Methods are described for measuring the light emitted by an emulsion of luminous bacteria of given thickness, and calculating the light emitted by a single bacterium, measuring 1.1 x 2.2 micra, provided there is no absorption of light in the emulsion.

At the same time, the oxygen consumed by a single bacterium was measured by recording the time for the bacteria to use up .9 of the oxygen dissolved in sea water from air (20 per cent oxygen). The luminescence intensity does not diminish until the oxygen concentration falls below 2 per cent, when the luminescence diminishes rapidly. Above 2 per cent oxygen (when the oxygen dissolving in sea water from pure oxygen at 760 mm. Hg pressure = 100 per cent) the bacteria use equal amounts of oxygen in equal times, while below 2 per cent oxygen it seems very likely that rate of oxygen absorption is proportional to oxygen concentration. By measuring the time for a tube of luminous bacteria of known concentration saturated with air (20 per cent oxygen) to begin to darken (2 per cent oxygen) we can calculate the oxygen absorbed by one bacterium per second.

The bacteria per cc. are counted on a blood counting slide or by a centrifugal method, after measuring the volume of a single bacterium (1.695 x 10–12 cc.). Both methods gave results in good agreement with each other.

The maximum value for the light from a single bacterium was 24 x 10–14 lumens or 1.9 x 10–14 candles. The maximum value for lumen-seconds per mg. of oxygen absorbed was 14. The average value for lumen-seconds per mg. O2 was 9.25. The maximum values were selected in calculating the efficiency of light production, since some of the bacteria counted may not be producing light, although they may still be using oxygen.

The "diet" of the bacteria was 60 per cent glycerol and 40 per cent peptone. To oxidize this mixture each mg. of oxygen would yield 3.38 gm. calories or 14.1 watts per second. 1 lumen per watt is therefore produced by a normal bacterium which emits 14 lumen-seconds per mg. O2 absorbed. Since the maximum lumens per watt are 640, representing 100 per cent efficiency, the total luminous efficiency if .00156.

As some of the oxygen is used in respiratory oxidation which may have nothing to do with luminescence, the luminescence efficiency must be higher than 1 lumen per watt. Experiments with KCN show that this substance may reduce the oxygen consumption to 1/20 of its former value while reducing the luminescence intensity only ¼. A partial separation of respiratory from luminescence oxidations is therefore effected by KCN, and our efficiency becomes 5 lumens per watt, or .0078. This is an over-all efficiency, based on the energy value of the "fuel" of the bacteria, regarded as a power plant for producing light. It compares very favorably with the 1.6 lumens per watt of a tungsten vacuum lamp or the 3.9 lumens per watt of a tungsten nitrogen lamp, if we correct the usual values for these illuminants, based on watts at the lamp terminals, for a 20 per cent efficiency of the power plant converting the energy of coal fuel into electric current.

The specific luminous emission of the bacteria is 3.14 x 10–6 lumens per cm2.

One bacterium absorbs 215,000 molecules of oxygen per second and emits 1,280 quanta of light at λmax = 510µµ. If we suppose that a molecule of oxygen uniting with luminous material gives rise to the emission of 1 quantum of light energy, only 1/168 of the oxygen absorbed is used in luminescence. On this basis the efficiency becomes 168 lumens per watt or 26.2 per cent.