Artificial Lighting by Fluorescent Lamps Part I. Fundamental Study On the quantitative comparison of fluorescent lamps with incandescent lamps

The fluorescent lamp, a new light-source which is called the "ideal light", was experimentally compared with the incandescent lamp on the quantitative basis, and the undermentioned conclusions were obtained. (1) In Lange's Standard-Beleuchtungsmesser, the photoelectric current is accurately proportional to the intensity of illumination, and the error is within a range of about 5% as checked with the Hefner's lamp, proving that it is an excellent photo-cell illuminometer. (2) In photometry, if there is a space more than double the length of mojor axis, the error is slight even though the inverse square law is applied. (3) The green fluorescent lamp is the highest in luminous intensity and lamp efficiency, which values lessen in order of the white fluorescent lamp, daylight fluorescent lamp, blue-white fluorescent lamp, pink fluorescent lamp and blue fluorescent lamp. (4) The comparison between the value of luminous intensity as measured by Bechstein's Flimmerphotometer and the values obtained by photometers of other types proves the following: (a) The error is generally small in Weber's photometer, but in case of green, blue-white or blue light, which differs in color of light, the measurement of R-G is impracticable. (b) The error in Bechstein's Luxmeter is less than 5% in any colors of light except green. However, in measurement of different colors other than the so-called "white", there is a trend that the judgement of equilibrium is apt to accompany difficulties. (c) Errors in Tavolux are rather great, but this is considered to be due to the variation caused by age. (d) In Lange's Standard-Beleuchtungsmesser, the error is small in case of white lights, but is fairly great in case of heterochromatic lights. Generally it is difficult to make the physical eye perfectly in accord with the international standard visibility curve, hence there is the need to obtain the light-meter correction factors with respective photometers. (5) Brightness is extremely small in fluorescent lamps, being 0.46c/c㎡ for the white fluorescent lamp, and 0.37c/c㎡ for the daylight fluorescent lamp, or so small as 1/70 of the 100W frosted incandescent lamp, whereas the lamp efficiency is so high as 5～7 times the incandescent lamp. (6) The bulb-wall temperature of a fluorescent lamp is very low, while its heat radiation is small. It is a so-called cool light. (7) It is desirable to light a fluorescent lamp at an ambient temperature within the range of 10°～30°C. The maximum luminous intensity is attained at the temperature of 20°～21°C. (8) In photometric measurement of a fluorescent lamp, it must be carried out after the lapse of more than 10 minutes from the time when the lamp is lit first. (9) One-fifths of both end-parts of a fluorescent lamp are darker by about 30% as compared with other parts of the lamp. (10) In the horizontal illumination curve of fluorescent lamp, there is witnessed a peculiar refraction at the horizontal distance of 50cm rigth beneath the light-source. (11) When measuring the luminous intensity and illumination, the location and direction of light-source must always be kept constant. (12) The variation in luminous intensity for every 1% of the change in the voltage of light-source is 1.8% with the fluorescent lamp, and such variations are lineal as shown in graph. In the incandescent lamp, the variation in luminous intensity grows large in proportion to the rise in voltage, and keep the relations between each other which are represented by a curve. The variations in intensity of illumination within any given time-unit are far less in frequency and degree compared with the incandescent lamp, and are less affected by the fluctuations in the voltage of the source of electric supply