 

 bell, start a clock, &c. Experiments made by the inventor 

 and Prof. Tacchini showed the extreme sensibility of this 

 seismoscope compared with others placed on a marble wall- 

 bracket upon which small weights were allowed to fall. Drawings 

 and descriptions are given of a method of using this seismoscope 

 to start a clock, the pendulum of which is retained ready to 

 vibrate by a short catch. The ring surrounding the falling bar 

 is supported on a jointed base and the falling bar displacing it, 

 and acting on a system of jointed levers the catch falls out of the 

 way of the clock pendulum. 



Next, Signor Brassart alludes to the different ideas current 

 about the best method of connecting seismic apparatus to the 

 ground, and shows the necessity of a series of experiments to 

 determine this by using stakes, walls, &c., and employing 

 seismographical instruments of different types. 



In the next memoir by the same author, a description and 

 drawings of the principal apparatus used in seismological re- 

 search in Japan is given, together with a criticism of each. He 

 concludes the review of these instruments by pointing out that it 

 is only those in which a pendulum is employed, together with 

 those possessing a rolling base, either spherical or cylindrical, 

 that do not require adjustment, beyond rendering the base 

 horizontal. His own choice would be a vertical pendulum, be- 

 cause not only does it not require adjustment and is constantly 

 ready to indicate the shock, but is always in the same condition ; 

 and then adds, "In other arrangements it is possible greater 

 stasis or neutrality of the mass is obtained, but notwithstanding 

 that, for the above-mentioned reasons I should show a propensity 

 for pendulums." This is a preference I fear few other seismo- 

 logists would share with the author, since he undoubtedly alludes 

 to vertical forms. The floating seismograph of Gray is favourably 

 mentioned, but troubles from evaporation of the liquid, &c., 

 pointed out. This is similar to the opinion expressed by the 

 writer of this article, s-ime years since, in a paper on seismo- 

 graphs published in Nature. 



Signor Brassart then proceeds to suggest an instrument con- 

 sisting of an annular pendulum suspended by a silken cord I 

 metre long, attached at its upper end in the middle of another 

 horizontal cord made taut by side screws in the frame. In the 

 middle of the bob, near the centre of gravity, is a pin which slides 

 in two slots placed at right angles to each other. These slots are 

 cut in the ends of levers of the first order, one being bent at 

 right angles so as to bring the opposite ends of the levers side 

 by side, to which are attached writing needles. These are 

 made to trace on a smoked glass ]3late held on a truck, which by 

 a seismoscope is liberated and drawn along at a uniform and 

 known rate by a clock. This proposition was soon carried out with 

 some modifications, and at the end of the paper is an appendix 

 describing it. The changes of design consist of the suspending 

 silk cord being attached to a stage supported on three vertical 

 cork cylinders, the object of which is not very evident, except, 

 perhaps, to absorb a part of the vertical component. This latter 

 is registered by a Gray's compensated bob attached to the same 

 frame, and by means of levers made to trace side by side with 



those of the two horizontal component styles on the same plate, 

 or, as the author suggests, in some suitable cases upon an end- 

 less band of paper. This seismograph has many defects which 

 it will be more convenient to discuss later on. This paper is 

 dated December 1887. 



It has dc ubtless been the ambition of most inventors of seismic 

 instruments to be able to register the three components of an 

 earth-movement employing but one steady-point or mass. Any- 

 one who makes such an attempt, however imperfect be his 

 result, deserves much credit for trying such a herculean feat 

 in mechanical construction. 



Signor E. Brassart's later production is described in February 

 of this year, and, in so far as the registration of the horizontal 

 components goes, is quite similar to that described in December 

 1887. Inserted in the suspending cord is a spiral spring inclosed 

 in a tube, which consequently also allows the bob movement up 

 and down. The peg or rod then slides in the slots of the 

 horizontal component levers, is ]3rolonged downwards a suitable 

 distance, and terminates in a ball, which is engaged in a short 

 tube or cup at the end of a bar. A gimbal is a mechanical 

 arrangement by which a bar supported in its centre can rotate 

 around any theoretical axis in the horizon of the gimbal, whilst 

 any motion along a line normal to the horizon or plane of the 

 gimbal would be resisted ; or, if the gimbal be suspended to a 

 lever, this will be moved in proportion to the vertical movement 

 applied to the gimbal arrangement. (The writer of this article 

 believes the gimbal was first applied by himself in seismological 

 instruments, but he is open to correction.) 



Signor Brassart takes advantage of this, and makes the sphere 

 at the end of the peg from the pendulum, form a ball-and-socket 

 joint with the tube or cup bar of the gimbal arrangement which 

 is supported horizontally at the end of a lever. Now all hori- 

 zontal excursions of the peg, joint, and gimbal bar, are quite 

 free, but vertical excursions will disturb the whole gimbal 

 system with its lever, the motion of which by suitable mechanical 

 arrangements is traced by the side of the horizonal component 

 styles. 



. It will be convenient now to examine the favourable points and 

 disadvantages of these instruments. In the first place, we have 

 the well-known defects of the vertical pendulum for registering 

 the horizontal components of an earthquake. Then, the author's 

 method of the peg sliding in the two slots of levers introduces 

 an amount of friction, even where the finest workmanship and 

 materials are adopted, as to modify the results to an important 

 extent. In regard to the gimbal arrangement, that answers 

 perfectly during vertical without horizontal movement ; but the 

 moment that lateral displacement takes place, the bar of the 

 gimbal, and the pendulum length, form two sides of a triangle 

 instead of a straight line. On the principle that two sides of a 

 triangle are always greater than the third, the more the horizon- 

 tal components are, the less would the vertical component be 

 registered. And even there is a point in which the motion of the 

 pendulum bob would simply resolve itself in a greater tilt of the 

 gimbal bar. 



The next study by Prof. G. Grablovitz is a kind of programme 

 of seismological ob.'ervations to be made in Italy. The paper 

 shows a clear conception of the known or probable relationship 

 of different seismic phenomena, and apparently the best methods 

 to adopt for their study. One most favourable point about the 

 author's writing is a total absence of that long-winded style, 

 charlatanism, and seismic magic, which has characterized the 

 writings of many Italian seismologists up to quite recently. 

 Very little, however, is new : similar suggestions may be found 

 scattered through the writings of many other seismologists, and 

 part, at least, were thrown out by the writer of this article in many 

 of his papers relating to Ischia. It is, in fact, in organizing the 

 new observatory on that island by Prof G. Grablovitz that has 

 led that gentleman to the considerations contained in his paper. 

 From observations made with the mercury cup in the temporary 

 Ischian Geodynamic Observatory, passing carriages and numbers 

 of people could be detected at a distance of loo metres ; there- 

 fore the author considers this as the minimum distance of isola- 

 tion such an observatory should. possess. 



The next is a report of a sub-comtnittee, drawn up by Prof. T. 

 Taramelli, dated December 1886. It consists of a review of 

 the principal earthquakes that have affected different regions of 

 Southern Italy, and the means then in existence for their study. 

 This was preceded by another report referring to Northern and 

 Central Italy. The author opens his report by discussing the 

 views of different writers as to the causes of earthquakes, with- 



