434 HAROLD F. BLUM 



number of lines in the ultraviolet below 334 m^u for most purposes, 

 but only scattered lines in the longer ultraviolet and visible. A mer- 

 cury arc operated at very low pressure may emit most of its energy 

 (about 85%) in the resonance line, 253.7 m/x; this is a useful source 

 of nearly monochromatic radiation for some purposes, but is obvi- 

 ously of little use by itself for determining action spectra. 



"Continuous" sources, of which the tungsten filament lamp is an 

 example, emit all wavelengths over a wide spectral range. To isolate 

 approximately monochromatic radiation from such a source requires 

 a very efficient monochromator. Monochromators of prism type 

 are the more commonly employed, but gratings that concentrate most 

 of the energy in a single order may be satisfactory. The latter have 

 the advantage that the dispersion is the same for all wavelengths. 

 Increased purity may be obtained by combining two prisms or two 

 gratings into a ''double" monochromator. Even with the largest 

 monochromators available the area that can be successfully illumi- 

 nated is limited to a relatively small size, inadequate for many biologi- 

 cal studies. 



Filters, on the other hand, permit much larger areas to be irra- 

 diated, but the older types of "monochromatic" filters (e.g., glass or 

 gelatin film) transmit wide spectral bands, and cannot be compared 

 to monochromators as regards purity. However, they may, if 

 properly chosen, serve well enough for isolation of monochromatic 

 radiation from line sources when the lines are not too close together. 

 During the past few years interference filters have been developed 

 that isolate narrower spectral bands than is possible with the older 

 types, and these will no doubt prove of considerable value for photo- 

 biological studies. As much as 45% of the incident radiation may be 

 transmitted, the width of the transmitted band being usually less 

 than 20 m/x, for half peak transmission. This makes them appro- 

 priate for use with line sources, although they may not provide suf- 

 ficient purity for successful use with continuous sources. At present 

 such filters are available only for wavelengths in the visible region, 

 but they may soon be produced for the near infrared and ultraviolet. 

 As is the case with older types, these filters may transmit elsewhere 

 than in the principal band of transmission, and this possibility must 

 be kept in mind. Further remarks about the use of filters will be 

 found at the end of Section C3. 



Generally, no matter what the object of the study, the principal 

 physical measurement that has to be made is the determination of the 



