F 
4 
90 Magnetic Dip in the United States. a 
The average of all the preceding errors is +5/.5, which may be 
taken as the probable error of a single reading entirely independ- 
ent of instrumental errors, and the error frequently amounts to 
about a quarter of a degree. What then are we to understand 
by this result? Simply this, that if the instrument be properly 
adjusted, and a number of different readings be made in the same 
position of the instrument and needle, the needle each time being 
raised from its supports and allowed to come to a state of rest, 
the readings will not be identical. ‘They will frequently differ 
+15’ from the mean, and on an average +5/.5. This is the con 
clusion derivable from Prof. Locke’s observations, and the result 
I presume coincides substantially with the experience of all who 
* have undertaken similar observations. My attention has been 
particularly directed to this very annoying and almost disheart- 
ening anomaly, and it has appeared to me that when the agate 
supports and the axis of the needle are carefully wiped clean of 
moisture and dust, the discordance of the readings arises mainly 
from the needle’s slipping upon the agates to the east or west; 
and that when the y’s which elevate the needle are so disposed 
as to allow the least possible motion in that direction, the accord- 
ance of the readings is the best. 
This uncertainty in the readings is of itself sufficient to entitle 
the dipping needle to the character of ‘one very ungrateful in- 
strument.’? Most of the other errors may be corrected by suit 
ble precautions and reversals ; but this cannot be thus annihilated, 
and the only remedy with which I am acquainted is to multiply 
observations. 1am accustomed to make five observations in ea¢ 
position of the needle and instrument, always reading at both 
poles. I thus obtain eighty readings with each needle. 
Another error, and one which equally affects both needles, 
arises from observations being made out of the meridian. At 
Hudson, the dip increases less than one minute from being ob- 
served two degrees out of the magnetic meridian. Where one 
has leisure therefore to determine the magnetic meridian with a 
curacy, this error may be pronounced insensible; but on a tout © 
where observations are usually hurried, the error from this source 
may become important. 
Another source of error is found in the imperfection of the 
axles of the needles. It has long been known that different in- 
struments would give different values of the dip at the same time 
* 
