C 96 ] 
has oftentimes been defeated by the index glafs being 
bent by the brafs frame that contains it: to prevent 
this, I have contrived the frame, fo that the glafs 
lies on three points, and the part that prefles againft 
the front of the glafs has alfo three points exa&Iy 
oppofite to the former. Thefe points are made to 
confine the glafs by three fcrews at the back, that 
a< 5 t exactly oppofite to the points between which the 
glafs is placed. This little contrivance may be of fome 
ufe ; but the principal improvements are in the me- 
thods of adjufting the glafies, particularly for the 
back obfervation. 
The method hitherto pradtifed for adjufting that 
part of the inftrument, by means of the oppofite 
horizons at fea, has been attended with fo many 
difficulties that it has fcarce ever been ufed ; for fo 
little dependance could be made on the obfervations 
taken this way, that the beft Hadley’s fextants made 
for the purpofes of obferving the diftances of the 
Moon from the Sun or fixed ftars, have been always 
made without the horizon glafs for the back obfer- 
vation ; for want of which, many valuable obferva- 
tions of the Sun and Moon have been loft, when 
their diftance has exceeded 1 20 degrees. 
To make the adjuftment of the back obfervation 
eafy and exaift, I have applied an index to the back 
horizon glafs, by which it may be moved into a 
parallel polition to the index glafs, in order to give 
it the two adjuftments, in the fame manner as the 
fore horizon glafs is adjufted. Then, by moving 
the index to which the back horizon glafs is fixed, 
exadtly 90 degrees (which is known by the divifions 
made for that purpofe) the glafs will be thereby fet 
6 at 
