170 MM. Ulster and C4eitel &n 



spirit-flame, in which direct contact of any metal with the 

 flame was likewise avoided. The conduction to earth took 

 place by means of a platinum plate dipping into the alcohol of 

 the lamp, and connected with the earth-conduction by a pla- 

 tinum wire. In addition to the electrode represented in fig. 3, 

 a wet string wrapped tightly round a glass rod was employed. 

 The electromotive force was then determined: — For the 



Water electrode in air . E = 24, 

 Wet string in air . . . E = 58; 

 For a platinum electrode E = 99. 



That lower values likewise result for the latter than in the 

 previous experiments cannot be surprising, since in this 

 arrangement of the experiment there is no second metal dip- 

 ping into the flame itself. 



A direct determination of the combination platinum, water, 

 alcohol, platinum gave a maximum of 11*6 for a daniell = 100; 

 so that the observed electromotive force cannot be produced 

 by this. 



In employing the wet string, care must be taken not to 

 place it tangentially near the flame ; for then small fibrils 

 might project into the flame itself, by which, for the reasons 

 above discussed, a reversal of the polarity of the flame might 

 easily be induced. 



The above-communicated values being so much lower than 

 on the employment of platinum electrodes may be accounted 

 for by the conduction to earth by distilled water or alcohol 

 being always very imperfect. On this account it seemed 

 advisable to examine the behaviour of a water electrode over 

 against a platinum electrode. If the apex of the flame is con- 

 ducted-froin to earth by a platinum wire, the water electrode 

 which is in air is positive only so long as the platinum wire is 

 completely enveloped by the flame ; if it be drawn so far out 

 that it also is entirely in the hot-air stratum, the polarity of the 

 electrodes is reversed — the platinum wire being positive, the 

 water electrode negative. For example, in an experiment of 

 this kind, by the drawing-back the electromotive force was 

 raised from — 142 to + 60 (the signs refer to the metal elec- 

 trode). From this it follows that metals in contact with hot 

 air become more strongly positive than liquids, but that liquids 

 in contact with heated gases exhibit nevertheless a similar 

 behaviour to that of metals. Accordingly we must not directly 

 infer, from the fact that flames show themselves electric even 

 when all metals are avoided, the existence of a peculiar elec 

 tricky of flame. 



The complete analogy of behaviour between metal and water 



