Terrestrial Magnetism. : 321 
stood. 'There is certainly no better authority for the dip at any 
particular time and place, than careful observations made with 
the best instruments. When artificial lines do not agree with 
observations, it is evident that those lines should be marked “zn 
error.” ‘The lines of equal dip, as obtained by actual survey, 
are not great circles nor uniform curves; they undulate irregu- 
larly, converging in some places and diverging in others, and 
sometimes, I believe, one line of equal dip will divide into two 
which will afterwards reunite. It is perhaps customary with 
those who make magnetical charts, to endeavor to equalize these 
natural irregularities, as the engineer after a survey for a road 
equalizes the hills and. hollows to obtain a less devious but more 
artificial surface. Such a line, although it is easily projected, and 
looks well in the chart, has no existence in nature, and is only 
an artificial approximation to truth; so far as it departs from the 
results of actual observations, the line itself should be marked in 
error, not the observations. ‘This, I say, would be philosophical ; 
a conventional mode of expressing the same relation in different 
terms may obtain a preference, and would be unobjectionable pro- 
vided it should be properly understood. 
The largest number of my magnetical experiments were made 
in Iowa and Wisconsin Territories during last autumn, and the 
general results, including both the dip and intensity, have been 
communicated both to Congress and to the American Philosophical 
Society. But I am reminded by the above suggestions that I 
ought to lay before the public a specimen of the details at large, 
that a judgment may be formed of the degree of credit to 
: ee ae dip. Be! a } 
compass was made by Robinson, of London, in 1837. Asa 
check upon errors, I make at each station, by means of two sep- 
arate needles, a double suite of observations. In each suite, all 
of the usual reversals are made, including the face of the instr 
ment. the face of the needle, and the polarity of the needle by 
saicochini: The dip is therefore determined by eight distinct 
readings of each needle; the two results almost always agreeing 
within one or two minutes of a degree. The mean of the whole 
of the sixteen readings is finally taken. The following are ex- 
amples : 
