164 <lttempt to fucititate Observations of 
From the first of these Tables we see that the values of the dip, obtained by the 
common and counterpoised needle, do not differ from one another by 16’, at any of 
the four stations ; while at two of the stations the difference is considerably within the 
limits of error of a single observation of dip, taken in the usual manner. 
The general agreement of the results of the two methods being thus established, the 
second series of observations, given in Table I, enables us to compare the methods 
with respect to accuracy ; for these observations being all taken at the same place, the 
consistency of the results will afford the means of estimating the probable limit of 
error in each method. 
Comparing then, the results of Table IT, it will be seen in the first place, that the 
mean of the four observations made with Needle I, gives 70° 55’.1 for the dip, in the 
Philosophy school of Trinity College ; while the mean of the observations taken 
with Needle H, is 71° 0.9, differing from the former by less than 6’. 
The results of Needle LI, are, however, by no means as consistent among themselves 
as those of Needle I. The greatest difference between any one result and the mean, 
with the common needle, is only 24’; while, with the counterpoised needle, the cor- 
responding difference amounts to 22.’ This, however, can scarcely be regarded 
as decisive against the accuracy of the method. The common needle was one of re- 
markable nicety of construction ; while that used with the counterpoises, was obviously 
erroneous. It was necessary for the consistency of the results that the points of appli- 
cation of the counterpoises, should coincide exactly with the magnetic axis, or cen- 
tral line of the needle. Now, no yery precise means were adopted to insure this coin- 
cidence, in drilling the holes for the counterpoises; and that the coincidence was in fact 
not effected, will readily appear on looking over the results. It will be seen that, when 
the second counterpoise is employed, there is a difference amounting sometimes to 4° 
between the angles read off with the face of the needle in opposite positions; and 
this plainly indicates an error in the place of the second hole on the northern 
arm of the needle. It is to be observed, further, that the magnetic state of the nee- 
dle was widely different in the several observations, having been altered before each 
trial by a pair of bar magnets ; so that the method was subjected to a more severe 
test than any to which it could be exposed in practice. Even with these disadvan- 
tages, however, the values of the dip obtained with this needle do not appear to 
differ from one another more than is common in such observations ; so that the result 
of the trial cannot be regarded as unfavorable. 
During the progress of the observations recorded in Table I, I made a cotempora- 
neous series of observations on the rate of vibration of Needle II, suspended horizon- 
tally ; with the view of ascertaining how far the results obtained with this needle could 
he relied on for the determination of the force. For this purpose the magnetism of 
the needle was altered by a pair of bar magnets, and its rate of vibration ascertained 
in the usual manner, after each change in its magnetic condition, by a good chrono- 
