[ Bo 1 
dipping-needle, are the lureft way to complete the mag- 
netic theory of this globe, analogous to the method e 
purfue in examining the tcr<>^eUa. But as all the dipping- 
needles which I had ieen, appeared to me to be very ill 
calculated, for the fea fervice at lead, 1 contrived one 
upon a different plan in 1764, and had it executed be- 
fore I left England, by Mr. sissox. 1 have called it an 
Lhiiverfal Magnetic Xeedle, or Obfervatir 1 Compafs; 
becaufe I can by it take the dip and amplitiidc, and even 
the azimuth, with only one aliilfant, to take the altitude 
For me. The needle is of the fame fliape and lize nearly 
as thofe ufed now for the compaffes of the royal navy, 
and plays vertically upon its own axis, which has two 
conical points, llightly fupported in two correfponding 
hemifpherical fockets, which are inferted into the op- 
jiofite fides of a fmall upright brafs paralellogram, about 
■one inch and a half broad and fix inches high. Into this 
parallelogram is fixed, at right angles, a flender brafs 
.circle, about fix inches diameter, filvered and graduated 
to every half degree, upon which the needle fliow^s the 
dip, by a vernier if you choofe ; and this, for the fake of 
ditfindtion, I lhall call' the circle of magnetic inclina- 
tion. This brafs parallelogram, and confequently the 
circle of inclination, alfo turns horizontally upon two 
other pivots, the one above and the other below, with 
xorrefponding fockets in the paralellogram. Thefe pi- 
(^a) Mr. SISSON thinks, that thefe fockets were conical as well as the ends of 
the axis, but more obtufe than them ; which feems moft likely to be the cafe, as 
they feem much more likely to anfwer well than hejnifpherical fockets. 
/ 
2 
vots 
