92 Mr R. Adie's Experiments with Thermo- and 



spring advanced. From May to the 21st June the two opposing cur- 

 rents were on the whole period so nearly counterbalanced, that I could 

 discern no action ; but at the beginning of May the night-current slightly 

 preponderated. In July, August, and September, the day-current was 

 slowly depositing silver. October there was no perceptible action. 

 November the current due to astral radiation again begun to deposit 

 silver, and this action may be expected to continue till next spring or 

 summer. 



These results differ considerably from the experiment where the vol- 

 tameter was exposed like the battery ; they shew a much feebler action 

 through June and July, when the sun was nearly two-thirds of the time 

 above the horizon, than in the first experiment (19) ; in it the opposing 

 currents developed by day showers, and by astral radiation, are nearly 

 powerless, on account of the reduced temperature of the voltameter (11). 

 In the other, where the decomposing cell is maintained at an equal and 

 elevated temperature, the current caused by day showers, or by astral 

 radiation, is as effective as the solar radiation current. To the showery 

 unsettled weather of May, June, and July, I can alone attribute the mi- 

 nute action observed ; for in September, when the weather had changed 

 to be remarkably fine, the solar current became more effective in depo- 

 siting silver than it had been in June. 



22. The galvanometer permanently attached to a thermo-electric bat- 

 tery (20), was at mid-summer exposed for two or three hours every 

 morning to the solar rays ; this deflected the needles of the instrument 

 as much as 20°, when the helix was not connected with the battery. M. 

 Becquerel and others have already noticed similar deflections, in experi- 

 ments of this kind, where they were found to be caused by small upris- 

 ing currents of heated air ; but, at the time, as I thought that a change 

 in the magnetic intensity of the upper needle, from its elevated tempera- 

 ture, and also local currents in the copper coils of wire, might operate to 

 produce a part of these deflections, I removed first the coils of wire, 

 which made no change j aiext the diaphragm or compass card interposed 

 betwixt the two needles, so that the rays of the sun now acted equally 

 on both of them ; this alteration increased the deflection, shewing streams 

 of heated air rising from the upper needle to be the active force which 

 produced the deflection of 20° in the complete instrument, when in the 

 sun's rays. 



23. While engaged with the last experiments, they appeared to me to 

 point out where to look for the explanation of a curious instrument, Mr 

 Watt's sun needle, which I had seen many years before, under my father's 

 care, rotating by the action of the solar rays ; and as I am not aware 

 that the cause of this rotation has yet been shewn, I will briefly state 

 the result of some experiments, which refer it to small uprising streams 

 of air, generated in their highest energy on metallic surfaces. 



Mr Watt's instrument consists of a light piece of wood, with agate 

 centre, in form resembling the needle of a mariner's compass ; on the 



