[Illustration:  Fig. 150.—­Smoke-making apparatus.]

A ring of smoke is projected from the hole in the card if the rubber diaphragm is pushed inwards.  A slow, steady push makes a fat, lazy ring come out; a smart tap a thinner one, moving much faster.  Absolutely still air is needed for the best effects, as draughts make the rings lose shape very quickly and move erratically.  Given good conditions, a lot of fun can be got out of the rings by shooting one through another which has expanded somewhat, or by destroying one by striking it with another, or by extinguishing a candle set up at a distance, and so on.  The experimenter should notice how a vortex ring rotates in itself while moving forward, like a rubber ring being rolled along a stick.

A continuous supply of smoke can be provided by the apparatus shown in Fig. 150.  The bulb of a scent spray is needed to force ammonia gas through a box, made air-tight by a rubber band round the lid, in which is a pad soaked with hydrochloric acid.  The smoke formed in this box is expelled through a pipe into the ring-making box.

Caution.—­When dealing with hydrochloric acid, take great care not to get it on your skin or clothes, as it is a very strong corrosive.

XXVII.  A RAIN-GAUGE.

The systematic measurement of rainfall is one of those pursuits which prove more interesting in the doing than in the prospect.  It enables us to compare one season or one year with another; tells us what the weather has been while we slept; affords a little mild excitement when thunderstorms are about; and compensates to a limited extent for the disadvantages of a wet day.

The general practice is to examine the gauge daily (say at 10 a.m.); to measure the water, if any, collected during the previous twenty-four hours; and to enter the record at once.  Gauges are made which record automatically the rainfall on a chart or dial, but these are necessarily much more expensive than those which merely catch the water for measurement.

This last class, to which our attention will be confined chiefly, all include two principal parts—­a metal receiver and a graduated glass measure, of much smaller diameter than the receiver, so that the divisions representing hundredths of an inch may be far enough apart to be distinguishable.  It is evident that the smaller the area of the measure is, relatively to that of the receiver, the more widely spaced will the graduation marks of the measure be, and the more exact the readings obtained.

[Illustration:  Fig. 151.—­Standard rain-gauge.]

The gauge most commonly used is that shown in Fig. 151.  It consists of an upper cylindrical part, usually 5 or 8 inches in diameter, at the inside of the rim, with its bottom closed by a funnel.  The lower cylindrical part holds a glass catcher into which the funnel delivers the water for storage until the time when it will be measured in a graduated glass.  The upper part makes a good fit with the lower, in order to reduce evaporation to a minimum.