194 
MR. LLOYD ON THE DIFFERENCE OF LEVEL BETWEEN 
foot, or a foot (which are the errors most likely to occur), without the trouble 
of bringing' the instrument and adjusting it two distinct times at the same 
spot. The mean of the four and the mean of the two observations ought of 
course to make 10,000 In the detail of the proof-levels, the correction for 
curvature is additive instead of subtractive ; and in fact the whole operation 
is reversed, + standing for — . 
In the course of my levellings, having instruments not generally used, I 
made some few notes, which I take leave here to transcribe. 
In adjusting the station-staff, it is difficult to know when the zero on the 
vane is made to coincide with the horizontal wire of the telescope. I have 
found the most convenient and correct mode to be, to observe 
with the wire forming a diagonal to the lines on the vane, by 
which, when the staff is near, the two small black lines at 
each end of the vane could be seen ; and when the vane was 
adjusted to the proper height, one of these lines was above and 
the other below the horizontal wire at equal distances, thus : 
— At a greater distance, the two little white right-angled triangles, formed by 
the edges of the vane and bounded by the black semicircle, are very distinctly 
seen, the one above and the other below the (now-placed) horizontal wire, and 
can be compared in size with great nicety. 
But in observing with the wire diagonally, great care must be taken that, 
by the vertical wire (the error of which, if any, will by practice be accurately 
known), the axis of the telescope shall bisect the centre of the staff. 
In an instrument with a very sensitive level, there is usually some difficulty 
in adjusting the level. It arises from no fault in the level itself, but from a 
$$$? jj 
1 

* In the proof-levels it will be found, that generally, upon adding them to the mean of the four 
observations, there will be a quantity of from .0010 to .00G0 to make up the 10,000. I could not 
discover the cause for some time. 
The wires being at right angles to each other, of course have been at different distances from the 
eye-glass ; hut the difference not being much more than the thickness of a hair, I did not alter the 
eye-tube : however, upon examination, I found that at the usual distance I took levels, when I 
altered from extreme distinct vision of the horizontal wire to extreme distinct vision of the vertical 
wire (which was the one I used for proofs), it made a difference of from .0020 to .0050, the distinct 
vision of the vertical wire being that quantity lower than when observed with the same adjustment as 
for the horizontal wire. 
