a 
Improvement in manufacturing Magnetic Needles. 15 
that while snow was melting away, these fits were the most 
frequent. 
But on comparing different compasses, I found that they 
frequently varied, not only from the common direction of the 
magnetic needle, but from each other. For example, when 
set in some directions, one compass would vary, while other 
Compasses would vary when set in different directions, and 
would not vary when set in the same directions. On exten- 
sive alluvial plains, where we could not suspect the presence 
of extensive ore-beds, all these difficulties occurred with equal 
force. And what appeared to be a still greater mystery, on 
changing needles, the variation seemed to be governed by the 
compass, not b the needle: 
After considering every proposed hypothesis, and trying 
every proposed remedy, I abandoned the subject, as totally 
inexplicable ; and contented myself with correcting these 
aberrations by ranging back-flags and using two compasses 
While exercising the students of Rensselaer school in land 
surveying, at the last summer term, the same difficulties re- 
vived the same enquiries. In a conversation with an ingen- 
ious artist, Mr Julius Hanks, of Troy, I learned that his best 
compasses had in some instances, been subject to those fits 
of aberration. He showed me a compass of most elegant 
and accurate workmanship, with a nonius and double levels, 
which had been returned by the purchaser on account of the 
frequency of those fits. I carried this compass to the school 
with a determination to search out if possible, the cause of 
its frequent fits. By applying ehneely suspended needles, | 
which might be called a suit of magnetrometers, I found a 
point in the limb, which attracted a fine needle at the distance 
of six tenths of an inch. This point caused the needle be- 
longing to the compass, to deviate at the distance of half an 
inch on each side; beyond that limit it was not affected. 
Consequently, when the course to be tak en brought the nee- 
dle within that limit it would deviate, and accurately in 
all other directions. Any practising surveyor will readily 
perceive, that in tracing the lines around a ‘field, the needle 
might come within this limit several times, or it might not 
fall within it in ranning a dozen fields. Hence the supposed 
irregularity of the fits. 
y conclusions from these experiments were, that a scale 
from a secrew-cutter or a punch, or a tooth from a file, &c. 
too minute for the eye, might have been lodged in that par- 
2 
® eet 
He 
