LIMITING FACTORS TO GROWTH MEASUREMENTS. 341 



lower throughout. The curve for Stifftia chrysantha, July 

 6th to 7th, was generally similar. 



The Light Factor. 



The cases of these shrubby plants are important as bringing 

 in to the question a factor which in considering the previous 

 cases of growth we have been able largely to ignore, namely, 

 light. In the case of Agave no retarding influence of light 

 could be at any time detected, for it was just when the shoot 

 was exposed to the hottest sun that the most rapid growth 

 took place. In Dendrocalamus, though it cannot be deci- 

 sively proved that the direct effect of light on the young shoots 

 is not in the direction of retarding growth, since the factors 

 of low humidity and intense light always occur together and 

 their effects cannot easily be separated, yet what evidence 

 there is on the question seems to show, as Lock (12) has pointed 

 out, that such an effect is not probable, and in any case must 

 be very small. The indirect effect of light acting through the 

 transpiration of the adult culms is, as we have seen, consider- 

 able, but of course in a quite different category from the direct 

 effect of which we are now speaking. It may be that the 

 absence of response to light in Agave, Furcraea, and Dendro- 

 calamus is due to the thick armour of bracts or leaf-sheaths 

 in which the young shoot is enveloped, though in this respect, 

 the flowering shoot of Agave, in which the absence of light 

 effect is most certain, is not so well off as Dendrocalamus. The 

 bracts which surround the growing inflorescence axis of Agave 

 do not form such a close and impenetrable armour as do the 

 leaf sheaths of Dendrocalamus. The cases of Stifftia, Capparis , 

 and other similar shrubs are in this respect quite different. 

 In these the growing region is exposed to the light with very 

 little protection. The extreme tip of the bud is indeed sur- 

 rounded by a number of developing leaves, but the growing 

 region extends considerably below this and is exposed for the 

 greater portion of its length with very little protection. It 

 might be expected therefore that here the effect of light would 



