RESPIRATION 15 



from the spiracles to the tracheoles. It was suggested long ago 

 by Treviranus (1816), and later by Thomas Graham (to whom wt 

 owe the laws of diffusion of gases), that it must be conveyed by diffu- 

 sion. But to the superficial observer of these fine tracheal tubes, this 

 idea seemed so inherently improbable that various other hypotheses 

 were put forward, and it is only within comparatively recent years 

 that the diffusion theory of insect respiration has been generally 

 accepted. 



The present acceptance of this theory is due to the work of Krogh 

 (1920), who, taking into consideration the mean diameter and length 

 of the tracheae, the oxygen consumption of the insect, and the diffu- 

 sion coefficient of oxygen, calculated what difference in partial pres- 

 sure between the atmosphere and the tracheal endings would be 

 necessary to maintain the supply of oxygen that was actually con- 

 sumed. He obtained the surprising result that, even in the case of 

 large caterpillars (assuming that the spiracles remain permanently 

 open), the partial pressure of oxygen at the tracheal endings need not 

 be more than 2 or 3 per cent, below that in the atmosphere; showing 

 that diffusion alone was adequate to supply the needs of the insect. 

 The same is true of the tracheoles : even in the active flight muscles of 

 insects, gaseous diffusion will account for all the oxygen which the 

 cells require. But when the muscle fibres exceed about 20 [x in dia- 

 meter the tracheoles indent the surface membrane and become 

 'internal'. 



The same argument applies to the elimination of carbon dioxide. 

 The amount of carbon dioxide produced by the insect is usually 

 rather less than the oxygen taken in; their rates of diffusion^ being 

 proportional to the square roots of their densities (v'22: V16), are 

 not enormously different. Therefore, since diffusion alone will ac- 

 count for the supply of oxygen, it will account equally for the elimin- 

 ation of carbon dioxide. But carbon dioxide diffuses through animal 

 tissues, and therefore presumably through the insect cuticle, more 

 rapidly than oxygen. Consequently, since carbon dioxide liberated 

 in the tissues must diffuse equally in all directions, the amount of 

 this gas which will escape directly through the cuticle of the body 

 wall (especially when this is thin) and through the walls of the larger 

 tracheal trunks, must be far greater than the amount of oxygen that 



