lve no more then what they are already acting upon, but
they carry up with them abundance of unctuous and sooty particles, which
meeting with that rag of the Week, that is plentifully fill'd with Oyl, and
onely spends it as fast as it evaporates, and not at all by dissolution or
burning, by means of these steamy parts of the filterated Oyl issuing out
at the sides of this ragg, and being inclos'd with an air that is already
satiated and cannot prey upon them nor burn them, the ascending sooty
particles are stay'd about it and fix'd, so as that about the end of that
ragg or filament of the snuff, whence the greatest part of the steams
issue, there is conglobated or fix'd a round and pretty uniform cap, much
resembling the head of a Mushrom, which, if it be of any great bigness, you
may observe that its underside will be bigger then that which is above the
ragg or stem of it; for the Oyl that is brought into it by filtration,
being by the bulk of the cap a little shelter'd from the heat of the flame,
does by that means issue as much out beneath from the stalk or downwards,
as it does upwards, and by reason of the great access of the adventitious
smoak from beneath, it increases most that way. That this may be the true
reason of this _Phaenomenon_, I could produce many Arguments and
Experiments to make it probable: As,

First, that the _Filtration_ carries the Oyl to the top of the Week, at
least as high as these raggs, is visible to one that will observe the snuff
of a burning Candle with a _Microscope_, where he may see an Ebullition or
bubbling of the Oyl, as high as the snuff looks black.

Next, that it does steam away more then burn; I could tell you of the dim
burning of a Candle, the longer the snuff be which arises from the
abundance of vapours out of the higher parts of it.

And, thirdly, that in the middle of the flame of the Candle, neer the top
of the snuff, the fire or dissolving principle is nothing neer so strong,
as neer the bottom and out edges of the flame, which may be observ'd by the
burning asunder of a thread, that will first break in those parts that the
edges of the flame touch, and not in the middle.

And I could add several Observables that I have taken notice of in the
flame of a Lamp actuated with Bellows, and very many others that confirm me
in my opinion, but that it is not so much to my present purpose, which is
onely to consider this concreet in the snuff of a Candle, so farr as it has
any