EXTERNAL COMMUNICATIONS OF THE AER1FLROUS SYSTEM 191 



of CO.,-assimilation, which is necessarily diminished almost to nil. The 

 formation of a cuticle and the regulatory action of the stomata is 

 essentially intended for the retention of the necessary amount of water, 

 and it is in view of this function that the special adaptive modifications are 

 to be regarded (cf. Sect. 38). 



It is hardly surprising that in a few cases true stomata, as well as water 

 stomata and the peculiar stomata of Marchantia, close only partially or not at all 

 when the plant becomes flaccid. According to de Bary, this is the case in the 

 stomata on the leaves of Kaulfnssia, and according to Schwendener in those of 

 Cynosurus echinatus and a few other grasses \ Schwendener's conclusion that non- 

 closing stomata were the rule in aquatic plants has not been found absolutely 

 correct, although Stahl found that frequently in cases of plants of marshy habitats 

 (Alisma planfago, Acorus calamus, &c.) the stomata remain open in flaccid leaves, 

 which is a property exhibited also by many woody plants which prefer a wet soil 

 (Salix, Alnus\ and by other plants as well 2 . 



The classical researches of Mohl have shown that the width of a stoma 

 varies according to the turgidity of its guard-cells, and the same author 

 has proved that stomata which open in water may be caused to close by 

 the plasmolyzing action of a sugar solution (1. c., p. 702). The movements 

 and changes of form of the guard-cells due to these variations of turgidity 

 have been more closely studied by Schwendener and his successors, but 

 Mohl was the first to prove that in the isolated stomatal apparatus an 

 increase of turgidity causes a widening of the stoma, a decrease leading to a 

 narrowing of the aperture, and ultimately to complete closure. By means 

 of comparative experiments, in which, when necessary, all the cells in the 

 epidermis surrounding the guard-cells were ruptured, it was found that the 

 antagonistic action of the neighbouring epidermal cells might more or less 

 markedly influence the movements of the guard-cells, so that in extreme 

 cases, instead of the attempted opening, a closure of the stoma might be 

 produced. 



Schwendener (1. c., 1881, p. 853), Haberlandt (1. c.), Schafer (1. c., p. 204) 

 recognized the influence exerted by the neighbouring cells, but did not attach much 

 importance to it, whereas Mohl (1. c.) and Leitgeb arrived in certain cases at a 

 contrary opinion. Thus both the latter saw that the removal or rupture of the sur- 

 rounding epidermal cells caused the stomata of many plants to open more widely, 

 while similar treatment caused stomata to open which had partially or entirely 

 closed owing to the immersion of the leaves in water. This closure in water was 

 seen by Mohl 3 to take place in various grasses, and also in Amaryllis formosissima, 



1 De Bary, Comp. Anat., 1877, p. 58 ; Schwendener, Sitzungsb. d. Berl. Akad., 1889, p. 69. 



2 See Kohl, Transpiration d. Pflanzen, 1886, p. 25 ; Haberlaudt, Flora, 1887, p. 100; Schaefer, 

 Jahrb. f. \viss. Bot., 1888, Bd. xix, p. 196; Stahl, Bot. Zeitung, 1894, pp. 123, 137. [Rosenberg 

 (Ueber die Transp. d. Halophyten, Kongl. Vetenskap. Akad. Fb'rhandlingar, 1897, No. 9, p. 531) 

 finds that numerous Halophytes are able to clo^e their stomata.] 



3 Mohl, 1. c. Confirmatory results by Unger, Sitzuiigsb. d. "VYien. AkaJ.. 1857, Bd. xxv. p. 468; 



