2 SMITHSONIAN MISCELLAXEOUS COLLECTIONS VOL. 87 



The Division of Radiation and Organisms of the Smithsonian In- 

 stitution has devoted consideral)le time and effort during its first three 

 years to the de\'elopment of equipment and methods for the compara- 

 tive investigation of the effects of hght upon plant growth and be- 

 havior. Since to a considerable degree the general plan of investigation 

 and the apparatus employed will be common to a series of experiments 

 it seems desirable to describe the developments in some detail. Such 

 is the purpose of this discussion. 



SOURCES OF RADIATION 



The type of set-up to be developed depends necessarily in a large 

 measure upon the characteristics of the sources of radiation which are 

 available. Practically one has three choices : first, the sun ; second, 

 standardized incandescent filament lamps ; third, electrical discharges 

 through gases. The chief objection to the first, namely, its unavoidable 

 fluctuation in intensity and distrilnition, has been sufficient to rule it 

 out in the present undertaking. In designing the apparatus provision 

 has been made to utilize both of the latter types. Incandescent lamps 

 have the definite advantage of furnishing radiation distributed continu- 

 ously over a large range of wave lengths. They can be obtained with 

 a minimum of expense and, with suitable control, made to yield reason- 

 ably constant conditions. If it is desired, however, to restrict the light 

 to a very narrow wave-length range, that is, pure color, one finds rather 

 rigidly-defined limitations. A monochromator type of illumination 

 must be ruled out in the case of most investigations of higher plants, 

 not only because of expense but also because it fails to furnish a 

 sufificient quantity of available radiation. The use of light filters, 

 however, enables one to modify and limit the distribution to a con- 

 siderable extent. 



An interesting set of filters, developed by the Corning Glass Works, 

 transmit continuously the radiation of long wave lengths from about 

 2.5/A in the infra-red to fairly well-defined limits in the visible and 

 ultra-violet. This short wave-length limit differs for the different 

 filters. A group of lo filters may be chosen, for which the short wave- 

 length limit varies at convenient intervals from the deep red to the 

 ultra-violet beyond 2,900 A. While the short wave-length limit is 

 not ideal, the change of transmission from 80 per cent or more to less 

 than 2 or 3 per cent takes place within a few hundred Angstroms. 

 The transmission characteristics of these filters are shown in Figure i, 

 curves i to 10. Percentage transmission is plotted against wave length 

 in microns. The infra-red transmission values have been determined 



