IV.T VARIOUS tHIlORlElS, <&€. 125 



placed at numerous successive levels in the stem, act 

 alternately as suction and force pumps : they absorb 

 water forcibly by osmosis, and they drive it out again 

 (also forcibl}^) by exfiltration at a higher level. 



The chief difficulties which face the hypothesis are 

 those which recur when we try to explain root-pressure, 

 and yet it is certain that the living cells of the root- 

 hairs, and root-epidermis, &c., absorb water from the 

 soil and force it into the axial vessels. 



Now the phenomenon of " root-pressure " can be 

 got with pieces of older roots, or even bits of stems 

 put into wet sand ; and Hofmeister, Russow, Kraus 

 and others have conjectured that the living cells of the 

 wood must co-operate in producing root-pressure, and 

 it is difficult to see how it can be otherwise in such 

 cases as the above. 



The next difficulty is to explain how a cell can thus 

 take up water by endosmose, and then drive it out under 

 pressure ; for it is impossible to accept Hofmeister 

 and Sachs's hypothesis that the cells are differently 

 constructed on two opposite sides. Moreover, the 

 apparatus designed by Sachs^ to explain root-pressure 

 will not work — it contradicts the conservation of 

 energy. 



^ TLis apparatus is repiesented in Fig 213, p. 276, of Sachs's Lee- 

 hins on the Physiology of Piants, English edition 



