596 



SCIENCE. 



[N. S. Vol. I. No. 22 



§ 16. A small quantity of fresh water or 

 salt water shaken up vehemently with air 

 in a corked bottle electrifies the air, fresh 

 water negatively, salt water positively. A 

 ' "Winchester quart ' bottle (of which the 

 cubic contents is about two litres and a 

 half), with one-fourth of a litre of fi'esh or 

 salt water poured into it, and closed by an 

 india-rubber cork, serves veiy well for the 

 experiment. After shaking it vehemently 

 till the whole water is filled with fine bub- 

 bles of air, we leave it till all the bubbles 

 have risen and the liquid is at rest, then 

 take out the cork, put in a metal or india- 

 rubber pipe, and by double-acting bellows 

 draw off the air and send it through the 

 electric filter. We find the electric effect, 

 negative or positive, according as the water 

 is fresh or salt, shown very decidedly by 

 the quadrant electrometer; and this, even 

 if we have kept the bottle corked for two 

 or three minutes after the liquid has come 

 to rest before we take out the cork and 

 draw off the air. 



§ 17. An insulated spirit lamp or hydro- 

 gen lamp being connected with the positive 

 or with the negative terminal of a little 

 Voss electric machine, its fumes (products 

 of combustion mixed with air) sent through 

 a block-tin pipe, four meters long, and one 

 centimeter bore, ending with a short insu- 

 lating tunnel of parafiin and the electric 

 filter, gives strong positive or strong nega- 

 tive electricity to the filter. 



§ 18. Using the little biscuit-canister and 

 electrified needle, as described in our ' com- 

 munication ' * to the Eoyal Society ' On the 

 Diselectrification of Air,' but altered to 

 have two insulated needles with varied dis- 

 tances of from a half a centimeter to two 

 or three centimeters between them, we find 

 that when the two needles are kept at equal 

 differences of potential positive and nega- 

 tive, from the enclosing metal canister, 

 little or no electrification is shown by the 



* Proceedings of the Royal Society, March 14, 1895. 



electric filter ; and when the differences of 

 potential from the surrounding metal are 

 unequal, electrification, of the same sign as. 

 that of the needle whose difference of po- 

 tential is the greater, is found on the filter. 

 When a ball and needle-point ai-e used, 

 the efl'ect found depends chiefly on the dif- 

 ference of potentials between the needle- 

 point and the surrounding canister, and is 

 comparatively little affected by oi^posite 

 electrification of the ball. When two balls 

 are used, and sparks in abundance pass be- 

 tween them, but little electricity is deposited 

 by the sparks in the air, even when one of 

 the balls is kept at the same potential as 

 the surrounding metal. [The communica- 

 tion was illustrated by a repetition of some 

 of the experiments shown on the occasion 

 of a Friday evening lecture* on Atmos- 

 pheric Electi'icity at the Eoyal Institution 

 on May 18, 1860, in which one-half of the 

 air of the lecture-room was electrified posi- 

 tively, and the other half negatively, by 

 two insulated spirit lamps mounted on the 

 positive and negative conductors of an elec- 

 tric machine.] 



(2) 'ON THE THEE3IAL CONDUCTIVITY OF 

 ROCK AT DIFFERENT TEMPERATURES.' 



Experiments by Lord Kelvin and Mr. 

 Erskine Murray were described, and the 

 apparatus used in them was shown, by 

 which it was found that the thermal con- 

 ductivity of specimens of slate, sandstone 

 and granite is less at higher temperatures 

 than at lower for each of these rocks. The 

 last tested was Aberdeen granite, for 

 which experiments of fau-ly satisfactory ac- 

 curacy showed the mean conductivity for 

 the range from 146° C. to 215° C. to be 86 

 per cent, of the mean conductivity in the 

 range from 81° C. to 146° C. They hope 

 to send a communication to the Eoyal So- 

 ciety describing their work before the end 

 of the present session. 



* ' Electi-ostatics and Magnetism, ' x vi. , ^ ? 285, 286. 



