o6 
of nature, the writer substitutes a restricted potential 
of his own imagining. 
(2) Having chosen to deal with an elementary portion of 
potential of nature, he fails — in comparing it with 
observation — to use, on the other side, only the 
corresponding elements of the variations of force, 
but compares with the whole complex variation of 
observation. 
(3) He describes such agreement of observation with his 
theory as is displayed in his comparisons in un- 
measured language. 
(4) He draws conclusions on questions of importance that 
are unwarranted by the evidence produced ; and he 
expresses them with a degree of confidence that 
would mislead any reader but the specialist, who is 
able to appreciate the evidence at its true value, into 
a belief they are better founded than they really are. 
As to the practical object which I have in view in thus 
exposing the illogical character of the writer’s treatment of 
this question, besides the correction of published fallacies, 
I may mention that Hr. Schuster is a member of a Com- 
mittee of the British Association appointed to consider the 
best means of comparing and reducing magnetic observa- 
tions, and in that capacity— as in the paper under dis- 
cussion-“he has advocated the application of Gauss’ 
method to the magnetic diurnal variations; and, not agree- 
ing with him that there is urgency in this point, I deem it 
important that the rest of the Committee should be made 
aware that his authority in this field does not stand un- 
questioned. My own opinion is that the effectual applica- 
tion of the harmonic analysis to the diurnal variations of 
terrestrial magnetism ^^ill be a work of such magnitude^ 
* How formidable an undertaking, even if the potential include only terms of the 
fourth degree— as in Gauss’s calculation of it from the absolute forces— may be inferred 
from the fact that, taking each of the periodical factors with four pairs of terms, there 
would be no less than 192 arbitrary constants to determine for the surface potential, 
and each of these would divide itself into two when the variations of vertical force were 
taken nto account. 
