I io8r ) 
by the attradion of the contiguous Annulus at C. For 
rhis, being equal to the upper Annulus at A, is capa- 
ble of llifiaining a Colamn of Water of the length A B, 
and confequently is more than fufficient for fupporting 
the Column of Water below it, CB From which it is 
plain, that no part of the Water contain’d in the Tube 
can poiTibly defcend, unlefs the upper part, aflifled by 
the weight of the Water below it, be fufficient to over- 
come the Attra(5fion of the Annuhis of Giafs at A. 
But in fuch a compound Tube as that made ufe of 
in our Experiment, Fig- ^th A CB, the cafe is very dif- 
ferent, and it does not eafily appear, why in a Facu- 
urn any part of the Water in the wider part of the Tube, 
as for Example at C, ffiould not leave that which is 
above it, and defcend ; fince the Annulus at C is by 
much too wide to fuflain a Column of Water of fo great 
a length as C B, 
The bed anfwer I can give to this difficulty is, that 
the Cohefion between the Water contain’d in the Ca- 
pillary and that below it, is fufficient to balance the 
weight of the Column fufpended. But how far this 
Cohefion may depend upon the PrefTure of a Medium 
fubtile enough to penetrate the Receiver, is worthy of 
Confideracion. For though fuch a Medium will pervade 
the Pores of the Water, as well as thofe of the GlaCs, 
yet it will adt with its intire PreiTure upon all the (olid 
Particles, if I may fo call chemi, of the futface of the 
Water in the Ciflern ; whereas fb many of the folid Par- 
ticles of the Water in the Tube, which happen to lie 
diredlly under rhe folid Particlesof the Water above them, 
will thereby be fecur’cl from this Preflure; and confe- 
quently there will be a lefs PrefTure of this Medium up- 
on any furface of the Water in the Tube below dig 
Capillary, than upon an equal furface of the Water in 
the Ciflern. So that the Column cf Water uvfpended 
