the same results.

The manufacture of the apparatus is very simple. The tube is closed above
or below, according to the system one desires to adopt, by means of a
perforated cork. The valve is made of a piece of kid skin, which is fixed
by means of a bent pin and a brass wire (Fig. 2). It is necessary to wet
the skin in order that it may work properly and form a hermetic valve. The
arrangement of the lower valve necessitates the use of a tube of
considerable diameter (Fig. 1). We would advise the adoption of the
arrangement shown in Fig. 2. Under such circumstances a tube half an inch
in diameter and about 3 feet in length will answer very well.

It is better yet to simply use one's forefinger. The tube is taken in the
right hand, as shown in Fig. 3, and the forefinger placed over the
aperture. The finger should be wetted in order to perfect its adherence,
and should not be pressed too hard against the mouth of the tube. It is
only necessary to plunge the apparatus a few inches into the liquid and
work it rapidly up and down, when the water will rise therein at every
motion and spurt out of the top.

This is an easy way of constructing the _Chinese Pump_, which is found
described in treatises upon hydraulics. Such a pump could not, of course,
be economically used in practice on account of the friction of the column
of water against a wide surface in the interior of the tube. It is
necessary to consider the pistonless pump for what it is worth--an
interesting experimental apparatus that any one can make for himself.--_La
Nature_.

* * * * *

THE WATER CLOCK.

_To the Editor of the Scientific American_:

Referring to the clepsydra, or water clock, described and illustrated in
the SCIENTIFIC AMERICAN SUPPLEMENT of December 20, 1884, it strikes me
that the ingenious principle embodied in that interesting device could be
put into a shape more modern and practical, doing away with some of its
defects and insuring a greater degree of accuracy.

[Illustration: Fig 1.]

I would propose the construction given in the subjoined sketch, viz.: The
drum, A (Figs. 1 and 3), is mounted in a yoke suspended in such a manner
as to bring no unnecessary, but still sufficient, pressure on the friction
roller, B, to cause it to revolve the friction cone, C (both cone and
roller being of wood and, say, well rubbed with resin so as to increase
adhesion).

[Illustration: Fig 2.]

The friction