80 
MR. A. W. RtiCKER AND DR. T. E. THORPE ON A MAGNETIC 
Instrument 83. 
Observers. 
Number of stations, with 
difference. 
Bias. 
Upper sign, Needle 1. 
Positive. 
Negative. 
Lower sign. Needle 2. 
Scotland . 
T. and R. 
6 
3 
+ 0T9 
England 
T. 
25 
16 
+ 0-20 
Ireland. 
T. 
11 
7 
± 0-20 
An analysis of tlie errors and corrections such as we have carried out appears to 
constitute a very delicate test of the observer and instrument combined. Thus it is 
curious to note that (although the differences are small) one of us has on the whole 
been more successful with the Declinations and the other with the Force observations. 
As has already been remarked, the observations in Scotland were often divided between 
us, so that we do not attempt to analyse them, but both in England and Ireland 
Dr. Thorpe’s Declinations agree the better by about 0'’26, and Professor Pucker’s 
Forces by about O'OOOll metric unit. The facts that in Ireland Professor Pucker’s 
mirror appears to have been generally in the better adjustment, and that in England 
Dr. Thorpe’s observations of the period of oscillation of the needle were the better, 
indicate that the causes of the differences in the results were probably a more 
perfect management of the torsion of the thread in the Declination observations, and 
a slight superiority in the deflection observations in the Force measurements. 
It will, we think, be granted that in all cases the observations are as good as the 
occasion requires, and we discuss these minute diflerences not because we think they 
produce any appreciable effect on the results of our survey, but as an illustration of 
the fact that our method of tabulating the observations, conveys information as to the 
personal equations of the observers which is not afforded by the results as ordinarily 
presented. 
We think it possible that if we had been able to carry out the work in a more 
leisurely manner, to observe azimuths only when the Sun was low in the heavens, 
and to wait for fine weather, the results might have been in closer accord, but on the 
other hand we do not think that any improvement so obtained would have been of 
practical importance. The uncertainty of the value of the secular correction at any 
given station, and the changes produced in disturbed districts by a slight alteration 
in the position of the instruments, are far more important than any residual errors 
by which our observations may possibly be affected. 
