THE CONDUCTION OF WATER. I 53 



from below upwards is by no means always regular, and further that the pres- 

 sure variations in the individual manometers are often quite independent of 

 each other. This phenomenon is explained especially by the fact that bleeding- 

 pressure under consideration may exist not merely in the root, but also in a number 

 of other places, in the stem, in the leaves, in the axis of inflorescence, and so on. 

 Although the regions where bleeding occurs may be quite near to each other, it 

 by no means follows that they are in uninterrupted communication with each 

 other, and this condition is applicable to two nearly related regions in the 

 xylem of a tree. Why intercommunication in this latter case is not unrestricted 

 we shall discover when we consider the external conditions of bleeding. 



The first and most general condition of bleeding is the occurrence of living 

 cells in the neighbourhood of vessels. Death of the plant stops all further 

 bleeding, and certain stimuli, which diminish its vitality without actually pro- 

 ducing death, temporarily, at least, retard the phenomenon. For instance, 

 WIELER (1893) brought the bleeding at once to a standstill by stopping the 

 supply of oxygen, that is to say, by inhibiting respiration, and chloroform has 

 the same effect. We may deduce, therefore, from these facts that bleeding 

 is a vital phenomenon. 



A second important condition of bleeding is that the cells which exhibit 

 this phenomenon must be abundantly supplied with water, and this is effected by 

 aiding absorption and retarding evaporation. The soil in which the roots lie must 

 be well moistened, and the air kept saturated with the object of reducing tran- 

 spiration. In our native trees bleeding may be observed most conveniently 

 in early spring before the leaves come out, because, owing to the activity of 

 the root, all the cells are saturated, and also because loss arising from transpira- 

 tion is at a miminum. If, on the other hand, the tree be felled in summer, we 

 find that, even after the soil has been well watered, not only does the cut 

 surface exude no fluid, but that water poured upon it is at once greedily ab- 

 sorbed. But if, finally, a plentiful supply of water is collected by the root- 

 system, then bleeding shows itself, and a positive root-pressure may be induced 

 where previously a negative pressure, that is to say, a pressure less than that 

 of the atmosphere existed. 



A third condition of bleeding is a certain temperature, differing markedly 

 of course in different plants. At o C. very few plants exhibit bleeding ; others, 

 such as the gourd, begin to bleed at 7 -o, C., whilst in all plants an increase 

 in the amount of fluid given off accompanies an increase in temperature. 

 Detailed investigations on this subject are, however, still wanting. 



In addition to the three factors referred to above, there is yet another 

 which is of some significance, viz. light, but this factor we need not discuss. 

 Variations in these conditions are accompanied by corresponding variations 

 in the amount of fluid excreted, and in the pressure exerted by the sap, 

 and it would be natural to attribute the periodic fluctuations described above 

 to the influence of these external factors. The researches of BARANETSKY 

 appear to support this view, but those of later investigators scarcely confirm it. 



The individual eccentricities are indeed quite incomprehensible. One of 

 the most striking illustrations of this is given by WIELER, who has recorded 

 diametrically opposite results arrived at from experimental observations made 

 on two plants of Alnus glutinosa, which were not only of the same age but were 

 similarly treated and studied under the same external conditions. While the 

 one exhibited a minimum exudation of sap in the forenoon and a maximum 

 in the afternoon, the other gave precisely converse results. When we recoiled 

 also that the same investigator was unable to demonstrate any periodicity at 

 all in the birch, we can have no hesitation in concluding that we are still tar 

 from having reached a satisfactory explanation of this phenomenon. 



To the factors above mentioned it is customary in many cases to add yet 



