either pair of classes, with the addition of class VII., may be formed into a tolerably perfect chain of genera, indeed with much less appearance of disconnexion than is observable on passing from either pair into the next pair,—a fact which attaches a degree of importance to the number three, on which, perhaps, at a future time, more may be said,—and thus a chain of relation would be established in each instance, leaving four whole classes entirely out of the question;—a chain which would steadily pursue its way, regardless and in open violation of all established laws of analogy, affinity and dichotomy; laws which I hope ere long to see pining away like Echo, until they also are really what I now fully believe them to be, vox et præterea nihil.

Mr. MacLeay found that in his quinary groups one of each five contained genera or species related to other genera or species in each of the other four groups. That I may be thoroughly understood, I will quote the author's own words:—"In almost every group which has been set before the reader, he must have perceived that one of the five minor groups into which it is resolvable, bears a resemblance to all the rest; or, more strictly speaking, contains types which represent each of the four other groups, together with a type peculiar to itself."[26] As far as my observation has extended, this is universally the case; and whether the total number of groups be five or seven, I think I am safe in asserting that the only possible way of making these types, thus representing groups, approach such groups, is to place the heterogeneous group in the centre, and the homogeneous groups around it; taking care that the type peculiar to itself be its very centre, its "heart's core." Such a heterogeneous group, then, is Neuroptera: its characters as given,[27] I believe, perfectly correct; and can any one say they are sufficient? Certainly not; but had I described it thus—Class VII. Neuroptera, central, partaking of the characters of all the others, I think a better character could not have been given. This class contains a type peculiar to itself—the genus Libellula of Linnæus: a genus so distinct, that several authors have supposed it to constitute one of the primary divisions of Insecta. It is, however, merely the Neuropterous type, the very essence of the class; and many of its species, Anax Imperator for instance, proclaim themselves by their imperial flight, their enormous size, their richly variegated colours, their despotic and cruel habits, emperors of the insect world. In this group we find the organs of sight, manducation, and locomotion, carried to a greater degree of perfection than we ever meet with, except in similar centres: like the king of birds, the dragonfly is unrivalled among his kind. From Libellula, the centre, we descend at once to Tinodes, or Psyche, on the circumference of the circle. Supposing Psyche to be the approaching genus to Lepidoptera, I think I need not enter very diffusely on the similarities. Passing to the right, we find that Diptera will next touch the central class; in which, after leaving the Phryganeæ, we have now arrived among the next group, or sub-class, Ephemeræ: and here, as we might expect, the inferior wings become much diminished—at the point of contact obsolete.[28] The flight, instead of being solitary, is in company, gracefully and gently rising and falling. The parts of manducation are become obsolete; while, in habit and appearance, the insect imitates the Tipulæ and Chironomi, so exactly that the naturalist is foiled in his endeavours to distinguish between them, as they joyously dance together by myriads in the rays of the setting sun.

We now approach mandibulated orders, and we shall see the loss of mandibles in Phryganea and Ephemera, although apparently resulting naturally enough from their distance from the type Libellula, has yet another cause—the proximity of classes that have no mandibles: in the city-building Ants, the mandibles are very perfect, and, therefore, we may expect them, and we find them in the city-building Termites. The opinion of philosophers, such as the authors of the Introduction to Entomology, is always worth having, although I am doubtful of assertions about insects, when unconfirmed by thorough entomologists; and I believe as yet no entomologist is sufficiently acquainted with the real history of white ants, to decide positively as to their different stages of existence. The following quotation contains also a corroboration of the propriety of this approach:—"The white ants, though they belong to the Neuroptera order, borrow their instinct from the hymenopterous social tribes, and, in conjunction with the ants, (Formica,) connect the two orders. Their societies consist of five descriptions of individuals:—workers, or larvæ; nymphs, or pupæ; neuters, or soldiers; males and females."[29] The class Coleoptera now approaches the Neuroptera, and on each side the boundary we find larvæ digging pitfalls in the sand to catch their prey, and having tubular mandibles to extract its juices when caught. We find them spinning silken cocoons, in which they change into quiescent pupæ, incapable of taking nutriment; which may fairly be supposed a symptom of approach; but there is no insect whose imago I would venture to place on the circumference of the neuropterous circle at the point.

When we find an insect so doubtfully situated between two classes, that Linnæus placed it in Neuroptera, Fabricius in Orthoptera, Latreille, in two of his works, in Orthoptera, and in two others in Neuroptera, MacLeay in Neuroptera, and Kirby and Spence in Orthoptera, I think it but fair to conclude, that the orders must approach very nearly to admit of this difference of opinion: such is Mantispa; and Mantis-like as it really is, it only borrows that appearance from being on the extreme circumference of the Neuropterous circle, and touching the Orthopterous one where Mantis must evidently be situated. Lastly, we see in Psocus the form, wings, and whole appearance of Aphis, so exquisitely imitated, that practised entomologists often, nay mostly, fail in separating them correctly: thus we find that class VII. contains five natural orders, the contents of which have been—and may be again, should the linear and dichotomous system continue in vogue—placed either in the class to which they truly belong, or respectively in classes I. II. III. V. and VI. at the mere option and caprice of the systematist. I have already admitted that I find no neuropterous insect sufficiently related, in its final state to class IV. to warrant my placing it in contact with that class; and that I may not be accused of assuming facts which exist only in my imagination, I am perfectly willing to conclude that no such insect is to be found; a conclusion that time and discovery, by falsifying, can only add yet one more buttress to a tower, which nature seems to point out as built by herself.

There are a few little insects which, like the spiders which crept across Richard's brain, are somewhat perplexing to the naturalist, yet he cannot dispose of them as the monarch did of his spiders; I mean Pulex, Stylops, Thrips, Forficula. But, in truth, the first attempt of the systematist should be to place classes properly, and these disconnected species will, after a time, find appropriate places: they were no more