I have here a piece of No. 24 platinum wire which has about the same resistance as iron wire when cold, but you notice that I may use a very much shorter length than I did of the iron wire because it will endure a very much higher heat without melting. Reducing the length would reduce the resistance, but reducing the resistance would allow more current to pass. If more current should pass it would make the wire hotter, and raising the temperature would increase the resistance, which would cut down the current, etc. By sliding the clip _c_ (Fig. 92), along, I finally reach a point where conditions balance so that I get a very brilliant light, dangerously near the fusing point of the platinum which is three thousand degrees above the boiling point of water. In 1879 Mr. Thomas A. Edison literally searched the whole world for something better than platinum for the filament of an incandescent lamp. He finally decided upon charred threads of a bamboo which he found in Japan. No research was ever more timely than this. Whereas there was practically no electric lighting before 1880, soon after that there began a phenomenal demand for carbon filament lamps. In 1890, 800,000 of these lamps were manufactured in the United States. In 1900 the number had risen to 25,000,000. In 1909 central stations were supplying electric current to 41,807,944 incandescent electric lights. By far the greatest number are still made with carbon filaments. [Illustration: Fig. 92] [Illustration: Fig. 93] We examined an ordinary 110-volt 16-candle-power carbon filament lamp, (Fig. 93). As near as we could estimate, its filament measured about eight inches in length. We broke open the bulb of this lamp by laying it upon the table and tapping it with a board. The bulb broke with rather a loud noise and the brittle carbon filament broke into many pieces. We found one of these pieces and measured its diameter with a wire gauge, (Fig. 94). It was the same size as No. 33 wire, which we also found by the wire gauge was the size of No. 90 sewing cotton. The diameter of No. 33 wire was given upon the wire gauge as .007 inch. When lighted, the filament of this lamp had looked to be about the size of No. 18 wire, which has a diameter of .04. That is, the filament when lighted looked six times as thick as it really was. Those who use sewing cotton learn quickly to know the size of the thread by its number. So those who have much to do with wire easily learn the system of designating sizes by numbers. Here are some selected figures easy to remember. A trolley wire is about one third of an inch in diameter. It is designated as No. 0. Notice in the following table that as the numbers rise by six the diameters are divided by two. Notice also that as the diameters diminish by two the resistance increases by four. [Illustration: Fig. 94] TABLE OF RESISTANCE OF COPPER WIRES _Nos._ _Diameter_ _Resistance_