208 Mr. T. Gray on Instruments for Measuring 



and the very great mass which can be put into the weights W 

 without increasing the friction, the only change required being 

 that the wire t should be lengthened. The question of mas- 

 siveness in the portion whose inertia gives the writing-power 

 is important, as the multiplication which can be used is prac- 

 tically proportional to this mass. Yery large multiplication 

 is not generally necessary for earthquakes of ordinary magni- 

 tude; but there is a large class of tremors which have a great 

 interest to investigators; and these may be recorded by giving 

 a multiplication of, say, from twenty to fifty, according to the 

 degree of minuteness aimed at. Of a considerable number of 

 machines which I have devised and tried, I think this is deci- 

 dedly the most sensitive; and it is probably the most sensitive 

 and at the same time accurate recording-instrument now 

 in use. 



It is of course to be understood that, in connexion with this 

 seismograph, some form of clockwork arrangement must be 

 used for the purpose of supplying a moving surface on which 

 the different motions of the pointers are to be recorded. 

 When such an arrangement is adopted, the horizontal ampli- 

 tude and direction and the period of each individual motion 

 are recorded, the period of course being reckoned from the rate 

 at which the drum is moved by the clock. The records are 

 most convenient for use when taken on smoked glass or trans- 

 parent smooth paper, because in that case, by simply varnish- 

 ing the plate or paper sheet, the record can be preserved and 

 used as a negative for obtaining photographs. 



VI. Hydrometer Vertical-Motion Seismograph. 



Some time ago I made several attempts to obtain an instru- 

 ment capable of recording the vertical movements in an earth- 

 quake shock. I propose in the present paper to describe two 

 of the most successful of these attempts. 



My first attempt in this direction was suggested to me 

 by observing the motion of a Nicholson's hydrometer when 

 immersed in a liquid and slightly displaced from its equili- 

 brium position. The period of up-and-down oscillation of 

 such an instrument can, by varying the thickness of the stem, 

 be made almost any length desired. I therefore proposed to 

 use an enlarged hydrometer, with the lower basket removed 

 and ballast placed in the bottom of the hollow foot. For the 

 sake of ease of regulation I attach the upper end, by means of 

 a thread, to a pulley through the centre of which, at right 

 angles to its plane, a wire has been passed and then soldered. 

 This wire is stretched between two springs, and can be twisted 

 round its own axis, thus allowing the force of torsion to regu- 



