The fourth sub-class, Tinea, far exceeds in numbers either of the others, and probably all of them together; and where such a multitude of species exists, great diversity in form and habit may be expected: the Pterophori are a most singular tribe, and greatly resemble the Tipulæ in many respects. I feel by no means certain that their situation would not be better between the lepidopterous sub-class, Tinea, and the dipterous sub-class, Tipulæ, thus throwing them completely out of the lepidopterous circle; but this I leave. I am now only sketching a rough and hasty outline from nature. If I attempt to finish my drawing as I proceed, I shall find occupation sufficient for a lifetime. I have observed that I considered the chain of relation entering from the last sub-class at Chilo, or about that genus; the same order must of course include Crambus, and its congeners; the next order will contain Yponomeuta, which I will place at the point of contact; and the next point being among the true Tortrices will drive Halias fagana as a decided departure from their typical form to the very circumference of the circle where it touches Noctua. The fifth sub-class, Noctua, seems to be but one mighty genus: we will enter it from Halias fagana, an insect so nearly allied to Noctua in its larva, its pupa, and its imago, that for a long time I hesitated to which sub-class it belonged; again, in Cymatophora,[38] subtusa and retusa, I was fearful that by considering them Noctuæ, I might deprive the order Tortrices of a genus on which perhaps many curious combinations might depend, and I now only place them in Noctuæ until I may have an opportunity of examining their larvæ, which I have not yet been fortunate enough to meet with. Towards the central sub-class there appear to be many genera which approach the line of contact; Agrotis and Chareas for instance:[39] I prefer taking the latter, and must mention the species Graminis, as I am fearful of encumbering my system with species to which I not only never intended to refer, but should probably place in some distant order, or perhaps sub-class. At the approach to Geometra, the genus Catocala, from its looping larva, seems to have a right to be placed: this I, however, look on with suspicion, as the larva appears to me any thing but a guide in the connexion of sub-classes; but I here succumb to customary usage in making this genus the approach to the real loopers, objecting, however, to the intervention of Phytometra, Euclidia, and Brepha. [38] Ochsenheimer places Oo in this genus, and I observe Mr. Stephens confines the genus to that one species. Mr. Curtis places Oo in the genus Bombycia: this confusion of genera is very puzzling, but I hope, by mentioning species, to make myself understood. Oo is not at all applicable to my purpose. [39] Perhaps Noctua Lambda.