Having made the Torricellian vacuum about five feet long, after the
manner described in the _Philosophical Transactions_, vol. xlvii. p.
370, if the excited tube be brought within a small distance of it,
a light will be seen through more than half its length; which soon
vanishes, if the tube be not brought nearer; but will appear again,
as that is moved farther off. This may be repeated several times,
without exciting the tube afresh.
This experiment may be considered as a kind of ocular demonstration
of the truth of Mr. Franklin's hypothesis; that when the electrical
fluid is condensed on one side of thin glass, it will be repelled
from the other, if it meets with no resistance. According to which,
at the approach of the excited tube, the fire is supposed to be
repelled from the inside of the glass surrounding the vacuum, and to
be carried off through the columns of mercury; but, as the tube is
withdrawn, the fire is supposed to return.