This system involves the use of two gauges for every size. The man employed upon making one-inch bars has a plate with a hole in it 1-1/1000 inches in diameter and another hole 999/1000 of an inch in diameter. One of these is the "go in" gauge; the other is the "not go in." So that all he has to do, in order to be quite sure that his work is right, is to see that it can be poked through one of these holes, but not through the other. No trouble at all, it will be observed, adjusting fine measuring appliances, simply a plate with two holes in it, and the workman can be sure that he is turning out articles every one of which is practically correct, with no variation beyond a slight inequality too small to matter. And probably at some other part of the factory there is a man making articles each of which has a hole in it, into which this bar must fit. How does he manage? He is provided with a gauge somewhat the shape of a dumb-bell, one end of which is slightly larger than the other. One is the "go in" end, the other the "not go in" end. If the hole which he makes will permit the former to enter, but will refuse admittance to the latter, then he knows that that hole is sufficiently near its reputed size to answer its purpose. [Illustration: _By permission of The Mining Engineering Co., Sheffield_ A MINERS' RESCUE TEAM These men are equipped with breathing apparatus which enables them to pass safely through the deadly fumes after an explosion, to rescue their unfortunate comrades] In the instances mentioned, a thousandth of an inch either way has been mentioned as the limit of inaccuracy, or the "tolerance," as it is sometimes termed, but often the limits are much narrower than that. The gauges themselves are a case in point, for they must be true within, say, a ten-thousandth, or even less. And they too are checked by master gauges of a finer degree of accuracy still, being made by the most laborious methods, and checked over and over again, so as to reach the utmost limits in the way of correctness. So this methodical "scientific" system of "limit gauges" is based upon the principle of having one gauge limiting the error one way and another defining it in the other. Anything simpler or more effective it would be impossible to conceive. It is due very largely to this system that many manufactured articles are now so much cheaper than they used to be. For it enables each individual part to be made wholesale on a large scale, by machines specially adapted to the work, operated by men specially trained to work them, with the practical certainty that these parts when assembled together will fit each other. In conclusion, there is another very interesting instrument which was first made for a purely utilitarian use--namely, the investigation of the methods of making coloured glass--but which has since been applied to some interesting problems in pure science. It is called the "ultra-microscope."