96 DIFFUSION AND OSMOTIC PRESSURE 



urated walls brings about all the transfer that is required. 

 The process may often be hastened by pits and protoplasmic 

 connections. In higher plants phenomena occur which are 

 more difficult of explanation. In order to discuss the trans- 

 mission of water in such cases, it will first be necessary to 

 consider briefly the ways in which water is lost by the plant. 

 (1) Evaporation. As has been stated already, by far the 

 greater part of the water absorbed by the plant is lost by 

 evaporation. Water is continually evaporating into the air- 

 spaces of the plant body and diffusing out into the external 

 atmosphere through stomata and lenticels, and to some extent 

 through the epidermis itself. The vacuoles of the leaf 

 parenchyma furnish this water to the cell walls, and thus 

 their solutions become more concentrated as evaporation con- 

 tinues. Of course this means that water must diffuse into 

 these cells from cells farther back, where the concentration 

 is not as great. Eventually these cells draw water osmotic- 

 ally from the xylem strands. The source of energy for this 

 process of concentration of leaf solutions is the heat of the 

 surrounding atmosphere, which causes the aqueous molecules 

 to break away from the films covering the parenchyma walls 

 next to the air chambers. 



(2) Water pores and nectaries. Water pores and nec- 

 taries are curious instances of the passage of liquid water 

 out of the plant body. No completely satisfactory explana- 

 tion of these occurrences has yet been given. Osmotic action 

 surely plays an important part here, but as yet no accurate 

 means of determining the exact process has been devised. 

 The difficulty lies in the fact that in these cells the move- 

 ment of the water is in the opposite direction from that which 

 would be expected from the principles of osmotic action. The 

 following hypothesis may help to bring the facts together : 



The modified cells bordering a water pore are perhaps 

 irritable in a peculiar way. It may be that the protoplasmic 



