Silbermann asserts that Auroræ are produced by the same general phenomena as thunder-storms, and concludes that the Auroræ of 1859 and 1869 assumed the character of thunder-storms which, instead of bursting in thunder, had been drawn into the upper parts of the atmosphere, and their vapour being crystallized in tiny prisms by the intense cold, the electricity became luminous in flowing over these icy particles. [Sidenote: Prof. Piazzi Smyth on monthly frequency of Auroræ and storms.] Professor Piazzi Smyth has observed that the monthly frequency of Auroræ varies inversely with that of thunder-storms. His Table of comparisons is as follows:— [Sidenote: His table of observations.] Month. Lightning. Auroræ. January 24·0 29·7 February 14·4 42·5 March 7·0 35·0 April 15·4 27·5 May 37·4 4·8 June 48·0 0·0 July 55·2 0·5 August 38·4 12·6 September 22·4 36·6 October 20·8 49·4 November 15·0 32·4 December 15·0 28·8 ---- ---- Mean of whole year 24·0 20·1 [Sidenote: Silbermann’s observations 15th April, 1869. 30th April, 1865.] Silbermann, on 15th April, 1869, observed a fall of rain (tiny crystals of ice) on the disappearance of an Aurora and its change into cloud forms (see section, “Auroræ and Clouds,” p. 53). He also observed a rain of little sparkling ice-prisms on 30th April, 1865, at Paris, the city being then enveloped in a cirrus of vertical fibres similar to that which frequently accompanies the Aurora. On the occasion of the Aurora seen by me at Guildown, 4th February, 1872, rain fell immediately succeeding the formation of the corona. The falling of rain as an immediate sequence of an Aurora seems, however, to be rather the exception than the rule; but possibly this may vary with the character of the Aurora itself—whether it be of the crimson class, passing into cloud and accompanied with much electric disturbance, or of the more quiescent white. [Sidenote: A falling barometer observed to follow Auroræ.] A falling barometer following a display of Auroræ has been noticed by Sir John Franklin and others; and in some cases (notably one in Sicily before referred to) storms and floods have accompanied this. [Sidenote: Professor Christison’s observations.] In a paper read before the Royal Society of Edinburgh in 1868, Prof. Christison mentioned, as a fact of importance to agriculturists, that the first great Aurora after autumn is well advanced, and following a period of fine weather, is a sign of a great storm of rain and wind in the forenoon of the second day afterwards.