I 



Thunder-storms, 9.S5 



which fire point-blank, with four others for ricochet firing, 

 with six mortars, and finally with a moveable battery of six 

 gun^. The practisings take place upon certain clays of the 

 week, from seven to ten o^clock in the morning ; and the 

 number of shots which are fired each day are about 150. As 

 the reports are still very loud at the observatory, it appears 

 to me that if they exercise upon the atmosphere that influence 

 which so many people believe, the sky should be more rarely 

 overcast on the days of practice than on the other days of the 

 week ; and this inquiry I have subjected to a minute investiga- 

 tion. 



General Duchan, the commandant of the Vincennes school, 

 has, at my request, kindly sent me an extract of the days on 

 which the artillery practised, from the year 1816 to the year 

 1 835. The total number of these days amounts to 662. Again, 

 the meteorological registers of the observatory furnish me, for 

 each of these days of practice, with the state of the sky at nine 

 o'clock in the morning. Of these ^^9, days, I find 158 during 

 which the sky at nine a.m. was completely obscured with clouds. 

 I put, then, the question — If the cannon had notjired^ would the 

 number have been greater ? 



I believe it will be conceded thati put the solution of this 

 problem upon the fairest possible footing when I take for each 

 preceding day, and for each succeeding the practice, the data of 

 the observatory register already alluded to, and then take the 

 mean of the two numbers as the normal meteorological state of the 

 practice days. By this method we must be entirely free from 

 all possible influence of the noise of the artillery. The follow- 

 ing are the results : — 



Of the 662 days before the practising days, 128 were cloudy. 



... 662 ... which were practising days, 158 



... 662 ... succeeding the practising days, 146 



The mean of 128 and 146, viz. 137, is so much a smaller 

 number than 158, that we might be tempted to conclude that 

 the noise of the artillery, instead of dissipating and dispelling the 

 clouds, actually condensed and retained them. But I know well 

 that the numbers with which we have calculated are too few to 

 allow us to draw any such conclusion. I shall only, then, af- 

 * 



