same fact. Hence, the various forms observable in the structure of the mouth of these insects, notwithstanding the uniformity of the nature of their food, are dependent upon the form of the flowers from which the nectar is procured; the means by which it is prepared and transported either for food or for store for the future young; and the various modes in which the different parts of the mouth are employed in the construction of the nests. For instance, if a flower be long and tubular, it is necessary that the tongue should be elongated and slender; whereas in those species which collect the honey from shallow or flat flowers, the tongue is shortened. In those species, again, which collect honey for store for their young, the tongue and its appendages are of large size; whereas in those which need only a very small supply of honey for their own support, it is short and narrow. The form of the mandibles, again, is entirely dependent upon the mode of construction of the nest, serving as trowels, rasps, &c. The order is also anomalous, from comprising species which are composed of three kinds of individuals, namely, males, females, and neuters. The latter occur in no other order of insects except in the single Neuropterous family Termitidæ ; and as they exist only in the social species, where a great share of the labour of the community is cast upon a certain portion of the inhabitants, it is necessary for the due performance of the labour by the latter, that they should not interfere with the duties of the individuals occupied in continuing the race; and their sexual organs and instincts are accordingly rendered singularly abortive. In all other respects they are females. These peculiarities will however more properly obtain notice under the respective families. The duration of the existence of these insects never exceeds one year, there being only one generation during that period. This order is of very considerable extent, being apparently inferior only to the Coleoptera. Messrs. Kirby and Spence have estimated the number of its species as averaging about one fourth of the insect population. In this country we probably possess 3000 species, of which two thirds are of minute size. These insects, judging at least from the number of large species sent home by travellers, appear to be far more abundant in tropical climates than in our country. The species seldom occur of a very large size, very few attaining or exceeding two inches in length, or three in the expansion of the wings. Messrs. Kirby and Spence seem to consider that, in respect to the mode of taking their food, the Hymenoptera can belong to neither of 66 the great divisions Mandibulata and Haustellata, but that they really lap their food, and might thence be termed lappers, their mandibles being employed in their economy. Dumeril however, (Considérat. Général., p. 9.), gives them "la double faculté" of masticating and sucking the food, considering the lower organs of the mouth as forming une sorte de tube et de langue." Hence Lamarck makes the Hymenoptera the connecting order between the Mandibulata and Haustellata. Latreille, attaching greater importance to the organs of flight than to the mouth, has placed the Hymenoptera between the Neuroptera and Lepidoptera, regarding Phryganea and Termes as forming the passage between the two former orders and the long-tongued bees as approaching the Lepidoptera. (Considérat. Général., p. 73. 76.). Another circumstance, confirming the relationship with the last-named order, occurs in the resemblance between the larvæ of the Saw-flies and the caterpillars of the Lepidoptera. Mr. MacLeay, on the other hand, places the Hymenoptera between the Coleoptera (with which they are supposed to be connected by the osculant order Strepsiptera), and the Trichoptera, the Tenthredinidæ being considered as Trichopterous, and the Uroceridæ as forming an osculant order, Bomboptera, between the Trichoptera and Hymenoptera, which last order is thus reduced to the species possessing apodal larvæ; thus, by means of the connection between the Ants (Formicida), and White Ants (Termitida); and the Caddice-flies (Phryganeidae), and Saw-flies (Tenthredinida), a strong relation is shown. to exist between the Linnæan orders Hymenoptera and Neuroptera. It seems to be admitted on all hands that the insects, which are the real analogues of the present order, exist in the Dipterous order, almost every Hymenopterous group having its representative in the latter. Mr. MacLeay has also noticed the apodal structure of the larvæ as analogous in both orders, adding also the incomplete or coarctate nature of the metamorphosis: no Hymenopterous insect, however, undergoes the latter kind of transformation. The order was established by Linnæus under the name of Gymnoptera (naked wings), in the 4th edition of the Systema Naturæ (1744); it had previously, by the old naturalists, been united with the Neuroptera. In this edition, as well as in the 1st edition of his Fauna Suecica (in which work the name of the order was changed to Hymenoptera), it was composed of only four genera, Tenthredo, Ichneumon, Apis, and Formica; but in the subsequent writings of the il * lustrious Swede, the genera Cynips and Urocerus were separated from Tenthredo; Sphex, Chrysis, and Vespa from Apis, and Mutilla from Formica. In his last edition of the Systema Naturæ, 314 Hymenopterous species were described; but in Gmelin's 13th edition of this work, their numbers were increased to 1241. Other genera were added by Geoffroy, Fabricius, Panzer, &c., but the first attempt to distribute these generic groups into primary sections was successfully made by Latreille, who divided the order into, 1st. The Terebrantia ; and, 2d. The Aculeata; the former including the Saw-flies (Securifera), and the Cuckoo-flies (Pupivora); and the latter comprising the Ants (Heterogyna), Sandwasps (Fossores), Wasps (Diploptera), and Bees (Mellifera). In the Terebrantia the construction of the abdomen and ovipositor and the antennæ offer the most satisfactory characters for classification, but which become almost useless from their uniform structure amongst the Aculeata; resource is therefore had, in the latter section, to the wings, the habits of the insects, and the nature of the food of the larvæ, &c. Mr. MacLeay, excluding Tenthredo and Urocerus from the order, divides it into five groups: 1. Anthophila (Bees), connected by the Wasps (the folding of the wings of which does not appear to him to be a character of so much importance as it possesses in Latreille's system), with 2. Rapacia (Sand-wasps); 3. Pupivora (Cuckoo-flies, &c.); 4. Heterogyna (Ants); and 5. Tubulifera (Ruby-tails). The removal of the Ants (and Mutilla), from their intimate allies the Anthophila and Rapacia, to a situation between the parasitic Pupivora and Tubulifera, cannot be maintained. In like manner, and for the reasons given by Messrs. Kirby and Spence (Introduct., vol. iv., p. 374.), and upon other considerations which I have detailed in my volume upon Insects in the Cabinet Cyclopædia of Dr. Lardner, I do not adopt Mr. MacLeay's removal of the Tenthredinidæ and Urocerida; and therefore propose the following arrangement, founded chiefly upon the views of Latreille: Sect. I. TEREBRANTIA, Latreille, (Hymenoptera Aberrantia), exhibiting very variable structure, but having the abdomen of the females furnished with an instrument employed as a saw or borer for depositing the eggs. Antennæ various. Sub-section 1. PHYTIPHAGA, SESSILIVENTRES, SECURIFERA, or It may be noticed that Linnæus thought of dividing the Hymenoptera into two sections, "secundum aculeum punctorium vel mitem." (Philos. Entomol., 4to., p. 29.) SERRIFERA, having the abdomen sessile, the larvæ feeding upon vegetable matter, with a well developed mandibulated mouth. Fam. 1. Tenthredinida. 2. Urocerida. Sub-section 2. ENTOMOPHAGA (Pupivora Latr.), having the abdomen attached to the thorax by a portion only of its transverse diameter: larvæ with slightly developed mandibulated trophi, for the most part feeding parasitically upon other living insects. 5. Division 1. SPICULIFERA, abdomen with an elongate plurivalve Sect. II. ACULEATA, Latreille (Hymenoptera Normalia, or the typical portion of the order), the abdomen of the females (and neuters), armed with a sting connected with a poison reservoir. Antennæ of the males, 13-; females, 12-jointed. Sub-section 1. PRÆDONES, Latr. (Heterogyna, Fossores, and Diploptera, Latr.) having the basal joint of the posterior tarsi cylindrical, not dilated, nor formed for collecting pollen: larvæ feeding upon other insects stored up, or upon animal or vegetable fluids provided by neuters. Fam. 9. Crabronidæ, 10. Larrida, 11. Bembecida, 12. Sphegidæ, 13. Scoliida, 14. Mutillidæ 15. Formicidæ, 16. Vespida. Sub-section 2. MELLIFERA, Latr., having the basal joint of the posterior tarsi dilated and pollinigerous. Larvæ feeding upon honey or pollen paste, deposited by the parent, or collected by neuters. Fam. 17. Andrenidæ, 18. Apida. A succession of affinities appears to exist amongst these families. The bees, which are the most perfectly organised and typical insects of the order, lead to the Vespida by means of such short-tongued bees as Hylæus, &c.; whilst the transition from the wasps by the solitary species (Odynerus, &c.) to the Crabronidæ, Sphegidæ, Bembecidæ, Scoliidæ, and Mutillidæ, is almost unbroken. The ants are of difficult location. Their introduction immediately preceding the wasps, according to the views of Latreille (Genera Crustaceorum, &c., vol. iii. p. 220.), interrupts the series founded upon the great resemblance between such genera as Philanthus and Vespa (Latr. Consid. Générales, p. 76.), and upon the fossorial habits of certain Vespida; whilst, by placing the ants at the head of the Prædones, as in the later works of the same author, they are far removed from the social bees and wasps with which they agree in their entire economy. I have adopted the former situation for this family, although I cannot but think it would violate fewer relationships were the ants to terminate the order; their supposed affinity with the Mutillide would, indeed, in this view be broken; but the relation between these two families is in reality very slight. By this means the series above-mentioned would be maintained, and the connection existing between the aberrant Chrysidida and certain Fossores, as evinced by their habits, maintained. In the construction of the ovipositor, as well as in the nervures of the wings, some Chrysididæ nearly approach the Proctotrupidæ, especially certain splendid exotic species which I have seen in the Royal Museum of Berlin, allied to Bethyllus *; the genus Stephanus seems to connect the Ichneumonidæ and Evaniidæ, which latter are regarded by Latreille as nearly related to the Urocerida. The Cynipidæ appear to me however, in their gall-forming and consequent herbivorous habits, to approach nearer to the Tenthredinidæ, some of which are similarly gall-formers. In the structure of the ovipositor they also approach Oryssus amongst the Urocerida, whilst in the parasitic habits of several recently observed species (Allotria victrix, Westw., &c.) they are closely connected with some of the minute Adscitous Ichneumons, especially Aphidius, &c. The Uroceridæ, in the structure both of the ovipositor and larvæ, admirably intervene between the Entomophaga and Tenthredinidæ ; which last are certainly farther removed from the types of the order than any other of the Hymenoptera. † The genus Bethyllus seems to be nearly allied to certain Tiphiæ. Mr. Haliday even seems to regard it as aculeate, and as closely connected with Stigmus; from his account of its habits it also appears to be fossorial. By reversing the groups composing the section Aculeata, as arranged above, the bees, which are certainly the types of the order, would be removed to the greatest distance from the adjacent orders of Trichoptera (with which the saw-flies from the connecting link) and Coleoptera (with which some species of ants, according to Mr. Mac Leay, form the connection). By this means, however, the circular succession would be prevented, the ants and saw-flies having no bond of union. |