Page:Encyclopædia Britannica, Ninth Edition, v. 13.djvu/159

 INSECTS (1) that in which the larva changes into a pupa which, as a rule, is inactive, and which never eats, and (2) that in which there is no true pupa state, the animal continuing active, and eating, from the egg to the perfect insect. In the former the larva changes its skin, or moults, several (often many) times before it changes to the pupa, the last moult (or ecdysis) happening when it is in what may be called the pseudo-pupal condition (which may last from a few hours to several months) immediately before the pupa state is assumed ; this division is usually characterized as &quot;metabolic.&quot; In the latter ecdysis goes on continuously at intervals from the egg to the winged-insect, and the form of the larval condition much resembles that of the perfect, the wings budding out gradually as the creature approaches maturity ; the term &quot; hemimetabolic &quot; is applied to this division. In the first division the conditions are rather more varied than in the second. In the pupa of some Diptera the larval skin hardens, and within this the true pupa is formed (such a pupa is termed &quot; coarctate &quot;) ; in other Diptera the pupa is not contained within the larval skin (which much resembles a true cocoon), but is free and even sometimes active, the various appendages not being connected with the body, as is usually the case in that of a lepidopterous insect. The pupcc of Hymenoptera and Coleoptera are also much in the same condition, but they are not strictly active. Many writers have attempted to draw a broad distinction between such a pupa as that of a moth and that of an ant-lion or caddis-fly, cited here as extremes, because in the latter the members are free, and the pupa is really active shortly before the change into the perfect state, and thus the metamorphosis is supposed to be in some respects intermediate between that of true Metahola and true Hemimetabola. But such distinctions are more apparent than real. In many of the small moths the limbs and other appendages are scarcely consolidated, with the body, but simply concealed iu sheaths of which the ends at any rate are free. Also in the second (or hemimetabolic) division distinction has been drawn between the larva of a May-fly and that of a bug (as instances), because the changes from an absolutely apterous condition to one in which the wings are rudimentary and from this to the perfect state are more marked in the latter. This is probably due to the number of moults being less ; the form with rudimentary wings is in no way a true pupa. The metamorphosis of the internal organs, and even of the mouth parts, is much more marked in the Metalola than in the Hemimetabola. Respiration is maintained by means of spiracles or branchiae, as in larvae. &quot;Hypermetamorphism&quot; is a term applied to certain con ditions in which the larva at one period of its life assumes a very different form and habit from those of another period. Such a condition exists in several Coleoptera, such as Meloe and Cantharis, in which the larva is at first very active, with long legs, slender form, and anal seta, and attaches itself to the bodies of bees, afterwards becoming almost apodal, short, and stout, and living in the bees -nests. Other Cantkaridse live in the egg-tubes of Orthoptera. In Sitaris a still more remarkable intermediate condition has been observed : the larva after having attained its second con dition assumes that of the coarctate pupa of a fly, from which it changes again to a state more analogous to the second condition before finally transforming to a pupa. This kind of metamorphosis has been closely observed by Newport, Fabre, Lichtenstein, Riley, and others. Brauer has recorded a somewhat similar condition in the larva of Mantispa (Neuroptera), which is at first free and very active, and afterwards becomes nearly apodal and obese, and lives parasitically in the nests of spiders. Advanced evolutionists hold the idea that larviv are only acquired conditions. Classification. It is necessary to reduce what may be termed the systematic portion of this article to the smallest possible limits. The various orders are noticed under separate articles, and similar articles are devoted to the con sideration of many of the more prominent, interesting, and familiar insects. All we can do here is to allude briefly to classification as a whole, with indications of the higher groups under each order. We also have nothing to do here with Crustacea, Arachnida, and Myriopuda, now considered as distinct classes, although American writers have recently again included the last two in Insecta as orders, placing the more subordinate groups (or orders in the general acceptation of the term) as suborders. Still more impossible is it for us to enter into an examination of the history of classification ; those of our readers who are specially interested in this subject cannot do better, than consult vol. iv. of Kirby and Spence s Introduction to Entomology, where a most full and painstaking &quot;history of entomology&quot; is to be found up to the date (182G) at which it was published ; or they may consult with equal advantage West wood s Introduction to Modern Classification, and Burmeister s Manual of Entomology (Shuckard s English translation). The different classifications proposed by authors mainly resolve themselves under three headings, the &quot; metamorphotic &quot; (of which Swammerdam may be considered the founder), the &quot;alary&quot; (or wing-system, due to Linnaeus), and the &quot; cibarian&quot; (or mouth-system, originating with, or at any rate elaborated by, Fabricius). The meta morphotic system divides insects into those that undergo complete and incomplete metamorphoses ; the alary is based upon the presence of two or four wings, or their absence altogether ; the cibarian depended upon the conditions of the mouth organs, and more especially as to their being fitted for biting or sucking (mandibulate or haustellate). But experience proved that each of these systems had its defects ; there were always some groups, of more or less importance and extent, that would never fit satisfactorily into any of the proposed systems. To remedy this varying means were adopted, such as a combination of the several systems into what has been termed the &quot;eclectic&quot; system, the erection of numerous orders for certain aberrant groups, and that most ingenious idea of MacLeay, the author of what is termed the &quot;circular&quot; system. We are disposed to consider that of all systems the one that combines the greatest amount of convenience with the nearest approach to being natural is the metamorphotic, and this we shall accordingly follow here. It is not intended to acknowledge the subsidiary orders, excepting the Collembola and Thysanura, which are probably scarcely true insects, but which it is necessary to place here, were it only to avoid the risk of their being overlooked alto gether, inasmuch as the writers on the other classes of Arthropoda are not likely to recognize them as coming within their scope. The stumbling-block of all systems has been the Linnrcan order Neuroptera, inasmuch as its members combine the characters of most of the other orders, and ingenious American writers have attempted to overcome this difficulty by considering it a collection of &quot;synthetic types.&quot; In adopting metamorphosis as the basis of classification, we prefer to take another course, and to follow Erichson, who (in 1839) boldly transferred all those Neuroptera with in complete metamorphoses to the Ortkoptcra as a suborder, although, in dealing with the Neuroptera in the light of a specialist, division into several orders appears the more natural course. The sequence of orders we propose to follow is as under :