250 RADIATION BIOLOGY 



The data for the spectral energy distribution of the standard lamp 

 given in Table 3-19 were obtained by multiplying the emissivity data 

 (column 2) by the relative spectral intensity of a complete radiator at 

 the true temperature of the filament (column 3) and arbitrarily reducing 

 the values to 100 at 560 m/i. The relative spectral energy distribution 

 for a complete radiator at 2854°K is given in column 4 for comparison. 

 The spectral-emissivity data for tungsten are those of Forsythe and 

 Adams (1945); similar data have been published by Ornstein (1936), but 

 the two sets of values differ markedly in some regions (see Sect. 2 for 

 references) . It is the uncertainty as to the spectral emissivity of tungsten 

 which limits the accuracy with W'hich energy-distribution curves, obtained 

 from color-temperature data, may be extrapolated into the ultraviolet 

 and infrared. In general, however, the values given in Table 3-19 closely 

 approximate those given by Forsythe and Adams for the direct measure- 

 ment of a similar lamp of the same color temperature (2848°K on the 

 1931 temperature scale and 1854°K on the 1948 scale). 



The temperature is not uniform throughout all parts of the filament 

 of a lamp, especially where contact is made with the supports. The 

 color temperature specified is the average for the whole filament and 

 introduces a small error in the calculation of spectral energy distribution. 

 In order to obtain a more uniform filament temperature for the emitted 

 radiant energy and to be able to extrapolate spectral energy distribution 

 further into the ultraviolet, Stair and Smith (1943) designed a special 

 lamp in a quartz envelope, with a tungsten wire filament arranged in a 

 series of four hairpin turns. The flux radiated beyond the lamp is limited 

 to the straight portions of the filament by metal shields over the support 

 wires and end turns. The color temperature of the exposed portions of 

 the filament is very uniform and subject to more precise extrapolation 

 than is possible for the glass-projection-lamp standard. Coolidge (1944) 

 has studied the mercury arc as a standard of spectral energy distribu- 

 tion in the ultraviolet. Although this source is not so stable and repro- 

 ducible as the incandescent lamp, its ultraviolet radiation is much more 

 intense, and the fines can be isolated with filters. 



SPECTROSCOPIC INSTRUMENTS 



When comparing the performance of monochromators, it is often essen- 

 tial to have data on the spectral transmission of the complete instrument. 

 If the monochromator is used with a detector to measure the spectral 

 energy distribution of a source, the combination must be calibrated for 

 spectral responsivity and accuracy of wave-length indication. Spectro- 

 photometers must be occasionally calibrated for transmission and wave 

 length (Gibson, 1949; Gibson and Balcom, 1947; Mellon, 1950; Normand 

 and Kay, 1952). These operations require the use of standards of wave 

 length, spectral energy distribution, and transmission. 



