24 



ler branch to the alimentary canal and a larger, outer one with branches in a star- 

 like manner; one of these branches is stronger than the others and runs into the 

 preceding segment. In the eighth segment a strong trachea branches off, running 

 into the anal segment, sending out branches to the gills. 



The function of this very peculiar tracheal system is difficult to understand. 

 I have tried to understand the mode of action of this peculiar tracheal system, it 

 being as far as I know unique among the insects. Later on Prof. KROGH has been 

 kind enough to read my manuscript and has discussed the problem with me. 



From the very broad band-like tracheal trunks it can only be concluded, that 

 the larvae must be able to produce a very powerful respiration. This may probably 

 be of great significance for the larva, because the air in the plant from a respira- 

 tory point of view is a very bad medium, by no means rich in oxygen. This has 

 been ascertained by EGE (1915 p. 183). The main result of his investigation is "that 

 the composition of the air in the intercellular spaces of different aquatic plants is very 

 variable. The oxygen percentage is low, rarely higher than 10 / and may especi- 

 ally in winter sink to about 1 2 /o and even still less". When we remember that 

 the larva probably fixes itself to the plant until September and probably detaches 

 itself in May, the animal living as imago or egg during the time May September, 

 it will be understood that the larva is dependent on the plant air, especially at 

 the season when the oxygen percentage is lowest. For a long time I therefore 

 thought that the tracheal bladders might be used to pump the air from the plant 

 into the body of the larva by pressure and dilatation. Prof. KROGH has now at my 

 request been kind enough to go into the matter and tells me that this however 

 cannot be the case. He pointed out that the larva, as far as he could see, made 

 no visible respiratory motions. He varied the composition of the air in the inter- 

 cellular spaces of the plant, but even then, if the amount of oxygen only was 

 three per cent., no respiratory motions were traceable. If the larvae got only nitrogen, 

 all motion ceased and the brushes stopped; if then atmospherical air was again 

 conducted through the plant, they were soon restored. If the larvae were exposed 

 to various different pressures, it could be shown that the tracheal system really 

 would be somewhat compressed; but the compression of the tracheae which was 

 the result of the pressure, possible for the larvae themselves to procure by 

 the contraction of their body, (not more than 0'2 atmospheres), is so inconspi- 

 cuous that it would be without any practical significance, even if the larva made 

 respiratory motions. Furthermore it could be shown, that it was only the flat band- 

 shaped tracheal trunks and not the bladders which were compressible. 



It must therefore be taken for granted that the transport of oxygen from the plant 

 to the larva only can take place by means of diffusion. The bladders on the tracheal 

 system are said in Tceniorhynchus to have quite another function. They augment 

 the respiratory surface and the amount of air which is at the disposal of the larva. 

 Further they act hydrostatically ; they diminish the weight of the animal, more especi- 

 ally of the anterior part, which would otherwise be much heavier than the posterior. 



