562 



KNOWLEDGE & SCIENTIFIC NEWS. 



[October, igo6. 



a weight at one end, and the rod was suspended from a 

 separate stand, by two unequal silk threads attached 

 to the ends of the rod respectively. The longer thread 

 — from the weighted end — was extended l)ackwards 

 and upwards to a point in the front surface of a vertical 

 support, and was there fastened. 'Hie shorter thread 

 w;is i-xlended from the other end of the rod, forwards 

 and upu.irds, to a point in the rear surface of the same 

 \ertical support, and was there fastened so that the 

 weighted rod hung freely, approximately horizontal, as 

 in an unequal bifnlar suspension (see Fig. 8). 



Clo.se found his pendulum so sensitive that it was 

 moved through an easily measurable angle if finely 

 adjusted, when the whole ground floor of the house, in 

 the basement of which his pendulum stood, was tilted 

 " over to leeward " by a moderate gale. I know of no 

 other observations made with it or results obtained 

 thereby. 



Zoll'ncr's Pendulum (1872). This, like I'crrott'.s, was 

 practically upon Hengler's principle. Zollner himself 

 credits I'errott with having invented the "horizontal 

 pendulum," but as he also quotes from Hengler's paper 

 and discredits the results therein reported, it is amazing 

 that he did not credit Hengler with the invention of it, 

 unless he meant to confine the credit to the modifica- 

 tions introduced by I^errott and added to by himself. 

 If so, Hengler's credit is the greater. 



Zollner used a separate stand, and made modifica- 

 tions in detail, besides using such long previously well- 

 known accessories as the Gauss mirror, etc. Vide 

 " I'oggendorff's .'^nnalen," 1873, wherein it may be 

 seen that Zollner's object was to investigate as to the 

 "origin of the earth's magnetism," etc., also the 

 results obtained by the use of his pendulum may thereby 

 be traced. 



On the 22nd February, 1880, a great earthquake in 

 japan, led to a meeting being convened there, and at 

 which the Seismological Society of Japan was estab- 

 lished, for the special purpose of investigating earth- 

 quakes. From that time, seismology, as a science 

 apparently dates. The importance of seismometry as 

 affecting gravitation research, and vice versa, has not 

 vet been "fully realised; nor has the action of bodies 

 outside the earth as affecting both been much more than 

 guessed at, and to a very slight extent definitely experi- 

 mented upon. 



No doubt seismological investigations had previously 

 been made in Italy and elsewhere, for in the BuUetino, 

 1876, page 5, Rossi describes instruments previously 

 used for that purpose; and the experiments of 

 M. d'.^bbadie, with a dish of mercury, of Plantamour, 

 with a level, and of Siemens, with the " Attraction- 

 meter," as reported in Phil. Traits., 1876, are distinctly 

 related thereto. But after 1880, a particular distinction 

 specially affecting so called " horizontal " pendulums 

 requires to be noticed. 



Previously the characteristic of astronomical pendu- 

 lums had usually been that the earth's surface in tlie 

 locality where they were respectively u.sed, together 

 with the building, or rigid support to which they were 

 suspended, w-as considered as at rest, in relation to the 

 oscillating pendulum weight. But the usual character- 

 istic of seismological pendulums is directly opposite 

 thereto, for the pendulum weight, by its inertia, is sup- 

 posed to remain usually in relative rest, and to serve 

 as a fulcrum for the pendulum rod which is considered 

 to ha\e its axis of oscillation moving with changes of 

 the earth's surface. In fact, the rod is simplv regarded 



as having the function of a lever in these instruments, 

 and the motion of the stand at one end is either recorded 

 by a writing point or ray of light at the other, or the 

 angular deviation of a ray of light or .scale image, by 

 means of the v\ell-known (iauss mirror method, is, :f 

 necessary, photographically recorded as magnifying 

 that of the stand, when the mirror is attached to the 

 rod or its weight. Oscillations of short-period earth- 

 quake waves, apparently, are chiefly sought to be re- 

 corded by means of seismographs in general u.se. 



Ewmg's Seismograph (about 1881), described in 

 Eiicy. Brit., N. S., Vol. 27, page 605, appears to have 

 a seismographic lever, which is therein called a " hori- 

 zontal pendulum," consisting of a frame of metal work- 

 ing gatevvisc upon two points, and carrying a large ball 

 of metal, which has a long writing pointer attached. I 

 am not aware of any publication prior to 1881 of such a 

 " two-point " suspension, though I had used such long 

 before without publication, and so far the credit of 

 prior publication of a " two-point " suspension is his 

 to the best of my information. 



The " two-point " lateral suspension is a step in ad- 

 vance upon the " ring and one point " lateral suspen- 

 sion used by Gerard, but I have been unable to find an 

 instance of any astronomical or cosmic observations 

 having been made by the use of this instrument, which 

 might benefit gravitation research, nor even of any at- 

 tempt to apply it for the purpose of such discoveries. 

 That, however, may be due to my having been ignorant 

 even of its existence until I heard of it in July, 1906, as 

 containing a " two-point " suspension. The " suspen- 

 sion " and " frame " are in principle like Fig. 10. 



Sharp's Pendulum, inxented by the writer hereof 

 about 1875, and used in gravitation research until date 

 of its first publication, in February, 1885 {vide Jour. 

 Upool Asiron. Sac, March, 1885), was essentially an 

 astronomical and inclined pendulum, expressly not hori- 

 zontal. It consisted of a tube or trussed rod (or boomi, 

 having a ball at its outer end, and a needle point at the 

 other, but I now prefer to use a rod having a star or + 

 shaped section when I cannot get suitable tube. A 

 cord or wire of brass, aluminium, or copper, was fas- 

 tened at the outer end of the rod and carried backwards 

 as a stay to a reversed needle point, forming part of 

 a steel bolt driven into the solid angle of the room 

 walls just below the ceiling. That wire ended in a 

 stirrup, the interior surface of which was hardened, and 

 had a minute indentation made therein with a fine 

 " centre punch." That stirrup was hooked upon the 

 reversed needle point so that the point entered the in- 

 dentation, and was therel>y prevented from sliding out 

 of place. The point of the rod below^ in the solid angle 

 of the walls, near the floor, rested in a similar inden- 

 tation in a hardened surface, which was mounted on a 

 compound lathe slide rest, giving definite and con- 

 venient control of its position, so that the lower needle 

 point could be brought as was requisite almost 

 vertically below, but a little in rear of the line contain- 

 ing the upper point (see Fig. 9). 



The rod formed the lower, and the stay formed the 

 upper, of two sides of an isosceles triangle, of which the 

 third side was wanting, and the pendulum oscillated 

 like a > shaped gate, upon the two needle points as 

 hinges. The needles, of course, were in line with the 

 pre-ssures they had to oppo.se. With adequate time 

 and patience no difficulty whatever was experienced m 

 obtaining the normal variation curve of /daily and 

 monthly lunisolar changes, by direct readings from a 

 scale of millimetres, taken by means of an ordinary 



