BY R. J. TILLYARD. 431 



Now let us follow the gas-flow along the dorsal trunks. It 

 travels steadily backwards to the rectal region. The primary 

 efFerents are first filled, then the secondary, then the smaller 

 branches, and, finally, with considerable rapidity, gas passes up 

 around the capillary loops, completely displacing all the liquid. 

 The gas also passes from the dorsal tracheae by means of short 

 connecting branches into the ventral and visceral trunks, so that 

 the whole system is very soon filled. The process is \ery easy 

 to follow, since the gas shows up like a black rod travelling 

 along the colourless tubes. 



The expulsion of the liquid from the capillaries and larger 

 tracheae, which can be clearly seen taking place, is not difficult to 

 understand if we assume (as is generally agreed) that the liquid 

 is blood. If so, it passes into the haemocoele, and the connection 

 between the dorsal tracheae and the aorta must be sought for in 

 the cephalic heart. 



As soon as the filling of the tracheal system with gas is com- 

 pleted, rectal respiration begins. The rectal valves immediately 

 come into play vigorously, so that the rectal cavity is soon 

 strongly distended with water. For a short time, the action is 

 exceedingly vigorous; then it settles down to the usual slow and 

 somewhat irregular rate. 



Thus, within a few minutes of the hatching of the larva, all 

 the conditions necessary for the passage inwards of oxygen by 

 diff'usion are fulfilled. For, whatever be the proportion of 

 oxygen to nitrogen in the gas in the tracheal system, it is quite 

 clear that the partial pressures of both these gases must always 

 tend to become equal on either side of the difi'usion-membrane. 

 We may take it that the water in the rectum is fairly well 

 aerated, i.e., contains oxygen and nitrogen dissolved in the ratio 

 of one to two.* If, then, the proportion of oxygen to nitrogen 

 in the tracheae be originally greater than this, it will be lowered 

 by the passage of nitrogen inwards from the water in the rectum, 



* Or, more correctly, out of every 100 parts of gas dissolved, 35 are 

 oxygen and 65 nitrogen. Tlie result is obtained from the ratios of the 

 products of the coefficient of solubility and the partial pressure of the gas 

 at the water-surface. 



