62 



Mr. G. J. Symons. On Barometric Oscillations [May 8, 



M. Teisserenc de Bort and Mr. Budd have suggested that the rise 

 may be due to the local compression of the air by the multitude of 

 falling rain drops. 



In the ' Annales ' of the French Meteorological Office for 1880, 

 M. Benou gives reproductions of some barograms from the Obser- 

 vatory at Pare St.-Maur. A very interesting one is reproduced in 

 fig. 1. M. Benou does not append any remarks to the plate ; but 



FIG. 1. 



Pare St.-Maur, Paris, August, 1878. 



from other sources it appears that there was a heavy thunderstorm 

 from 10 till 11 P.M. on August 18th, 1878. The total rise may be 

 taken as O'lO in. and the fall as nearly O'lo in. 



Dr. Fines, of Perpignan, established a Bedier barograph in 1875, 

 and in a memoir published in ] 883 gives reproductions of the traces 

 during several storms. He states that, considering the present im- 

 perfect knowledge of the real conditions and origin of thunderstorms, 

 it seems useless to try to explain records frequently influenced by 

 very distant storms. One fact, however, is certain, that thunder- 

 storms are accompanied by great condensation, which must cause 

 variations in the density of the air, and therefore should affect the 

 barometer. He gives - 144 in. in 20 minutes as the greatest varia- 

 tion, but says that it rarely exceeds 2 mm. or 3 mm. (0'08 in. or 

 0'12 in.) in an hour. The only facts which are certain from the 

 records at Perpignan are that there is usually (1) before heavy rain, 

 decrease of pressure and temperature ; (2) with the rain, sudden 

 increase of wind, rapid rise of barometer, and fall of temperature ; 

 (3) at the end of the storm-rain, reversal of the last three phe- 

 nomena. 



Dr. Ciro Ferrari, who has devoted great attention to the progress 



