36 METABOLISM 



at the ways in which the existence of this phenomenon may be proved, and 

 the more exact methods employed to estimate minute losses of water, not 

 merely the grosser evidence presented by the process of withering. 



Thanks to the large number of experimental researches that have been 

 carried out on the subject from the days of Hales (1748) up to the most recent 

 times (e. g. Burgenstein, 1887-1901), we have become acquainted with so 

 many methods that we must limit ourselves in their enumeration. Evapora- 

 tion from a plant may be demonstrated in the clearest and simplest way 

 by observing the dampness deposited on a bell jar placed over it and kept at 

 a low temperature. The reason for this deposition of moisture is the same as 

 that for the dimming of a window pane when one breathes on it ; it is nothing 

 more or less than the deposition of dew on a cold surface. The most exact, 

 uniform, and quantitative method of proving the existence of transpiration 

 is to employ a balance. If proper precautions be taken to permit of water 

 being given off from the plant only, and not at the same time from the 

 earth in which it is rooted, it may be shown that the decrease in weight 

 from hour to hour is due to loss of water. It is true that there are other pro- 

 cesses taking place in the plant which lead to change in weight, still, quantita- 

 tively speaking, they are insignificant when compared with the change in weight 

 due to movements of water. The data given on p. 25 as to the amount of tran- 

 spiration have been obtained by weighing. A third method, extremely con- 

 venient and useful for demonstration, consists in making use of the alteration in 

 colour which many substances undergo when they absorb water. Stahl (1894), 

 to whom we owe the application of this excellent method of investigation, used 

 strips of filter paper soaked in cobalt-chloride. ' Cobalt paper ' is deep blue 

 when dry, but becomes red when wet. The method of use is to place a small 

 piece of the blue paper on the subject of investigation — say a leaf — covering it 

 up with a glass plate so as to eliminate the influence of atmospheric moisture. 

 According as the leaf gives off much or little water the paper changes in colour, 

 after a few seconds, hours, or days. In place of change of colour we may employ 

 bodies which exhibit hygroscopic movements, such as gelatine (Benecke, 1899) or 

 awns of frorftMW (Darwin, 1898), for the demonstration of the same phenomenon. 



Many authors, e. g. Vesque (1877), Moll (1884), Bonnier and Mangin 

 (1884), and Kohl (1886), have employed the potometer figured on p. 30 for 

 demonstrating transpiration. It will be remembered that with this apparatus 

 it was the amount of water absorbed — not the amount given off — that was 

 measured. If transpiration be kept within moderate bounds, however, one 

 can make out that the two amounts are equal — that the loss due to evaporation 

 is covered by the amount absorbed. The potometer method has many ad- 

 vantages ; it is very easily demonstrable, more so if coloured water be employed 

 in the capillary tube ; it is very convenient, especially if the influence of external 

 factors on transpiration have to be studied ; it does not necessitate the presence 

 of roots on the plant — isolated branches are quite sufficient for the purpose. 



With the aid of one of these methods let us, first of all, study the effect which 

 the plant's structure has on transpiration. Observation alone teaches us that the 

 external walls of the epidermal cells are the parts of the plant first concerned 

 in the giving off of water vapour. Like all other cell-walls these contain water 

 of imbibition, and this water is retained with a certain amount of force. Every 

 particle of water lost by evaporation is replaced by another which the wall 

 attracts from the protoplasm. The protoplasm in turn abstracts water from 

 the cell-sap. But the cell-sap also holds the water firmly, and, in consequence 

 of its osmotic properties, together with the imbibitive energy of the cell-wall 

 and protoplasm, the outer surface of the plant gives off less water vapour than 

 an aqueous surface of equal extent under the same conditions. Aubert 

 (1892) found that, taking the evaporation from a water surface, per unit of 



