94 Col, Beavfoy^s Summary of the Magnetical and [Feb. 



flannel upon the hole at the joining of the tube, and upon this I 

 poured tne gunpow^der, using two, three, four, and sometimes 

 five grains at once. After each trial, I found a scorched brown- 

 ish mark in the centre of the flannel about the size of the hole of 

 the tube. 



A variety of experiments were also performed with flannel, 

 paper, and other substances, placed between the joinings ; in all 

 which cases it was found that the flannel had been forced 

 through, generally leaving a hole in the substance used ; and a 

 pale-coloured flame was observed to dart to a considerable dist- 

 ance below the bottom (B) of the tube. 



In a future paper, I hope to have the honour of laying before 

 the Society a fuller account of these curious experiments. In 

 that paper I propose to enter upon the cause of the results which 

 present themselves ; and more particularly to attempt an expla- 

 nation of that extraordinary one wherein we have the gunpowder 

 remaining apparently inert to the flame which passes through it. 



Article III, 



Summary of the Magnetical and Meteorological Observations, 

 during a Period of Three Years and Nine Months, By Col. 

 Beaufoy, F.R.S. 



(To Dr. Thomson.) 



MY DEAR SIR, BusJiey Heathy Stanmore^ Jan. 1, 1821. 



In the Annals of Philosophy for May, 1820, you did me the 

 favour of publishing a table containing the monthly mean varia- 

 tion of the magnetic needle for three years ; and in the letter 

 accompanying it, I expressed my opinion that the maximum of 

 the western variation at this place occurred in the month of 

 March, 1819. With the view of demonstrating that this conclu- 

 sion had not been precipitately drawn, I continued the observa- 

 tions to the end of last year ; and as the corresponding monthly 

 mean variation in every case shows a diminution, I infer that the 

 variation has been retrogade for the last 21 months. 



In the latter part of the protectorship, the true and the mag- 

 netic meridians coincided. If 24° 4 V 42", the greatest variation, 

 be divided by 162 (^the number of years since that period), the 

 quotient 9' 09'^ will be the mean annual increase ; it is reasonable 

 to suppose, therefore, that after the same lapse of time (from 

 1819) these meridians will again coincide ; but by inspecting 

 Table II. it will be seen that the mean annual decrease in lieu of 

 being 9' 09" is V 57''; consequently, an acceleration must take 

 place, or the supposition is erroneous. 



I remain, my dear Sir, very sincerely yours, 



Mark Beaufoy* 



