ON THE EARTHQUAKE AND VOLCANIC THENOMENA OF JAPAN. 121 



black filament of straw. To balance the weight of the reed and straw a 

 light arm, resting upon a pivot on the underside of the brass tube, carries, 

 at its extremities two small weights. Assuming the pivoted mass, which 

 has a considerable moment of ineitia and but little weight, to be the centre 

 of oscillation of the system, which is held in a horizontal position by u 

 thin wire tie, then the arrangement is that of a conical pendulum seismo- 

 graph, having a multiplication of about tweiaty, and without the friction 

 of a writing pointer. The black hair at the extremity of the boom floats 

 above the slit in the box containing the drum, carrying the photographic 

 film, which moves at a rate of 2 inches per houi-. The resulting diagram, 

 is a white line showing the position of the shadow of the hair-like- 

 filament. 



A modification of this instrument was to reduce the length of th& 

 boom to 2 feet, and because the trace given by the shadow of the filament 

 at the end of the boom was wanting in sharpness, the filament was re- 

 placed by a thin plate of mica about half an inch square, with a small slit 

 in its centre, similar to that used in the larger apparatus. Because the 

 floating plate attached to the boom does not cover the whole of the slit in 

 the box above the drum carrying the photographic film, the diagram is a 

 black line given by the spot of light from the crossed slits, bounded on the- 

 right and left by a black band, the irregularities on the inside edges of 

 which correspond with the irregularities on the central line. Every houi' 

 a separate clock depresses on one end of a balanced lever, at the other end 

 of which there is a light vane which rises in front of the lamp, and cuts 

 tlie light oS for one minute. The result is that the central line (whem. 

 there are no tremors), and the two bands at all times, are transversely 

 marked by a distinct white line. Not only do the bands indicate time, 

 and repeat the sinuosities and other movements shown on the central line, 

 but by the presence or absence of striations they immediately show whether 

 the clock driving the drum has been working regularly. 



Instead of cutting otf the light from the lamp, the light is now cut oft" 

 the edge of the fixed slit by the hour hand of a watch moving horizontally 

 across the same (fig. 3). 



Fig. 3. 



I 



An experiment made at this station was to obtain diagrams on films, 

 which only moved at the rate of 75 mm. in twenty-four hours, or 3 mm. 

 per hour, whicli, from the results obtained, appears to be a suitable speed 

 for recording diurnal waves and earth tremors (fig. 9). The western elonga- 

 tion is completed rather suddenly between 7 and 8 a.m. (nineteen and twenty 

 liours). The eastern movement, which is performed with extreme irregu- 

 larity, is ended about 4 p.m. The amplitude of this wave is about 30 mm. 



