aa 
We f ‘rez. J 
the legs of a triangle, fo as not altogether to be a much greater 
incumbrance than a common walking-cane. 
As the ftop-cock cannot otherwife be conveniently fecured 
from falling out, a collar of brafs is turned to fit a fhoulder left 
on it for that purpofe, and {crewed over it, as in the Fig. 3. 
Tue greateft inconvenience which occurs to me in the ufe of 
this barometer is the chance of fhaking out too much mercury 
in adjufting it, or the danger of its being difturbed by the wind. 
But a little practice will prevent any uncertainty from thefe 
caufes, and that without much trouble ; befides the barometer 
may be faftened in a perpendicular direction. 
A GREATER objection feems to arife from the danger of 
forcing air into the mercury which is to fill the bafon from the 
bag below. But from the {mall quantity neceflary to make it 
overflow, little inaccuracy feems to be apprehended ; and even 
what is confined with the reft in the box when the inftrument 
is to be carried about will hardly injure to any perceptible 
degree that in the tube, as lefs agitation can take place in this 
than any other barometer; for the air in maffes or fenfible 
bubbles is more completely excluded, and the cavity being lefs 
irregular does not allow any air to lodge in corners as may 
about the floating gages, and various ftop-cocks of thofe on a 
different conftruction. 
AMONG 
