PASSAGE OF GASES THROUGH CELLS AND CELL-WALLS 185 



rapid, gas bubbles may appear on the submerged surface of a diosrnotic 

 membrane. 



All the experimental evidence shows that the more soluble carbonic 

 acid diffuses much more rapidly than oxygen does through cell-walls 

 saturated with water, and that oxygen diosmoses more rapidly than 

 nitrogen, which is still less soluble. This is admirably indicated by 

 the rapid collapse of a moist bladder filled with carbonic acid gas and 

 surrounded by air. Devaux's l researches upon diosmosis through entire 

 cell-layers of living aquatic plants gave rates of transference precisely 

 similar to those for water-lamellae, and probably the same would hold good 

 for all completely saturated membranes capable of pronounced imbibition. 



Gases are unable to diffuse at the same rate through a cuticularized 

 membrane impregnated with wax as through unaltered cellulose, for the 

 rate of transference depends upon the nature of the imbibed substances. 

 It appears, however, that carbon dioxide diffuses more rapidly through the 

 cuticle, or through a lamella of india-rubber, than oxygen does, and the latter 

 more rapidly than nitrogen, although no decisive conclusion can be made 

 from the researches mentioned 2 . 



The impregnation of the cuticle and cork markedly decreases their 

 permeability for gases, and hence in general the more pronounced the im- 

 pregnation and the consequent depression of transpiration, the slower the 

 diosmosis of gases will be (cf. Sect. 21.) The relationships are similar to 

 those shown when a piece of paper is impregnated with fat, and although 

 comparative experiments have not as yet been made, there can be no doubt 

 that in the cuticle also the permeability for gases decreases less rapidly 

 than that for water, as the impregnation becomes more and more complete. 

 It is not impossible that in this way a membrane may become totally 

 impermeable to water, and yet allow gases to pass through to a considerable 

 extent, as is actually the case with a film of india-rubber. 



In any case, it is of great importance that oxygen shall be able to 

 diosmose to a sufficient extent even through a strongly developed cuticle 

 (Sect. 29). This is shown by the fact that when the basal ends of hairs 

 of Tradescantia, Momordica, Urtica, &c., are imbedded in vaseline, the 

 protoplasm may continue to show fully active streaming movements, for 

 streaming is dependent upon respiration, and in this case the whole of 

 the oxygen necessary for respiration must pass through the cuticle 3 . 



certainly an error, although it is quite possible that by compression the pressure may be insufficiently 

 increased to cause perceptible gaseous diosmosis under the given experimental conditions. 



1 Devaux, Ann. d. sci. nat, 1889, vii. ser., T. IX, p. 63. 



2 Wiesner, 1879, 1- c -> P- 74 > Mangin, Ann. d. sci. nat., 1887, T. civ, p. 1809 ; X. J. C. Miiller, 

 Jahrb. f. wiss. Bot., 1869-70, Bd. VII, p. 169; Barthelemy, Ann. d. sci. nat., 1874, v. ser., T. xix, 

 p. 138. Cf. Winkelmann, Physik, 1891, Bd. I, p. 657; Kayser, Ann. d. Physik u. Chemie, 1891, 

 Bd. XLIII, p. 544. 



3 [This proof is only satisfactory when sti earning continues under such conditions for prolonged 



