430 Mr. Newport on the Anatomy and 



and metathorax, below the second, or inesothoracic spiracles, while the other 

 projects from the anterior of the metasternuin. The sixth, like the fourth, is 

 only a single pair of sacs in the soft membrane of the articulation of the coxae 

 of the third pair of legs {h) ; while the seventh and eighth, also single pairs 

 of sacs, project from the inferior lateral surface of the first and second ab- 

 dominal segments, in situations nearly corresponding to the usual place of 

 spiracles in other insects. 



M. Pictet has described the branchial tufts in the larva of Perla bipunctata 

 very precisely, and has successfully combated the opinion put forth by M. Bur- 

 meister, in opposition to his view, that the branchial filaments are only stiff 

 hairs. A careful examination of the tufts in Pteronarcys has confirmed to 

 me the correctness of M. Pictet's observations as regards their true nature. 

 Each tuft or sac (fig. 3) is an extension outwards of the soft tegument from 

 which project an abundance of delicate csecal filaments. Each filament (fig. 4) 

 is a simple, unarticulated, uniform structure, slightly tapering and closed at 

 its extremity, and in the interior of which there is an extremely minute tra- 

 cheal vessel (c). On examining some of these filaments taken from the 

 branchiae of my specimen of Pteronarcys formerly, in company with Professor 

 M.- Edwards, we were unable, at that time, to satisfy ourselves of their true 

 branchial function ; but longer-continued, repeated, and more carefully con- 

 ducted investigations have now most fully satisfied me of their real import- 

 ance as active organs in the imago. The uncertainty of former examinations 

 arose, as I now find, from the branchial filaments being greatly altered in 

 their appearance by the contraction of their fibrinous tissue, together with the 

 coagulation of the circulatory fluid and blood-corpuscles within them, occa- 

 sioned by the insect having been killed and preserved in spirits. I have 

 since recognised corresponding appearances, induced by a similar cause, in 

 the branchiae of other insects killed in like manner. 



The number of filaments produced from each sac varies from about twenty 

 to fifty or more. It is greatest in the sacs of the meso- and metathorax, and 

 smallest in those of the neck and of the abdomen. The filaments originate in 

 little bundles, four or five in each, from the distal border of the sac, but not 

 all on precisely the same line. Usually each filament is simple and distinct ; 

 but in a few instances, as in some (fig. 4) from the external sac at the anterior 



