818 EEPOET— 1888. 



taining air), united by a U-shaped tube containing mercury in the bend, and 

 having sealed into it two thin platinum wires whose inner ends are in contact 

 with the mercury, one deeply and the other slightly submerged beneath its surface. 

 There passes through the bulbs the two tubes (one open and the other closed), 

 containing the two platinum wires, difference in the temperatures of which de- 

 termines the action of the apparatus. 



The current which passes through the lamp filament traverses also the indicator 

 wires, and the wires in contact with the mercury ; it wUl therefore be easily 

 understood that, in the absence of fire-damp, as the wires contained within the 

 thermometer bulbs will be equally hot, there will be no movement of the mercury, 

 and that in the pi-esence of fire-damp, as the exposed wire will become hotter than 

 the enclosed wire, the air in the two bulbs will be unequally heated, and conse- 

 quently the mercury will move away from the hotter bulb, and if its movement 

 is sufficiently great to uncover the end of the least submerged wire, the circuit will 

 be broken, and the light will go out. A switch is pro\-ided to enable the broken 

 circuit to be remade, and the lamp lighted after the indication has been given. 



The indicating wires are so short as not to consume a large proportion of the 

 energy of the battery. This is the more necessary as it is intended that the current 

 shall continually pass through them during the time the lamp is in use. 



The other form of the apparatus to which he referred is similar in principle to 

 that described, but instead of the diflerential nir thermometer, use is made in it of 

 a differential metallic thermometer, constructed on the principle of Immisch's 

 pocket thermometer. In this case the thermoscopic part of the apparatus consists 

 of a pair of helical tubes of oval section filled with liquid. 



The pair of helices are in every respect alike except that one is a right-hand 

 screw and the other a left ; they are arranged one over the other with their free 

 ends lightly touching, so as to be in electrical contact with each other. 



Variation of temperatiu-e, acting equally upon both spirals, dees not disturb 

 the contact, but if one helix be heated more than the other, there will be separa- 

 tion of the free ends and breaking of the electrical contact. In adapting these to 

 the purpose of the detector, the enclosed platinum wire is fixed in the axis of one 

 helix, and the open wire in the axis of the other. The electrical connections are 

 80 made that normally the current from the lamp battery passes through the lamp, 

 the exposed and enclosed platinum wires, and the helices ; and, as in the other 

 form, there is a switch to close the circuit and relight the lamp when it has been 

 extinguished by the action of fire-damp. 



The apertures by which the air gains access to the exposed platinum wire are 

 carefully guarded by wire gauze, and they can be closed entirely when, by the 

 going out of the light, there is evidence of the presence of fire-damp. 



TUESDAY, SEPTEMBER 11. 

 The following Papers were read : — 



1. An improved Seismograph. By E. A. Cowpee, M.Inst.G.E. 



In order to register the motions of the earth, north, south, east and west, it is 

 of course necessary to have a heavy weight as little as possible affected by the 

 motion of the earth ; parallel motions, or ' compound pendulums,' have been used 

 with good effect, but a simpler and more efi'ective plan would be to place a heavy 

 flat weight on a plate of glass with three small * steel bicycle balls ' below, resting 

 on another flat or slightly hollowed plate, so that when the vibrations of the earth 

 ceased, the weight would slowly return to its normal position. In place of glass 

 plates, ground porcelain or metal plates might be used. The weight might be 

 brought back to its normal position by a light pendulum, and then all the plates 

 might hejlat and oi com7non jtlate-glass. Then two levers, acted on from the centre 

 of gravity of the weight, and multiplying the motions, would register on long strips 



