l8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 92 



Studying these tables, it should be borne in mind that the fourth ex- 

 periment was in progress for gi period of 45 days, whereas the fifth 

 experiment was of 16 days' duration. The difference in time probably 

 accounts for the greater number of colorless cells and cells in which 

 carotin had begun to appear in the fourth experiment. It should be 

 noted that cultures growing in light where the wave lengths were 

 cut off at 3700, 4000, 4500, 4600, 4800, and 5000 A were in especially 

 good condition. 



The cultures that showed the most disintegration of the chloroplasts 

 were those where the light was cut off at 5200 A in the fourth experi- 

 ment, 5600, 5800, and 5900 A, with the exception of one culture 

 in the fifth experiment, and 6000 A, 



GROWTH AS INDICATED BY MULTIPLICATION OF CELLS 



The results show that cell multiplication ranging from twofold to 

 fourfold occurred in all the complex beams of radiation. 



The third intensity experiment indicated that within the limits of 

 intensity here employed multiplication is proportional to the intensity 

 of illumination. 



If it be assumed that this law holds for each of the complex beams 

 employed, means are found for separating the propagating influences 

 of different wave lengths. For if in the energy curves of two com- 

 plexes whose included areas are equal a part P is common, then if 

 Q be the total area of either curve and M the growth ratio due to the 



O - P 

 complex of longer wave lengths, '^ M would be the growth 



ratio due to the part of the long-wave complex remaining in the 



shorter-wave complex. If A'' be the observed growth ratio of the 



- P . 



shorter-wave complex, N — ^^^—^ M will be the growth ratio due 



to the shorter wave lengths not found in the longer-wave complex. 



Working on this plan, growth ratios have been computed for many 

 narrow ranges of wave lengths, and by inspection of the overlapped 

 energy curves, approximate values of their effective wave lengths have 

 been estimated. (See tables 4-9.) 



In this way it is found that a wide red and hifrared complex of 

 wave lengths from 0.6 to 1.4 microns is moderately effective in pro- 

 moting multiplication of algae. It is impossible to know from these 

 experiments which of its wave lengths are the most eft'ective. The 



