of certain species of Willow. — Part 2nd. 247 



unicolor? Beau v., collaris Say, arvensis Say, hicolor Beauv., similis 

 Nort., and two others,) the anterior branch of the submarginal bullar 

 streak, instead of unitingwith the posterior branch, as in Fig. 1, A A', 

 fades out suddenly in the blackish or subhyaline membrane, which re- 

 places in this genus the 2nd submarginal cross-vein A A', so as to 

 form no connection whatever with the other branch ; though in a few 

 specimens there is seen in certain lights an indistinct fold adumbrating 

 the obsolete cross-vein, which fold, however, is not colored white like 

 the streaks. Surely, if this anterior branch were a true vein or artery, 

 it would form such a connection. What is very remarkable, though 

 Dosytheus apricus Nort. (= J), aprilis Nort.) has wings no more hya- 

 line than those of similis Nort. and sericeus Say, which exhibit dis- 

 tinct bullar streaks, in all my eight specimens of this species there are 

 no bullar streaks whatever perceptible, though the folds in the wing 

 are as distinct as usual. 



"Without venturing the assertion, that the bullae and the bullar 

 streaks have nothing whatever to do with the circulatory system in the 

 wings, it becomes, I think, sufficiently evident that they cannot per- 

 form the function of the veins in Vertebrata, the so-called wing-veins 

 acting as arteries, or vice versa. More than this, in the present state 

 of our knowledge upon this subject, it would be unsafe to assert. But 

 even assuming that they form some of the definite channels for the cir- 

 culatory system, spoken of by Prof. Clark as being without any appa- 

 rent walls, yet this is quite a different fact from their being coloration- 

 ally distinguished from the rest of the wing. If the bullar streaks 

 form such channels, it is reasonable to infer that similar channels exist 

 in all Hymenopterous genera, which have visible bullae but no bullar 

 streaks, and again, in all Hymenopterous genera wdiich have neither 

 visible bullae nor visible bullar streaks. Their supposed function as 

 blood-conducting channels without determinate walls, is a structural 

 fact; their being sometimes colored in a peculiar manner is a colora- 

 tional fact; and the two facts, as is abundantly shown by the phenom- 

 ena exhibited in the Hymenopterous wing, have no necessary connec- 

 tion with each other. Hence, whatever views we may adopt as to the 

 the circulatory system in the Hymenopterous wing, the peculiar color- 

 ation of both the bulla: and the bullar streaks, in such species of cer- 

 tain genera and subgenera as have blackish wings, (Fig. 1,) and even 

 in a few species (Dolerus sericeus Say and D. si7nilis Norton) which 

 have wings that are almost hyaline, the peculiar coloration of the bul- 

 lae alone in other genera, (Fig. 2,) and the total absence of any such 



