19G Dr. T. Williams on the Mechanism of Aquatic 



penetrated by a single trachea and an advancing and returning 

 current of blood. In the Agrionida (fig. 5, a) they assume the 

 character of lancet-shaped processes attached to the sides of the 

 abdomen at the points of the future spiracles. Examined care- 

 fully as transparent objects, the trachese of these branchiae 

 divide and subdivide much more elaborately than is commonly 

 supposed (fig. 5, B, 6). It is only the larger- tracheae that are 

 accompanied by a current of blood (fig. 9, i, h, fig. 12, e). The 

 latter is much less profusely subdivided than the former. This 

 fact seems incontestably to prove that the tracheae, not the blood- 

 channels, extract the air from the surrounding water. In the 

 anal branchiae of the LibellulidtE, M. Leon Dufour exhibits the 

 tracheae (fig. 7) as terminating in Z'M^/ec? extremities (fig. 7, 13, a). 

 The author will only state that he has never, in the course of his 

 numerous researches, in any instance met with this mode of ter- 

 mination. In the filiform branchiae of the larva of gnats each 

 trachea tapers to the finest extreme. 



In Pteronarcys regalis Mr. Newport describes the branchial 

 filaments as consisting " each of a simple, unarticulated, uniform 

 structure, slightly tapering and closed at its extremity, and in 

 the interior of which there is an extremely minute tracheal ves- 

 sel* terminating in delicate caeca.'' In no one of his writings is 

 it evident that Mr. Newport is aware to what an extreme degree 

 of capillary subdivision the tracheae are carried in the flat branchice 

 (fig. C, B, c), if not in the filiform, of the larvae of Insects. In 

 those of the Agrionicke, cut off and examined separately under 

 the microscope, they cannot be followed by the highest powers of 

 the microscope. The blood-current turns back at the larger 

 branches : it does not ramify in network streams. It is obviously 

 not designed to fulfil the office of breathing : this function falls 

 upon the tracheae. This conclusion is opposed to the views of 

 Mr. Newport. " The blood-corpuscles of the whole body circulate 

 through the branchiae for the purposes of respiration. The 

 current of the blood is always in the vicinity of the tracheae, 

 absorbing oxygen by endosmose and giving out carbonic acid. 

 This takes place in every form of hranchiat f." The author is fully 

 satisfied that this is an erroneous interpretation of the respii*a- 

 tor^ process as it occurs in the branchiae of Insects. The rami- 

 fications of the tracheae in these organs are far more elaborate 

 than Mr. Newport and other observers have ever yet supposed. 

 They render the inference irresistible that the branchial respi- 

 ration of the Insect is really atmospheric in type. The air does 

 not, as in fish-breathing, enter immediately into the blood. 



In the vessels of Insects in every phase of life, there seems to 



* Liniiffian Transactions, 1851. t 0/j. ei/. p. 432. 



