It remains shortly to notice the arrangements recently proposed by Stephens, Saint Fargeau, Dahlbom, and Hartig; the three last of whom have devoted almost exclusive attention to this order. The arrangement of Mr. Stephens agrees with that given above, except that a third section (inappropriately termed Tubulifera) is formed of the families Chrysididæ, Chalcididæ, Proctotrupidæ, and Cynipidæ, and placed after the bees. In adopting this plan, Mr. Stephens was doubtless influenced by the remarks of Mr. MacLeay upon the supposed affinities between the Chrysidida and some splendid foreign bees, and between the Chalcidida and the order Strepsiptera, which Mr. Stephens places immediately after the Hymenoptera. The removal of the Cynipidæ, Chalcididæ, &c., from the Ichneumonidæ, is clearly unnatural, as well as the juxtaposition of the latter and the ants. The views of M. Saint Fargeau, as developed in various articles in the Encyclopédie Méthodique, Magasin de Zoologie, Annales de la Société Entomol. de France, and more especially in his Hist. Natur. des Hymenoptères, are highly interesting, being the result of a scrupulous examination of "toutes les parties des corps de l' Hyménoptère adulte, dont la forme est l'expression de ses habitudes morales et mème de sa vie sous la forme de larve" (Hist. Nat. Hym. p. 89.). The adoption of these views has necessarily required a minute inquiry into the habits of the various groups, and an equally careful examination of the variations of structure dependent thereupon; and its effects are perceived in the arrangement of the families. The order is divided into two sub-orders, corresponding with the two primary sections given above, but to which the names of Hyménoptères Ovitithers (or Aculeata), and Oviscapters (or Terebrantia) are applied, in allusion to the mode of depositing the eggs. In the former there is said to be no exterior elongation of the oviduct, the eggs passing from the aperture of this organ into an anal cavity*, which opens horizontally, and discharges the egg near the food prepared for the future larva, the female being moreover armed with a sting. In the Oviscapters the eggs are introduced into the interior of various bodies, and it is necessary therefore that the oviduct should Figured by Reaumur (Mém. tom. v. pl. 29. fig. 1.). This figure is at variance with the more elaborate figures of Swammerdam (pl. 18 and 19.); hence, as well as from the slight degree of attention hitherto bestowed upon the mode in which the egg is discharge! in these tribes, I feel inclined to regard Saint Fargeau's definition with distrust. be externally prolonged for this purpose; which external elongation is named the oviscapt. The sub-order, Ovitithers, is divided into Phytiphages (the larvæ of which feed upon vegetable fluids), and Zoophages (the larvæ of which feed upon other insects, larvæ, or spiders). The Phytiphages are divided into either Nidifians (nest-makers), or parasites. The Nidifians are divided into the social and solitary species, and the social species into those whose communities exist several years, and those which are annual. The Phytiphagous, nest-making, social, perennial, Ovitithers, comprise the family of the Ants, and the two genera Apis and Melipona. The genus Bombus, and the family Polistides (including Vespa, Polistes, Epipone, and some new genera of social wasps) are annual. So far only have the details of this system been developed. That this view of the subject is highly interesting; and likely to lead, when fully worked out, to important results in the classification of the order, cannot be doubted. At the same time when we see by this mode of arrangement, insects widely separated, which are most intimately allied in general structure, although varying in the form of those particular organs which are employed in constructing a nestwhen, for instance, Psithyrus and Euglossa are removed far from Bombus, Odynerus from Polistes, &c., we cannot, as it seems to me, but question whether too great an importance has not been bestowed upon the "habitudes morales" of these insects. I have, however, elsewhere entered more fully into this question (Brit. Cycl. vol. ii. p. 874.), and shall only add that, in many cases, as for instance in Psithyrus, Nomada, Ceropales, &c., the parasites, as they are not quite correctly termed, merely deposit their eggs in the already provisioned nests of other insects, and that the progeny of the intruder being first hatched, consume the food stored up for the real inhabitant. There is, therefore, no real difference between the constitution and more important organisms of these miscalled parasites and the species upon which they are parasitic. It is therefore as absurd to place them apart, as it would be to separate the cuckoo, as a distinct primary division, from other birds. Dr. Dahlbom, a most assiduous Swedish Hymenopterologist, has recently published a very interesting sketch of the distribution of this order in his Clavis Novi Hymenopterorum Systematis (Lundæ, 1835, 4to.). He considers the fossorial Hymenoptera as the analogues of the Mammalian Primates, on account of the very imperfect or undeveloped state of the young and the great activity of the predaceous imago, whilst the herbivorous Tenthredinidæ, from their more perfectly organised and active larvæ and sluggish imago, appear to represent the Pecora at the end of the order. He considers the order as containing four principal divisions, which, from their habits, he names Raptatoria, Parasitica, Ædificatoria, and Plantivora, each being connected with the other three by means of transition groups. His more detailed view of the distribution of the families "anatomia externa, metamorphosi moribusque simul consideratis," does not appear to be quite in accordance with the quaternary division mentioned above, being evidently artificially constructed, as appears from the numbers attached to the families. The following is a concise abstract of it : I. Imago with petiolated abdomen. Larvæ apod, subvermiform. A. Imago predaceous, aculeate, solitary, fossorial. erucivorous. Larvæ insectivorous, or Fam. 1. Pompilini, 3. Mellinii, 7. Bembicini, 5. Pemphredonides, 4. Sphegides, 2. Larrates, 6. Crabronides. B. Imago styliferous, tubuliferous or aculeated, solitary. Larvæ erucivorous, or pupivorous. Fam. 8. Ichneumonides, 9. Braconides, 10. Evaniales, 20. Masarides, C. Imago nest-building, living in society. Larvæ omnivorous. D. Imago nest-building, solitary or social. Larvæ mellivorous. Fam. 22. Andrenides, 23. Anthophorini, 24. Apiariæ (sociales). E. Imago living in galls, with the abdomen compressed. Larvæ feeding on galls. Fam. 25. Cynipseæ. II. Imago with sessile and depressed abdomen; oviduct spiral. Larvæ gallivorous?. Fam. 26. Oryssini. III. Imago with sessile and depressed abdomen. Larvæ eruciform, pedate, plantivorous. Fam. 27. Siricides [Urocerida], 28. Tenthredinides. With the exception of the different location and juxtaposition of these primary groups the arrangement here given is not materially unlike that of Latreille. The section B appears to be a most artificial assemblage, sufficient to prove, in connection with the admission of Dahlbom (that certain parasitic species occur in the raptatorial, mellivorous, and gallivorous groups, which nevertheless agree completely in general habits, as, for instance, Psithyrus amongst the bees, &c.) that a distribution of the Hymenoptera, based upon natural habits alone, is as completely artificial as any arrangement resting upon any other single character. Thus the Eumenides and Mutillariæ amongst the Parasitica are strictly raptatorial; and, if the parasitic Psithyrus, &c., be admitted into the Ædificatoria, there is no ground for excluding the Nomadini. In this respect the views of Dahlbom are not so precise as those of Saint Fargeau, which are based, not only upon the natural habits but also upon the precise structure of those individual organs which are employed in performing such habits. Dr. Theodore Hartig, in a memoir published in Wiegmann's Archiv. 1837, No. 2., and in his Die Aderflugler Deutchslands, has proposed an arrangement founded upon the external characters of the imago, and originating in a dichotomy, nearly agreeing with that of Latreille, well characterised by the structure of the trochanters, a peculiarity not previously adopted for this purpose. The Phyllophaga comprises the family of the saw-flies; the Xylophaga that of the Urocerida; the Parasitica, those of the Evaniidæ, Ichneumonidæ, Chalcididæ, Proctotrupidæ and Cynipida?; the Rapientia consists of the Sand-wasps, Wasps, Ants, and Chrysidida; and he Anthophila, the solitary, social, and parasite Bees. The TEREBRANTIA, or first general and aberrant section of the Hymenoptera, is distinguished by having the posterior trochanters 2-jointed, and the abdomen in the females furnished with a lamellate or filiform, auger-like, and generally more or less exserted instrument, employed for the purpose of depositing the eggs in the various bodies destined for their reception. This instrument is connected with glands which do not secrete a highly concentrated poison, although it is evident that, in some species (as the Gall-flies, and some Tenthredinida), the act of oviposition is accompanied by the emission of an irritating and analogous fluid.* The antennæ are very variable in the The Ichneumonidæ, when alarmed, endeavour to use the ovipositor as an organ of defence, and certainly emit a fluid. (Sce E. W. Lewis, on Pimpla stercorator ; and my additional observations, Mag. Nat. Hist. vol. vi. p. 414.) number and form of their joints, both in the various species, and in the sexes of the same species. The females do not lay up a store of food, either of pollen paste, or other insects, for the supply of their progeny. The section Terebrantia comprises the two subsections, a. Spiculifera. form. b. Tubulifera. Abdomen not tubular at the extremity; ovipositor spiculi Abdomen tubular at the extremity; ovipositor sting-like. Messrs. Saint Fargeau and Serville have proposed another mode o. distribution of the Terebrantia, founded upon variations in the form of the ovipositor, proposing five families; namely, 1. Serrifera (Tenthredo), 2. Spirifera (Cynips and Oryssus), 3. Terebellifera (Chalcis, Ichneumon, Evania, and Sirex), 4. Canalifera (Proctotrupes), and 5. Tubulifera (Chrysis). I have not adopted this arrangement for the reasons subsequently detailed. The first sub-section, Phytiphaga, Securifera, Sessiliventres, or Serrifera, as it has been variously named by Latreille and Saint Fargeau, is distinguished by having the abdomen sessile, hiding the base of the posterior legs. The ovipositor in the majority consists of two saws, which are alternately protruded and employed in preparing a place for the reception of the eggs, as well as in conducting them to their destination. In a few species (Terebellifera), the ovipositor is similar in its construction to the ovipositor of the Ichneumonidæ, &c., whilst in one genus (Oryssus), it is spiral. The larvæ feed entirely upon vegetable matters, for the most part upon leaves; a few, however, are internal feeders, and others reside in galls, in the manner of the Cynipidæ. The larvæ are furnished with completely developed organs of manducation; they also for the most part possess six short articulated legs, and a greater or less number of anal prolegs or other appendages. The Phytiphagous subsection consists of two tribes, each composed of a single family, namely, a. Serrifera, TENTHREDINIDE. with a pair of saws. b. Terebellifera, UROCERIDE. with a borer. Abdomen of the female furnished Abdomen of the female furnished |