314 Profs. Percy Frankland and Marshall Ward. 



rodlet, a comparatively easy task with the small drops used and with 

 this power, and noted the commonest average length which prevailed 

 among the little plantlets developed. The following summary will 

 give the best idea of the results. (Table, p. 315). 



The averages were got by taking the sum of all the lengths 

 measured, and dividing by the number of rods measured. This last 

 number is placed in brackets next each average. 



These results seemed to show very clearly that (1) the differences 

 in the rapidity of growth are very large, and (2) they are not to be 

 referred directly to the changes in temperature. 



They also show that the growth in total darkness under the given 

 conditions is much more rapid than the growth in any kind of light 

 experimented with, except the red. It is eventually more rapid than in 

 the red light, but not so at first, and I attributed the difference here 

 to the fact that so long as the light was acting the dark heat rays in 

 the red can make themselves felt to such an extent that the young 

 filaments grow more actively during this period; when the light 

 fails, however, between 6 and 7 P.M., the curve in the red falls the 

 effect of a falling thermometer, in fact. 



In the present case the blue light seemed to be distinctly a retard- 

 ing agent throughout, but it was not sufficiently intense to exercise 

 its full effect, and the screen must be regarded as having acted 

 more as a shade to the relatively low intensity of light employed 

 than as a transmitter of any large proportion of the presumably 

 injurious rays. 



With the ordinary glass and the pale olive (which really trans- 

 mits almost as much light) the retarding effects of the light seemed 

 very evident. I attributed the more effective retardation behind the 

 pale olive to the fact that it transmits about as strong a light as 

 the ordinary glass, but with relatively more blue in its composition. 

 The effects with green glass screens are often puzzling, as I have 

 found in other cases. On the whole it seemed to act like a shade to 

 the intensity of the light, but what does traverse has a lot of blue in 

 it. It lets the dark heat rays through, and consequently the fall 

 about 7 P.M. is like that of a fall of temperature. 



In both cases that of the red and that of the green it will be 

 noted that the fall of the curve would be aided by the fact that the 

 temperature was also falling slowly all the time, and this latter fact 

 also probably prevented the other curves (blue, ordinary, and pale 

 olive) from ascending more rapidly from 7 P.M. onwards. 



Without pressing this explanation too far, it seemed not unlikely 

 that it was approximately correct, though the matter is undoubtedly 

 a complex one. 



In any case the above results seemed to show very clearly that 

 whereas the mean growth of the filaments, in darkness, or behind 



