35 



wing being furnished with a simple costal, a 5-branchecl post-costal, 

 a 4-branched median and a simple anal vein. Now this is the typical 

 number of branches which a lepidopterous wing ought to possess, 

 according to the theory of Mr. Edward Doubleday, that we are to 

 suppose the existence of a discoidal vein traversing "the middle of the 

 discoidal cell, and that this discoidal vein, as well as the post-costal 

 and median, are respectively furnished with three branches. Ac- 

 cording to this theory therefore, the two branches of the post-costal 

 vein which run to the tip of the fore-wing of Endromis, together with 

 the first branch traversing the front of the disc of the apical portion 

 of the wing, are the only real branches of the post-costal vein ; the 

 two foUovnng branches of the post-costal vein, as I have regarded 

 them, and the first branch of the median vein, are the branches of the 

 supposed discoidal vein, and the three remaining branches of the me- 

 dian vein are its only true branches. I do not intend in this place 

 to enter into a detail of the reasons which induce me to refuse assent 

 to this theory ; I may however observe, 1st, that mth regard to the 

 functions of these branches, it is evident that the fourth branch of the 

 median vein, where present, must form a portion of the system of cir- 

 culation effected by the branches of the mechan vein, just as in like 

 manner the three branches of the post-costal vein of Saturnia, which 

 traverse the apical portion of the fore-wing, must be considered as ef- 

 fectually forming a portion of the post-costal vein ; 2ndly, that it seems 

 to me contrary to analogy to admit the existence of fully-developed 

 branches of a vein, the base of which has no real existence ; and 3rdly, 

 that instances occur (e. g. Psyche Stettinella, Cochhophasia tes- 

 sellea) in which the number of branches exceeds the supposed typical 

 number of nine {i. e. three post-costal, three discoidal, and three me- 

 dian), those insects having ten branches, in which case one of the 

 veins must have an extra branch ; whilst in Saturnia for instance, the 

 supposed discoidal vein can only have two branches, — hence I see no 

 reason why cases may not be supposed in which one vein should have 

 more, and another vein fewer, than the typical number of branches ; 

 or, in other words, why the median vein in Fmdromis should not 

 have four branches, whilst there are only five branches for the post- 

 costal and supposed discoidal veins. 



The antennae also of Endromis, as well as its transformations, are 

 quite different from those of Aylaia and Saturnia ; indeed the tribe 

 Endromides of Boisduval seems to possess no single connecting cha- 

 racter. 



Hiibner, in his • Verzeichniss bekannter Schmetterlinge*,' has at- 

 tempted an arrangement of these insects which appears to me unna- 

 tural, so far as the primary divisions are concerned, whereas his inferior 

 groups {Coitus), founded almost entirely upon the form and marking 

 of the wings, appear to bring together the closely allied species. His 

 first tribe of the Bombycoid Nocturnal Lepidoptera is termed Sphin- 

 ffoides, and contains five stirpes: — 1st, Dimorphce {Endromis, Chao- 

 nia, Petasia, &c.) ; 2nd, Ptilodontes (the Prominent Moths) ; 3rd, 



* Augsburg, 1816, 8vo. 



