of dark Heat-rays by Gases and Vapours. 13 



the siphon-shaped glass tubes. After standing 24 hours, the 

 sulphuric acid was blown out the same way, and the end of 

 the glass tube carefully cleaned. 



In 0, on the other hand, pieces of marble were placed which 

 had been cleaned on the surface with hydrochloric acid, then 

 washed and dried, and moistened with a solution of caustic 

 potash. This kind of purifying-apparatus is mainly adopted 

 from Tyndall *, and should be employed in all researches in 

 which the chemical purity of gases is required, instead of the 

 usual Bunsen's wash-bottles. Lastly, the tube x must be 

 mentioned, which was filled with pure cotton-wool and the 

 upper portion with asbestos, to separate the organic germs, 

 which, according to Tyndall f, can only be separated in this 

 way. 



The galvanometer, an ordinary thermo-multiplier, the copper 

 wire of which was covered with white silk free from iron, was 

 placed several rooms away from the rest of the apparatus. 

 It was observed, in fact, that the galvanometer altered its zero- 

 point considerably with change of temperature ; so that the 

 room in which it was placed, on a small table attached to the 

 wall, was kept with the windows closed in order to avoid change 

 of temperature as much as possible. The instrument was 

 provided with a mirror, and a scale at a distance of about 

 3 metres from the mirror. A board was placed horizontally in 

 front of the the scale, on which a small gas-flame could be 

 moved by the observer at the telescope by means of cords, so 

 as to adjust the illumination of the scale. In this way the 

 scale was easily illuminated, even when the deflection was very 

 great, without any elaborate arrangement. 



The coil of the galvanometer consisted of copper wire 

 3*5 millims. thick. No binding-screws were employed. At 

 the points where this wire was to be connected with the wires 

 leading from the thermopile, the wires, after being carefully 

 polished, were twisted together ; then the joint was carefully 

 varnished and plunged in water coming from the tap, so as to 

 prevent the production of any thermo-current. 



One of the two conducting-wires was completely insulated; 

 for we had occasion to confirm an observation already made 

 by Lamont^, that perfectly dry wood is no insulator for 

 feeble currents. 



* Loo. tit. p. 74. t Loc. cit. Memoir, 10, p. 341. 



\ Lamont, Der Erdstrom, Leipzig (published by Voss), p. 7. When 

 the one wire of our galvanometer was connected with the water-pipes, 

 and the end of the other (insulated according to ordinary views) was held 

 in the hand, a considerable current was produced in the apparently open 

 circuit, producing a deflection of nearly 100 divisions of the scale. 



