ABSORPTION OF RAIN AND DEW BY GREEN PARTS OE PLANTS. 169 



and herbaceous stems were exceptionally flaccid at the beginning of the 

 period of darkness. ■ 



Now Duchartre always weighed his plants early in the morning after 

 this statical condition was fully attained ; so that it is not at all 

 surprising to find that he could not detect any increase of weight ; hence 

 his experiments seem to prove conclusively that at night dew is not 

 usually absorbed in any appreciable degree. 



Dew, however, does not disappear suddenly from leaves at sunrise ; 

 and it is only after sunlight and heat begin again to affect leaves that 

 dew is actually absorbed. Herein, however, is involved a practical 

 difficulty, for the balance will no longer help us. But I believe that as 

 soon as transpiration recommences, then any part that may be the first 

 to become dry will begin to transpire, and so cause an indraught of dew 

 in any neighbouring spot where it may have been retained ; so that there 

 will be an influx and efflux accompanied by the usual root-supply, which 

 probably furnishes the main source of water for transpiration. Hence it 

 will be seen that it is generally impossible to detect the absorption of dew 

 or rain by leaves with mathematical accuracy or to prove it to demonstra- 

 tion. On the other hand, the " prcof " that such is the case may be 

 arrived at indirectly by accumulating probabilities based upon observed 

 facts. Such is the methcd I have attempted by aid of the following 

 experiments. 



The conclusion I have arrived at is that, while there is no objection 

 that I know of which cannot be met, there are ample reasons for believing 

 that dew and rain are, when absorption has been deficient, absorbed and 

 utilised to supplement the nrrmal root-supply. 



2. Experiments illustrating the Power cf Absorption op 

 Water by the Epidermis of Herbaceous Internodes. 



A shoct of first year's growth of elm had three internodes wrapped 

 up in saturated blotting-paper on June 12. By the 15th the leaves were 

 flaccid and nearly faded ; but the terminal bud and a leaf adjacent to it 

 remained quite fresh. By the 20th all four leaves were almost dead, with 

 the exception of the bases of the blades. The terminal bud remained 

 perfectly fresh until the 29th, when the whole was dead. Duration, seven- 

 teen days. A similar specimen not moistened totally perished in two days. 



When herbaceous plants, especially those with tolerably large leaves, 

 as Borago officinalis, Rumex crispus, Sisymbrium Alliaria, Lychnis 

 dioica, &c, have only their internodes wrapped up in saturated blotting- 

 paper, the leaves generally soon wither and perish, but the internodes 

 remain green and fresh for long periods. 



A branch of borage, having two internodes, was wrapped up on J une 8. 

 On the next day the leaves were much faded, but the stem was quite 

 firm. On the 10th the upper part of the leaves was brown, brittle, and 

 dead, but no change had taken place in the stem. By the 12th the leaves 

 were entirely withered, excepting a small portion at their bases. On the 

 13th the leaves were quite dead. The internodes remained firm, green, 

 and fresh. They thus continued until July 10. They then decayed 

 slowly. Duration, five weeks and three days. 



