ON THE MAGNETIC SURVEY OF ENGLAND. 259 



instrument emploj'ed in the earlier survey was a circle and two needles by 

 Gambey, free from any appreciable error, and in the later survey. No. 30 

 and its two needles already described : the results were as follows : — 



Margate. Lew Trenchard. 



Lat. 51° 23' N., Long. 1° 23' E. 



Nov. 9, 1837 69° 02'-9 



Oct. 5, 1860 68° 05'-9 



Lat. 50° 39' N., Long. 4-° 11' W. 



July 30, 1838 69° 19'-0 



July 9, 1859 68° 11 H 



Secular change in 21 years 1° 01'*6 

 Annual change 2''93 



Secular change in 22*9 years 57''0 

 Annual change 2''49 



This comparison shows that the mean annual secular change in the interval 

 between the surveys was a decrease of 2'"49 at Margate on the east coast, and 

 of 2'*93 at Lew Trenchard on the borders of Devonshire and Cornwall. 



The amount of the difference in the rate of secular change on the east and 

 west coasts corresponding to the change in the value of u, may be further 

 and more fully exemplified by comparing the values of the dip at the two 

 epochs 1837 and 1860 at Lowestoft on the extreme east of England, and at 

 the Land's End at the extreme west ; the values in 1837 being taken from 

 the map of the isoclinal lines for January 1837 accompanying the report of 

 the survey of that epoch, and those in 1860 being computed by the formula 

 obtained by the survey of 1860. 



Lowestoft. Land's End. 



Lat. 52° 30' N., Long. 1° 45' E. Lat. 50° 05' N., Long. 5° 40' W. 



Dip in the Isoclinal map 1 o / o < 



of 1837 J 69 34.-5 69 21-0 



Dip in January 1860, com- ' 

 puted by the formula 

 0=68°59-2+O'-1993a 

 + 0'-591l6 



168 40-2 68 10-0 



Secular change in 23 years 54*3 1 11*0 



Annual change 2*36 3*09 



If we now bring together the values of the annual secular change during 

 the 23 years preceding 1860 as shown by these four comparisons, placing them 

 in order from East to West across our island, and introducing in its proper 

 place 2''63, the annual secular change at Kew in the 21 years preceding 1859 

 as known from other sources (Proceedings of the Royal Society, vol. xi. p. 158), 

 we have as follows : — 



Lowestoft 2-36 



Margate 249 



Kew 2-63 



Lew Trenchard 2-93 



Land's End 3-09 



The increase in proceeding from east to west is shown consistently. The 

 annual values derived from determinations including intervals of above 

 20 years, are of course mean values. The surveys furnish no direct means 

 of judging whether the secular change has been uniform or otherwise at any 

 of the stations. At one of the stations only, i. e. Kew, we have reason to 

 believe, from the observations recorded in the ' Proceedings of the Royal 

 Society ' referred to above, that the change has been uniform during the 

 whole period from 1837 to 1860 (and also for several years preceding 1837) ; 

 but we are not entitled to assume a similar uniformity at any of the other 

 stations. 



Proceeding now to the increase in the value of r in the interval between 

 the two surveys, — the difference in the rate of secular change of dip in the 



s2 



