TRACHEAL RESPIRATION 135 



brought down from the. surface by diffusion of oxygen from 

 the water, the air-water interface acting as a kind of gill. 

 This conception was first put forward by Comstock (1887), 

 who noticed some curious reactions in Notonecta which in badly 

 aerated water apparently attempted to increase the surface 

 of its air supply, but it was worked out quantitatively by Ege 

 (1915) who also defined the conditions and limitations. 



When an insect carrying a sheet of air on the outside stays 

 under water for some time, a difference in oxygen tension 

 between the water and the air carried down will be set up by 

 the metabolic processes, and consequently oxygen will diffuse 

 in from the water. At a certain tension difference the rate of 

 oxygen uptake may be sufficient to cover the metabolic re- 

 quirements of the animal, but Ege showed that even in this 

 case the air will have to be renewed from time to time at the 

 surface. This is due to the tension difference set up for nitro- 

 gen which will cause the total quantity of air to decrease all 

 the time. The 2 used up is replaced with C0 2 , but this 

 diffuses away so rapidly that the nitrogen percentage rises. 

 Supposing that the composition at equilibrium is 5% 2 and 

 1% C0 2 tjae nitrogen percentage will be 94%, correspond- 

 ing at the surface to 94/100 X 760 mm = 713 mm and at 

 a depth of 1/2 meter to an additional pressure which is 



1 ' ( 



94/100 X -TyTT = 36 mm or 749 mm in all, while the tension of 



the nitrogen dissolved in the water is only 79/100 X 760 = 633 

 mm. Ege made a number of micro-analytic determinations 

 of the actual composition of the air on Corixa, Notonecta, and 

 some of the Dytiscidce which fully bear out his calculations. 

 They show a rapid reduction of the 2 content with only a 

 slight increase in the C0 2 . The experiments made show that 

 a Corixa normally gets so much oxygen out of the water by 

 diffusion into its air-covering that the supply will last 10 to 30 

 times as long as would otherwise be the case. One experi- 

 ment is specially instructive. A Notonecta was put in water 

 saturated with oxygen and allowed to breathe also from an 

 atmosphere of pure oxygen. In this case no perceptible diffu- 



