should be received with the greatest caution, because the second group in which the character is absent[22] is sure to be too comprehensive. Mr. MacLeay,[23] himself no great friend to dichotomies in general, is completely led away by this particular one. He considers the classes I. II. and VI. of the foregoing table to constitute one grand order, and the classes III. IV. V. and VII. to constitute another; and, after Clairville, he calls the former order, Haustellata, and the latter, Mandibulata. Mr. MacLeay's name is a tower of strength to any theory; and his authority, added to the plausibility of the idea, has really given such a truth-like appearance to this division, that we see it now universally adopted. Let us examine its worth. First, I would ask, Can distinctive characters, thus drawn from part only of the external anatomy of insects, be sound, when to enforce them we are compelled to neglect various other characters which we have been accustomed to consider all important? Scopoli has said, "_Classes et genera naturalia non sola instrumenta cibaria, non solæ antennæ nec solæ alæ constituunt_;" but our dichotomizing entomologists tell us, that neither antennæ, nor wings, nor habit, nor metamorphosis, are to be regarded at all, but "_sola instrumenta cibaria_;" at least, they infer this by separating Orthoptera and Hemiptera, by the intervention of several orders totally unrelated to either of them, a disruption which no nature-loving naturalist could for a moment admit. The truth is, there are seven kinds of mouth in insects, so distinct that good classes could be built on them,[24]--classes which would confirm those which Aristotle appears to have derived chiefly from other characters: of these seven, three are mandibulate, three are haustellate, and one without the rudiments of mandible or haustellum. The three which are mandibulate are somewhat similar, the three which are haustellate bear no more resemblance to each other than that which they all may be said to bear to that haustellated quadruped an elephant; and the tie which holds Haustellata together as a group is about as strong as one formed to bend in a genus Blaps mortisaga, Acrida aptera, Cimex lectularius, and the female of Bombyx antiquus, with the one sole character of being destitute of wings. [22] And, be it observed, Haustellata merely means _not mandibulate_; it does not propose to assert that the contents of the tribe so named need have a particular kind of haustellate mouth, or any mouth at all. [23] Mr. MacLeay has written a little pamphlet on the impropriety of the dichotomous system, which I recollect reading, when published, with considerable pleasure. I forget its title.