The earth has different climates, and these have different temperatures. As we approach the poles the variation of seasons is more perceptible,--all bodies contract with cold and expand with heat. This effect is more readily measured in liquids, and is particularly noticeable in spirituous liquors. This fact suggested the idea of the thermometer. The wind strikes our faces; air is therefore a body, a fluid; we feel it though we cannot see it. Turn a glass vessel upside down in water, and the water will not fill it unless you leave a vent for the air; therefore air is capable of resistance. Sink the glass lower, and the water rises in the air-filled region of the glass, although it does not entirely fill that space. Air is therefore to some extent compressible. A ball filled with compressed air bounds much better than when filled with anything else: air is therefore elastic. When lying at full length in the bath, raise the arm horizontally out of the water, and you feel it burdened by a great weight; air is therefore heavy. Put air in equilibrium with other bodies, and you can measure its weight. From these observations were constructed the barometer, the siphon, the air-gun, and the air-pump. All the laws of statics and hydrostatics were discovered by experiments as simple as these. I would not have my pupil study them in a laboratory of experimental physics. I dislike all that array of machines and instruments. The parade of science is fatal to science itself. All those machines frighten the child; or else their singular forms divide and distract the attention he ought to give to their effects. I would make all our own machines, and not begin by making the instrument before the experiment has been tried. But after apparently lighting by chance on the experiment, I should by degrees invent instruments for verifying it. These instruments should not be so perfect and exact as our ideas of what they should be and of the operations resulting from them. For the first lesson in statics, instead of using balances, I put a stick across the back of a chair, and when evenly balanced, measure its two portions. I add weights to each part, sometimes equal, sometimes unequal. Pushing it to or fro as may be necessary, I finally discover that equilibrium results from a reciprocal proportion between the amount of weight and the length of the levers. Thus my little student of physics can rectify balances without having ever seen them.