20'8* VARIATION OF THE COMPASS. 



Table II, contains the mean monthly variation for the ahove^ 

 mentioned times of the day contained in Table I. 



Table III, contains, besides the mean monthly true varia- 

 tion, and mean monthly diurnal alteration of variation, for the 

 sixteen above-mentioned months, the mean monthly true va» 

 riation, and diurnal alteration of variation for many months irt 

 fheyear, between the years 1786 and 1805 inclusive. 



The numbers put down in Table I, are eachofthemamean 



©f five observations, and often more. 



Those in Table II. depend on Table I. 



As the observations from which the true variation has been 



given in Table III, between the years 1788 and 1805, vi^ere 



t-oo numerous to be all inserted, it has been thought sufficient 



to give the mean monthly true variation, and mean monthly 



diurnal alteration of variation only; and they were determined 



from a mean of the observations made at those times of the 



day when the variation was considered least, and greatest; 



which variations for each month, may generally be considered 



as a mean of 600 observations. 



Remarks on From the observations made by the late Dr. Heberden and 



the changes of others, about the year 1775, the variation was found to increase 



t^e variation, annually nearly 10'', since that time to the present, its rate of 



increase has been considered as gradually diminishing*, and 



for 



* An exception to the progressive increase appears between the 

 years 1790 and 1791, as the observations between these two years 

 make it to decrease 2'' or 3', and subsequent observations to increase 

 again. To what this should be attributed, I am at a loss to account, 

 unless it arose from the alteration which took place in the iron-work of 

 the room in December 1790, four strong iron braces having been ap- 

 1 lied to the girders in the floor of the great room of ihe Royal Academy 

 (which is over ihe Society's meeting-ioom), in consequence of 3 

 cracking noise made from the great pressure of a number of persons in 

 the room during the time that Sir Joshua Reynolds was delivering a 

 lecture : these braces were apphcd two on each side of, and equidis- 

 tant from, the compass, the ne.irest about 18 feet from it. It may be 

 proper to mention, however, that having been favoured with the varia- 

 tion observed both by Mr. Cavendish and Dr. Heberden, in the above- 

 mentioned years, the alieration of the variation was by the former 

 nearly the same as in my own, but by those of the latter, greater in 

 both cases. 



An .-^Iteration took place between the observations made with thtf 



