ELECTRICITY, MAGNETISM, AND GRAVITATION. 



161 



indicates 4°. More clouds intervene directly, and it returns 

 to zero again. 



December 15th., at 10 a.m. The sun is now in a direct 

 line with the helix. The indicator 5° to the right hand. 

 Placed a screen between the sun and the instrument, and 

 in a few seconds it has returned to zero. 



It may be noticed that in all these experiments the sun's 

 rays had to penetrate the glass of the window as well as 

 the glass shade of the instrument. 



I now used a large lens to condense the rays upon some 

 of the wires at the south end of the helix, but found that 

 this condensation deprived the rays of their influence upon 

 the galvanometer. 



That the results were not due to thermo-electric action is 

 manifest. In these experiments there was no electric circuit 

 formed. There was no wire of communication between the 

 two mercury cups forming the terminations of the helix, and 

 therefore, upon all commonly admitted principles, there could 

 be neither ordinary voltaic, or thermo-electric phenomena 

 manifested, for the latter are never seen unless the electrical 

 circuit is complete. 



I now completed the circuit by a connecting copper wire, 

 one extremity of which was introduced into each mercury 

 cup. When the sun became partially unclouded, the gal- 

 vanometer indicated 3*^. I removed the connexion, and 

 completed it again and again several times, but without any 

 sensible change. 



It is now 4°. I break the circuit, but no effect ensues ; 

 still 4°. The sun shining brightly, at lOh. 45m., 5°. 

 The wire of connexion is removed, and after one minute 

 the galvanometer still indicates 5'^. Clouds intervene, 

 and the needle declines. The introduction of the con- 

 necting wire produces no alteration during the decline of 

 the needle. 



Shortly, although clouds of a fine texture intervene, the 



