.'{tin SMTTB UTI.n \ri<>s OF Till-: THEORY OF 



moisture, The measurements of Shibata (20) on Phyllos- 

 tachySj of Lock (12) and myself on Dendrocalamus and my 

 ou ii observations on various other plants have not yet shown 

 ,i clear instance of unmistakable ligb.1 effect, while effects of 

 temperature ami moisture have in different cases been verj 

 marked. There lias in fact been nothing in our results which 

 needed the introduction of any other factor besides water 

 supprj and temperature. The case of Agave inflorescence 

 axis shows thai very rapid growth may take place even while 

 the growing organ is exposed to the full intensity of a tropical 

 mid-day sun. It may be of course that even here the growth 

 would be somewhal greater bul for the retarding influence of 

 lightj hut this, considering the rapidity of the growth, is im- 

 probable. It certainly seems thai compared with the other 

 environmental factors the effect of light is negligible in some 

 B and may he generally so. The mechanism of the 

 retarding lighl effect is of course entirely unknown to us. but 

 whatever it may he there seems to he some difficulty in re- 

 garding lighl as a "limiting factor" ingrowth. Suppose for 

 instance a growing organ was subjected to constantly increas- 

 ing water supply in a lighl of given intensity, the temperature 

 being throughout high enough for much greater growth than 

 is taking place. The growth being in this case limited by 

 water supply would increase with that factor. If the light LS 

 of the nature of a limiting factor the growth will increase with 

 the humidity lip to a certain point when this increase will no 



longer be observed. At this point the intensity of the lighl 



may l><- supposed to limit the more rapid growth while it did 



not affect the Blower growth. In other words there is not 



sumoienl • absence of lighl or ' darkness " for the rapid growth 

 while there had been sufficient darkness for the slowei 

 growth. Such a conception is highly artificial, and we must 

 suppose thai the retarding effect of light, however large 01 



mall it ma\ be acts equalll "ii both slowly and cpiicUv 



