REPORT OF THE SECRETARY 119 



influence would be carried around so that it was felt equally from 

 all sides. Ultiniatel}', this only proved completely effective when 

 the space between the double walls of glass was filled with water. 



In order to determine the wave-lengtli range transmitted by the 

 filtei-s used in the phototropic experiment, spectrograms, secured 

 through the cooperation of the Fixed Nitrogen Laboratory, were 

 taken. These showed the varying amounts of light transmitted by 

 the filter from an incandescent bulb of the type used in the experi- 

 ment. In order to interpret the spectrograms, a self -registering 

 microphotometer, previously secured by the Astrophysical Observa- 

 tory, was set up in one of the small constant-temperature rooms. 

 Curves were thus obtained, representing the relative transmission of 

 these fdters for the various wave lengths, and so determining defi- 

 nitely the region of the spectrum, or colors, used. 



Completion of this preliminary experiment in September of 1930, 

 yielded the following facts: First, that red, or infra-red light, pro- 

 duced no measurable effect; second, that yellow light of the type 

 used, produced a small, but measurable bending ; third, that the green 

 light was one thousand times (+ or —2 per cent) more effective than 

 yellow; and finally, that blue light of the range used was thirty 

 times more effective than the green, or thirty thousand times more 

 effective than the yellow. 



The outstanding points of interest resulting from the preliminary 

 experiments are two: First, the quantitative reproducibility of the 

 ratios obtained, that is, within 5 per cent of the magnitudes quoted; 

 second, the tremendous ratio observed between yellow and green, 

 together with rapid increase of the effect as one proceeds to shorter 

 wave lengths, tends to indicate a definite threshold wave length at 

 which the phototropic influence shows itself. This is typical of 

 photochemical reactions involving an electronic change of energy. 

 Tiie conclusions drawn from this observation are of far-reaching 

 importance, as this constitutes crucial evidence against any tiieory 

 which seeks to explain phototropism as merely a thermal effect due 

 to unequal absorption of radiant energy. 



Tiie preliminary results of this experiment proved so interesting, 

 even at the early stages, that steps were immediately taken to prepare 

 for a more elaborate experiment. Mr. Hoover, returning from 

 extensive field experience with the Astrophysical Observatory, was 

 assigned to the work of the division in December, 1929. An ex- 

 perienced observer himself, naich of the successful carrying out of 

 this phototropic experiment is due to his efforts. 



The things which immediately suggest them.selves as desirable 

 in a new experiment are, flrst: To use a narrower wave-length range, 

 or a purer color, in order to determine what wave lengths have a 

 particular effect, and second; to use a larger number of different 



