For the same reason it will be assumed that electric glow-lamps of low intensity (nominally of 8 candle-power or less), aggregating 70-80 candle-power, will practically serve, if suitably distributed, equally as well as 100 candle-power obtained from more powerful sources of light. Electric glow-lamps of a nominal intensity of 16 candles or thereabouts, and good flat-flame gas-burners, aggregating 90-95 candle-power, will similarly be taken as equivalent, if suitably distributed, to 100 candle- power from more powerful sources of light. Of the latter it will be assumed that each source has an intensity between 20 and 30 candle-power, such as is afforded by a large oil-lamp, a No. 1 Welsbach-Kern upturned, or a "Bijou" inverted incandescent gas-burner, or a 0.70-cubic-foot-per- hour acetylene burner. Either of these sources of light, when used in sufficient numbers, so that with proper distribution they light a room adequately, will be taken in the tabular statement which follows as affording, per candle-power evolved, the standard illuminating effect required in that room. The same illuminating effect will be regarded as attainable by means of candles aggregating only 35 per cent., or small electric glow-lamps aggregating 77 per cent., or large electric glow- lamps and flat-flame gas-burners aggregating 90 to 95 per cent. of this candle-power; while if sources of light of higher intensity are used, such as Osram or Tantalum electric lamps, or the larger incandescent gas- burners (the Welsbach "C" or "York," or the Nos. 3 or 4 Welsbach-Kern upturned, or the No. 1 or larger size inverted burners) or incandescent acetylene burners, it will be assumed that their aggregate candle-power must be in excess by about 15 per cent., in order to compensate for the impossibility of obtaining equally well distributed illumination. These assumptions are based on general considerations and data as to the effect of sources of light of different intensities in giving practically the same degree of illumination in a room; it would occupy too much space here to discuss more fully the grounds on which they have been made. It must suffice to say that they have been adopted with the object of being perfectly fair to each means of illumination. COST PER HOUR AND HYGIENIC EFFECT OF LIGHTING BY DIFFERENT MEANS The data (except in the column headed "cost per 100 candle-hours") refer to the illumination afforded by medium-sized (0.5 to 0.7 cubic foot per hour) acetylene burners yielding together a light of about 100 candle- power, and to the approximately equivalent illumination as afforded by other means of illumination, when the lighting-units or sources of light are rationally distributed.