ABSORPTION OF WATER BY LEAVES 237 



subsequent chapter (Chap. X., II., A.). The principal function of 

 hydathodes is the secretion of water, but many of them are also capable 

 of absorbing liquids. It is, in fact, impossible to draw any sharp dis- 

 tinction between secretory hydathodes and water-absorbing trichomes. 

 After a certain stage in the life of a leaf the hydathodes not infre- 

 quently devote themselves exclusively to absorption. Again, in closely 

 related species which differ greatly in their mode of life, homologous 

 trichomes may in one case serve as hydathodes, and in another act as 

 absorbing hairs. 



In spite of a number of statements and suggestions to the contrary, 

 it may be regarded as an established fact that leaves never absorb 

 water through their stomata, at any rate under normal conditions. 

 Apart from the strictly localised " water-pores," stomata serve solely 

 as the external openings of the ventilating system, and for this reason 

 are actually furnished with numerous devices for preventing capillary 

 occlusion of the stomatic pores and access of water to the inter- 

 cellular spaces. Leaves that have been immersed in water for purposes 

 of experiment may become translucent in places, owing to the partial 

 injection of the intercellular spaces with water which has entered by 

 the stomata ; but this phenomenon must be regarded as a result of the 

 unnatural conditions to which the leaf is exposed and not as an illus- 

 tration of the normal action of the stomata. 



Absorbing hairs are found on the leaves of many plants of hot, dry 

 climates, but especially among species inhabiting desert regions ; it is, 

 in fact, the possession of such hairs that enables desert-plants to make 

 use of the nocturnal dews which in the Egypto-Arabian desert, for 

 instance, occur almost nightly from November to April, and often col- 

 lectively represent a very considerable amount of precipitation. Many 

 members of the Mediterranean flora are also provided with absorbing 

 hairs, and similar organs doubtless occur in a number of our native 

 plants which inhabit dry, sunny situations. As regards epiphytes, 

 water-absorbing trichomes have hitherto only been recorded in the 

 Bromeliaceae, where, however, they attain a very high degree of 

 specialisation. 



In respect of their external form, water-absorbing trichomes may 

 be developed as ordinary simple hairs or bristles, as capitate or clavate 

 hairs, or, finally, as peltate hairs or scales; both unicellular and multi- 

 cellular forms are known." 



Unicellular absorbing hairs occur, according to Volkens, in Diplotaxis 

 Harm and in certain species of Heliotropium inhabiting the Egypto- 

 Arabian deserts. The author's observations upon material collected by 

 himself confirm Volkens' statements in all essentials. The leaves of 

 Diplotaxis Harm (Fig. 92 a) bear a number of stiff, spreading bristles, with 



