430 RECTAL GILLS IN LARViE OP ANISOPTERID DRAGONFLIES, 



The method of filling the trachece with gas in the newly -hatched 

 larva. — It has been usually stated that, soon after hatching, 

 isolated globules of gas appear in the tracheae at various points, 

 and that these, by increase and coalescence, gradually drive the 

 fluid out of the tubes. My own observations!?), which support, 

 in their chief points, those previously made by Calvertd), show 

 that this is not so. In the case of Anax papue7isis Burm., I 

 found that the embryo, just prior to hatching, had its tracheae 

 filled with liquid, as is agreed by all observers. The emergence 

 takes place very rapidly. Owing to the interpolation of a pro- 

 nymphal stage, the larva has to escape not only from the egg- 

 shell, but also from the pronymphal sheath almost immediately 

 afterwards. Thus there is a period of at least half a minute 

 during which observation of the changes going on in the tracheae 

 is exceedingly difficult, if not quite impossible. However, at the 

 moment the larva becomes free, gas can be seen travelling evenly 

 down the two dorsal trunks from the region of the midgut hack- 

 wards. It is, of course, impossible to analyse this gas; but, as we 

 shall see later, the point is immaterial. The importance of the 

 observation lies in the fact that the gas does not form in the rectal 

 regio7i, i.e., it does not, at the start, pass in from the rectal cavity 

 to the tracheae. As the gas only travels down the dorsal tracheae, 

 and not down the ventral or visceral trunks, it seems almost 

 certain that it is derived from the head or thorax, since these 

 portions are entirely supplied by the dorsal tracheae. We might 

 mention, in this connection, the head-vesicle and the cephalic 

 heart as offering a probable solution to this difficult question. 

 Certain it is, that the pulsations of the cephalic heart synchronise 

 closely enough with the period of the filling of the dorsal trunks 

 with gas, to suggest that there is more than a chance-connection 

 between the two. But this problem cannot yet be definitely 

 solved. It is sufficient, for removing the obstacle supposed to 

 stand in the way of an acceptance of the Diffusion-Theory, that 

 we have shown that this obstacle does not really exist. We can 

 take it as proved that, within a minute or two after hatching, 

 gas fills the tracheal system of the larva, and that it comes in, 

 not via the rectum, but from somewhere in the anterior end of 

 the animal. 



