ON THE RESPIRATORY FUNCTION OF STOMATA. 17 



given conditions across the isolated cuticle ; and secondly, 

 of the quantity of this gas absorbed or exhaled by the organ 

 to which the cuticle belongs, in full activity. If the former 

 is less than the latter, stomata must come into use, but if 

 the former is the greater, then osmosis may be sufficient to 

 account for all the exchange. 



The diffusion experiments are carried on in an ingenious 

 apparatus invented by the author, in which the cuticle under 

 investigation serves as a diaphragm between two chambers, 

 one of which is kept tilled with CO,, while the other is 

 closed, contains some other gas, and is connected with an 

 adjustable manometer, the movements of which record the 

 amount of gas diffusing into or out of the chamber through 

 the diaphragm of cuticle. 



The method appears excellent, but the materials used 

 seem to be open to objections. The cuticle of various 

 leaves is isolated, by allowing the leaf to remain macerating 

 for some weeks in a culture of Bacillus Amylobacter, which 

 gradually destroys the rest of the tissues. Pieces of cuticle, 

 two to four centims. square, are thus obtained and employed. 

 In order to render them more resistant, they are painted 

 over with a 10 per cent, warm solution of glycerine jelly. 

 Of this it is alleged that it produced no appreciable increase 

 of the resistance of the cuticle to the passage of gases, and 

 that it will serve to block up the stomata and cracks which 

 have hitherto been the oreat stumbling-block in such dif- 

 fusion experiments. Surely it must still be objected, that, 

 since it is so permeable, gas will pass much quicker through 

 those spots where there is nothing but a film of glycerine 

 jelly covering a stoma or crack than through the unper- 

 forated cuticle. The osmosis numbers obtained one would 

 expect to be much too high. Thus, prepared cuticle gives 

 numbers for the rates of diffusion of equal volumes of 

 different gases agreeing closely with those obtained by 

 Graham for caoutchouc. 



Fallacious as may be the absolute numbers, the relations 

 between different cuticles are of interest. The cuticle of 

 the lower surface of the leaf is always found to be more per- 

 meable than that of the upper, but never more than five 



