nnot help looking upon them as
more satisfactory tests for the microscope than diatoms and other
things of the real shape of which we know nothing whatever.

Since figures as large as a million cannot be realized properly, it
may be worth while to give an illustration of what is meant by a fiber
one-millionth of an inch in diameter.

A piece of quartz an inch long and an inch in diameter would, if drawn
out to this degree of fineness, be sufficient to go all the way round
the world 658 times; or a grain of sand just visible--that is,
one-hundredth of an inch long and one hundredth of an inch in
diameter--would make one thousand miles of such thread. Further, the
pressure inside such a thread due to a surface tension equal to that
of water would be 60 atmospheres.

Going back to such threads as can be used in instruments, I have made
use of fibers one ten-thousandth of an inch in diameter, and in these
the torsion is 10,000 times less than that of spun glass.

As these fibers are made finer their strength increases in proportion
to their size, and surpasses that of ordinary bar steel, reaching, to
use the language of engineers, as high a figure as 80 tons to the
inch. Fibers of ordinary size have a strength of 50 tons to the inch.

While it is evident that these fibers give us the means of producing
an exceedingly small torsion, and one that is not affected by weather,
it is not yet evident that they may not show the same fatigue that
makes spun glass useless. I have, therefore, a duplicate apparatus
with a quartz fiber, and you will see that the spot of light comes
back to its true place on the screen after the mirror has been twisted
round twice.

I shall now for a moment draw your attention to that peculiar property
of melted quartz that makes threads such as I have been describing a
possibility. A liquid cylinder, as Plateau has so beautifully shown,
is an unstable form. It can no more exist than can a pencil stand on
its point. It immediately breaks up into a series of spheres. This is
well illustrated in that very ancient experiment of shooting threads
of resin electrically. When the resin is hot, the liquid cylinders,
which are projected in all directions, break up into spheres, as you
see now upon the screen. As the resin cools, they begin to develop
tails; and when it is cool enough, i.e., sufficiently viscous, the
tails thicken and the beads become less, and at last uniform threads
are the result. The se