VULCANOLOGY AND SEISMOLOGY. 227 



apon a moving smoked-glass plate. The proportions of the frame and 

 pointer are such that the line of attachment of the bob is the center of 

 percussion with respect to the vertical axis of support of the frame. 

 Most of tbe important graphic records obtained in Japan have been 

 made with some modification of this instrument. In its improved form 

 the frame is made triangular and the bob is a truncated cone of cast- 

 iron. Gray's conical i)endulum, in which the upper pivot is replaced 

 by an elastic wire, is also described, and a modification is suggested by 

 which the lower pivot also may be dispensed with in a similar way. 



The third chapter describes and discusses several previously used 

 seismographs dependent on rolling spheres or cylinders, and various 

 forms of pendulums making their records either directly or through the 

 intervention of multiplying levers, and especially the author's "duplex" 

 pendulum, which combines an ordinary with an inverted pendulum, so 

 as to make tbe equilibrium of the system neutral. 



The fourth chapter is devoted to instruments for recording vertical 

 movement, which have already been described elsewhere. 



The fifth chapter gives the results of instrumental observations in the 

 case of nine earthquakes, selected from a much larger number recorded 

 in Japan between 1880 and 1883, accompanied by fac-similes of the trac- 

 ings made by various instruments, and a discussion of them. 



The sixth chapter describes a number of miscellaneous instruments 

 and methods of observation, none of which are new; while the seventh 

 contains a brief statement of the author's results from a comparison of 

 the records obtained with different instruments. He concludes that 

 the only seismometers of value are those "which aim at giving a steady 

 point or line during the disturbance." 



In a paper read before the Seismological Society of Japan {Trans, of 

 8. S. of Japan, vi, 22), C. D. West suggests a new type of seismograph, 

 which is also described in an appendix to E wing's memoir. It consists 

 of a heavy weight supported by a system of links similar to those used 

 in the Richards indicator to secure rectilinear motion. A weight so sup- 

 ported would be free to move in a horizontal line through a considerable 

 amplitude without losing its astatic properties. 



In the Journal of the German Geological Society (vol. xxxvi, 29), 

 G. R. Lepsius describes a modification of Oacciatore's seismometer, in 

 which the fluted dish is replaced by a watch-glass containing mercury, 

 supported on the elevated central part of a porcelain dish whose outer 

 part forms a ring of sixteen deep hollows, into some one of which the 

 spilled mercury must fall. The whole apparatus has a diameter of 

 lOl'"'" and a height of 60°"". 



Several forms of recording apparatus for earthquakes were suggested 

 by Johnston-Lavis. (Nature, xxx, 009.) For the steady point in the 

 registers of the horizontal component a pendulum was usually employed, 

 and the direction and amount of its motion relative to the rest of the 

 apparatus were recorded (a) by the pendulum pulling upon a cord 



