MAGNETIC SUBVKY OF GREAT BRITAIN. 123 



sio-ned to the point of less than one mile in latitude, and half a 

 mile in longitude. We may hence estimate the probable limits 

 of inaccuracy in the extreme cases alluded to. It is obvious that 

 ^^•hen the fractional minutes in the observation are less than 

 thirty, these limits are proportionally reduced ; and it is further 

 plain that errors thus occasioned will be of a contrary nature to 

 each other, according as the fractional minutes are in excess or 

 in defect of the degree which the line represents. When, there- 

 fore, the observations are numerous, and fall on both sides of 

 the lines, as is the case in this survey, a mutual compensation is 

 afforded, and whatever small inaccuracies there may be in the 

 values of u and r, their ultimate eifect on the lines may be re- 

 garded as wholly insensible. 



If the observations at each station were free from instru- 

 mental defect and local influence, — and if they were continued 

 sufficiently long at each station to furnish its mean dip inde- 

 pendent of diurnal and irregular fluctuations, — the points com- 

 puted from them and transferred to a map would require merely 

 to be connected in order to form the isoclinal line. As might 

 be expected, however, the results of the observations are far 

 from presenting this perfect accordance, especially in Scotland, 

 where the prevalence of igneous rocks produces much disturb- 

 ing action. An examination of the map, however, in which the 

 points, and the stations they are derived from, are inserted, will 

 show that, notwithstanding the disturbing causes referred to, 

 they do arrange themselves in such manner as to leave very 

 little uncertainty in any quarter in tracing the position and 

 direction of each isoclinal line. Each line thus becomes an 

 independent determination, derived from observations which be- 

 long to itself alone, and unhifluenced by those which differ more 

 than thirty minutes from the degree which the line represents*. 



By this method of combination, any departure from system- 

 atic arrangement which might exist in any one of the lines 

 passing across the British Islands, would become manifest at 

 once to the eye. Individual stations there are, particularly in 

 Scotland and the north of Ireland, which throw their points to 

 some distance from their respective lines. In some very few 

 cases, a group of neighbouring stations appears to be similarly 

 affected. The most prominent instance of this is in North Wales, 

 where there appears a decided disposition of the majority of the 



* This has been strictly adhered to in the table everywhere ; and in the map 

 everywhere over the surface of the land. The lines are extended in the map a 

 short distance heyond the land ; and as the observations which justify this ex- 

 tension are few in comparison with those in other parts of the map, the determi- 

 nations which fall nearly midway between two lines have, in these few cases, 

 been given a bearing on the lines on either side of them. 



