Intelligence and Miscellaneous Articles. 239 



of the difference of potentials. This explains why the phenomenon 

 is almost imperceptible at small potentials, while with greater ones 

 there is a regular, and, under favourable circumstances, a very con- 

 siderable depression. 



Some experiments which I have made seem to confirm the 

 theoretical law, at any rate for not very great depressions. But with 

 greater deflections on the contrary, (from 10 to 40 inillim.), there 

 was apparently a simple proportionality between the increases of 

 difference of potential and the depression. This deviation may, 

 however, be referred to the imperfection of my experiments and to 

 different disturbing actions. 



In order to keep the inner surface of the areometer in conducting 

 communication with a source of electricity, without interfering too 

 much with its mobility, I constructed a kind of hydrostatic balance 

 from fine capillary tubes. A very fine silver wire passed through 

 these tubes, and terminated at one end in the areometer, and at the 

 other in a separate vessel of water. 



It is also very essential that the conducting liquid shall not 

 moisten the areometer-tube, otherwise this rises against the sides 

 on electrification, and the capacity of the condenser is increased 

 without a perceptible depression. If ordinary water is used as con- 

 ducting liquid, the air in the upper part must be replaced by an oil 

 under which water does not moisten the glass. The sensitiveness 

 of the apparatus is at the same time considerably increased. 



After more accurate measurements the comparison of the 

 observed and calculated depressions might give a datum for detei- 

 mining the alterations of capillary forces at high potentials. 



If, however, this influence of the capillary forces were defined 

 and eliminated, it might be possible to use the electroareometer as 

 an absolute electrometer. Its indications could not however lay 

 much claim to great accuracy. — Wiedemann's Annalen. No. 86, 

 1888. 



ON THE SECONDARY ARCS WHICH ACCOMPANY THE RAINBOW. 

 BY M. BOITEL. 



The old theory of Descartes does not explain the formation of 

 complementary arcs. Young, in 1804, was the first to account for 

 the production of these arcs by means of the theory of interference. 

 Airy showed afterwards how the position of these arcs could be 

 determined by applying the principles of diffraction, and he fixed 

 the position of the two former. Miller proved experimentally that 

 Airv's theory gave very approximate results for the first two arcs. 

 Stokes, lastly, made the calculation by a more rapid method than 

 that of Airy, which enabled him to obtain the position of the first 

 fifty bands. * 



But Airy in his calculation assumed that the equation of the 

 emergent wave (which is shaped like the letter S) in reference to 

 the tangent, and the normal to the curve at the point of inflexion 



a? 

 y = ^, and he has not given the expression for the constant a 2 . 



