RAPIDITF OF TRANSPIRATION CURRENTS. 



235 



a distance of 50 to 80 or 100, occasionally even of 200 centimetres, is frequently 

 traversed within an hour by a particle of water in the wood. It is obvious 

 that when the transpiration at the leaves is feeble, and the quantity of water 

 consumed small, the rapidity of the flow in the w^ood also diminishes; and that 

 with the complete stand-still of transpiration, the movement in the wood likewise 

 ceases, or at any rate sinks to a minimum in the event of a small supply of water 

 being necessary for the purposes of growth of parts situated higher. Formerly it was 

 sought in various ways to obtain an idea of the velocity of the ascending current of 

 water, but all the different methods employed for this purpose have been shown to be 

 faulty in the highest degree. I have shown that researches with cut-off branches are 

 for our purpose to be rejected once for all; and that likewise the experiments 

 made for centuries, in which 

 coloured solutions were allowed 

 to ascend in the wood, must ne- 

 cessarily give velocities which are 

 too small. This is easily seen on 

 reflecting that the colouring of the 

 wood cell-walls simply consists in 

 their seizing upon and holding fast 

 the colouring matter, while the 

 water of solution thus freed from 

 the colouring matter speeds on 

 in advance. This may be very 

 clearly illustrated by means of the 

 apparatus here represented. A 

 strip of filter paper hangs with its 

 lower end in a solution of colour- 

 ing matter {a), e.g. of aniline violet 

 or indigo. Even a few minutes 

 after the immersion a disassocia- 

 tion of the fluid is noticed : the 

 upper limit of the coloured part 

 {(/) only ascends slowly in the pk; j^^, 



paper strip, while the medium of 



solution {Ö c), freed from the colouring matter, ascends much more rapidly. The 

 like must also happen when a cut-off branch is placed with the cut surface in a solu- 

 tion of colouring matter; only in this case it is not observed how quickly the water 

 speeds on in advance of the colouring matter. Moreover in this mode of experimenting 

 still further errors are introduced through the rarefaction of the air in the wood, 

 to which I shall refer later. Really useful observations on the velocity of the ascend- 

 ing water current are obtained, however, when weak solutions of lithium nitrate are 

 allowed to be absorbed by the uninjured roots of a transpiring plant. For this purpose 

 the plants must have been previously grown for some time in earth in a flower-pot, or 

 have developed all their roots in a nutritive solution, according to the method to be 

 described later ; since it is impossible to obtain uninjured roots by digging them up 

 and washing them. The lithium solution possesses, as I convinced myself with the 



