Affinities of the Elements of Water. 307 
0010 ee .. YS S10 Season eee: ==)202".0 
PEO mice wet, “asian cn (Ok gee Ne tla ulon- sia 2OD <0 
Bg eh ge-8)\2s 0) weds Lil arate epee re 202°: 0 
Meta erate) Psi) ts ie EOE tree ee ee OD 
The probable error of the mean e here is + 4.6; the set is more consistent, and 
gives the change of e = 67.9 for 140°, or 48.5 for 100°. The change of /.o is 
24.2. 
One only objection occurs to me against this conclusion: it may be said that 
heat acts merely by expanding the evolved gases, and thus facilitating their 
escape from the electrodes. If so a similar effect should be produced by removing 
the pressure of the atmosphere ; and, to try this, I arranged the voltameter under 
the receiver of an excellent air-pump. Some difficulty was found from the rapid 
effervescence which occurs at the first contact, and would, without precaution, 
blow the fluid out of the voltameter. This was obviated by adding 700 to the 
circuit, removing 400, and then the rest; thus moderating the initial burst of 
gases. The pressure was reduced to 0'.6 before each observation, but the extri- 
cation of gas increased it to about 1'.1 before the deflection was read. The 
barometer = 29.55; temperature of room = 64°.6. The results are : 
BODO Oe ere Y= IA Si, ee) Be 
Gage Rr. ee et Oe ne 
GOGMEM Rint Shas) es eA Scomeet oF pe are 
DO POMEL. Cs a ale Bon os ce ee ee mee ee 
GURNEE, re cea AO aS ye ee te eee 
OOM MER site hte DUS) Saket ci +) etiam eee 
iO Ameer rs a DOS el is | can eee 
GUS Deeb. a AOR vg beeen LO 
SOG Mec ee (AOS oy ae mmr gel 
SOD OM tes ees (ASO Sar mmm ee les 
DON iee gern agers 2) nt VAG Senate ee mee cor dL 
* The conducting power of the water also increases with the heat; but y is, of necessity, less 
accurately determined than e. It almost seems as if the resistance of it would become insensible 
by a little elevation of temperature. 
+ The connecting wires were not sufficiently amalgamated. ‘This does not, however, affect ¢ ; 
it was corrected immediately. 
