adhering to the surface of the drop. By regulating the distance from
the plate the tar is slowly dried out and the button becomes solid. I
then once more dip the button in tar and hold it again over a plate
until the tar is evaporated, leaving only a hard mass which firmly
binds the crystals. When a larger button is required I repeat the
process several times, and I generally also cover the filament a
certain distance below the button with crystals. The button being
mounted in a bulb, when a good vacuum has been reached, first a weak
and then a strong discharge is passed through the bulb to carbonize
the tar and expel all gases, and later it is brought to a very intense
incandescence.

When the powder is used I have found it best to proceed as follows: I
make a thick paint of carborundum and tar, and pass a lamp filament
through the paint. Taking then most of the paint off by rubbing the
filament against a piece of chamois leather, I hold it over a hot
plate until the tar evaporates and the coating becomes firm. I repeat
this process as many times as it is necessary to obtain a certain
thickness of coating. On the point of the coated filament I form a
button in the same manner.

There is no doubt that such a button--properly prepared under great
pressure--of carborundum, especially of powder of the best quality,
will withstand the effect of the bombardment fully as well as anything
we know. The difficulty is that the binding material gives way, and
the carborundum is slowly thrown off after some time. As it does not
seem to blacken the globe in the least, it might be found useful for
coating the filaments of ordinary incandescent lamps, and I think that
it is even possible to produce thin threads or sticks of carborundum
which will replace the ordinary filaments in an incandescent lamp. A