1889.] A Magnetic Survey of the British Isles. 547 



exhaustive survey of the United Kingdom. They have observed at 

 200 principal and a number of secondary stations, and at all the 

 principal stations, except three or four, all the magnetic elements 

 have been determined. 



The two sets of instruments employed have been carefully com- 

 pared with each other at the Kew Observatory in 1884, 1886, and 

 1887, and are in remarkably good accord. All the observations were 

 made by the authors except the dip observations at eight stations in 

 Scotland, for which they have to thank Mr. A. P. Laurie, Fellow of 

 King's College, Cambridge. 



The chronometers were frequently compared with Greenwich by 

 means of the 10 A.M. time signal, for leave to receive which the 

 authors are much indebted to the good offices of Mr. Preece, F.R.S. 



The probable errors of the observations are as follows : 



Decimation 0''699 



Horizontal force + 0*00028 (M.Y.) 



Dip + 0''51 



In this computation, only declination observations which are in all 

 respects independent are included. The horizontal force observa- 

 tions are also as independent as is possible for nearly simultaneous 

 observations, and the dips compared are those taken with the two 

 needles. 



The authors propose the name isomagnetics for the class of curves 

 which are drawn through points at which the values of one of the 

 magnetic elements are constant, and in which isogonals, isoclinals, 

 &c., are included. 



To determine the form of the isomagnetics, they divided the area 

 of the survey into nine overlapping districts, and found for each a 

 linear formula which connected the value of the element with the 

 latitude and longitude. By means of this formula they calculated 

 the value of the element at points where the lines of longitude corre- 

 sponding to whole degrees east or west of Greenwich intersect lines 

 of latitude which correspond to half degrees. Where several 

 districts overlap, the mean value was taken. From the values thus 

 obtained at a series of points regularly distributed all over the 

 country, the isomagnetics were approximately determined. The 

 forms of these curves were slightly irregular, and equations were 

 framed to represent smooth curves which passed through their 

 mean directions. These were the equations to the terrestrial or un- 

 disturbed isomagnetics. 



These are compared with those obtained in the earlier surveys, and 

 the secular change is fully discussed. 



The calculated values of the elements are then obtained for every 

 station, and by comparing these with the observed values, the magni- 



