The situation of antennæ must next be considered. In this respect it seems necessary that they should be so situated as to be under the direction of the eyes: for if you examine ten thousand insects (except, as was before observed[1527], where there are four eyes), you will not find one in which these organs are situated either above or immediately behind them; their station being always either somewhere in the space between the eyes or that below them. In Ptinus F. they are placed near the vertex; but in Gibbium, which is so nearly related to that destructive genus[1528], they are beneath them. In many Melittæ K. they are in the middle of the space between the eyes; and in many other Hymenoptera and Coleoptera (Staphylinus &c.), in the anterior part of it. In many Lamellicorn genera (except in some Acridæ, as A. viridissima) as Melolontha, Cetonia, Lucanus, &c. they may be regarded as planted in the lower surface of the cheek before the eyes; but in Copris &c., in which they are inserted further under the shield of the head, they are properly in the prone surface of the front. In the Capricorn-beetles (Cerambyx L.) and Cnodalon F. they may be termed inocular, or placed in a sinus of the eye; in the former tribe in its interior, and in the latter its anterior side. In the Rhynchophorous or rostrum-bearing beetles (Curculio L.) they vary in their situation. Thus in Macrocephalus Oliv. they are inserted at its apex; in Anthribus in its middle, and in Calandra at its base[1529]. In the water-scorpions (Nepa, Belostoma, &c.) they may be called extraocular, being placed under the head in its prone part, outside the eyes[1530]. In Nirmus Fringillæ, a kind of bird-louse, they appear to be oral, being situated, according to De Geer, under the head near the mouth, at a great distance from the eyes[1531].
In their proportions, both as to length and thickness, antennæ vary extremely. Thus sometimes they are very short—much shorter than the head; as in the aquatic beetles Gyrinus, Parnus, and the water-scorpion; and some land-beetles, as Anthrenus, &c. At other times they far exceed the length of the insect: the males of many Capricorn-beetles are so distinguished. In that of Lamia ædilis they are more than four times as long as the body; and every intermediate length between these two may be found amongst them. They vary also greatly in thickness: in Paussus, whose antennæ emit light in the night[1532], and Cerapterus, they are nearly as thick,—at least their knob, which forms the chief part of them,—as the body of the insect[1533]; while in Mantis, Acrida K. and Psocus, they are as slender as a hair. The antennæ in many of the Prioni, especially in P. imbricornis, are thick from base to tip; while in other Capricorn-beetles they are quite the reverse.
It will not be necessary to enlarge here upon the general form of these organs: I shall therefore only notice the two principal divisions of them in this respect.—Antennæ, regard being had to one of their uses, may be divided into two sections, distinguished by forms extremely different: those, namely, that are employed by insects as tactors to explore their way, and those that cannot be so employed. The great majority are of the former kind; but those that may be denominated setigerous,—as the antennæ of the Libellulina, Ephemerina, of the Homopterous Hemiptera, and of many Diptera, the last joint of which terminates in a bristle, or is furnished with a lateral one, and of some gnats that have short feathered antennæ,—appear not fitted to be used as tactors to explore by touch, and form the latter description. This difference in these organs, as I shall have occasion to prove more at large hereafter, furnishes a strong presumption that their primary function is not touch. Were this the case, it would be common to them all.
As to their structure, antennæ consist in general of a number of tubular joints; each of which having separate motion, the animal is thereby enabled to give them every flexure necessary for its purposes. The scape, or first joint, by means of the bulb inosculates in the torulus, or is suspended to it; and the others, sometimes by a similar, though less pronounced knob at their base, inosculate in the preceding one; but in some cases the inosculation seems not so perfect, the joints being simply suspended by ligament. In pectinated or lamellated antennæ, the branch is usually a lateral process of the joint from which it issues; but in Phengodes (Lampyris plumosa L.) its involute plumose branches appear to articulate with the apex of each joint[1534]. I have a specimen of one of the Cleridæ, of a genus undescribed, in which each branch is forked. In some tribes of the Capricorn-beetles (Stenocorus, &c.) the antennæ are often armed at their apex with spines, sometimes on the upper side and sometimes below. In some aquatic beetles (Gyrinus, Parnus) they are furnished with an auricle at their base, which, like the lid of a box, shuts them in when unemployed, and protects them from the water[1535].
The portions into which antennæ may in general be considered as divided, have been sufficiently explained to you above; but it may not be amiss to add here a few words on the principal variations in their structure that I have had an opportunity of observing. The scapus[1536] or first joint, which includes the bulbus, is usually the most conspicuous joint in the antenna (exclusive, I mean, of the capitulum, in those in which that organ terminates in a knob), it being thicker and often longer than the succeeding ones. In the Capricorn and Darkling beetles, indeed (Cerambyx and Tenebrio L.), the third joint is the longest, but the scape is still the thickest; and in the stag-beetles (Lucanus L.), many of the weevil tribes (Curculio L.), and those of the bees (Apis L.), except in the males, it is as long nearly as the remainder of the antennæ, which forms an angle with it. In shape it is generally somewhat curved and subclavate, or increasing in size from the base to the summit; but it is sometimes straight and filiform, at others oblong or square, at others again triangular, in several instances three-sided: in one (Cetonia cruenta F. Genuchus K.) it is, as it were, broken, the upper part forming nearly a right angle with the lower; in Cerocoma Schæfferi it is foliaceous; and it is occasionally suborbicular: and probably many other forms might be enumerated.
The Pedicellus[1537] is the second, and may be deemed the least conspicuous joint of the antennæ. Though more slender than the scape, it is generally thicker than that which immediately follows it. In broken antennæ it is the hinge or pivot on which the clavola or upper member turns: it is usually very short, campanulate or bell-shaped, or obconical; but in a species of bug (Tetyra, from New Holland—T. pedicellata Kirb. MS.) it is nearly as long as all the rest of the joints taken together. In those species of Lycus, a genus of beetles related to the glow-worm, that have flattened antennæ (as L. reticulatus, fasciatus, &c.), this joint is almost received into the socket of the scape, so that their antennæ appear at first to have only ten joints, but in those which have those organs filiform (as L. minutus, Aurora, &c.) it is more conspicuous.
The Clavola[1538], or remaining joints of the antennæ taken together, constitutes the principal part of the organ, which, especially at its extremity, exercises its functions of touch, or any other sense. The principal variations, as to form and structure, that occur in this part will be mentioned in another place. I shall only here observe, that in many instances the first joint of this part is longer than the rest; but in Tetyra pedicellata just mentioned, it is by far the shortest, and shaped like the pedicel of most insects. In the Libellulina, the Homopterous Hemiptera, and those flies whose antennæ terminate in a bristle, the clavolet is represented by the bristle. But in the flies which have a lateral bristle, on the last joint, and those with triarticulate antennæ that have no bristle, the terminal joint represents it. The clavolet often terminates in a knob, or in several joints thicker than that which precedes them. This varies greatly, not only in its form, but also in the number of joints of which it is composed. Thus in Paussus, Platypus, and many Calandræ, it consists of only a single joint[1539]; in Anthrenus, Ditoma, &c. of two; in Nitidula, Geotrupes, &c. of three[1540]; in Tetratoma, the Silphidæ, of four[1541]; of five in Scaphidium[1542]; of six in one species of Languria, of seven in the common cockchafer (Melolontha vulgaris[1543]); of eight in Diaperis Boleti, in which the whole clavolet forms the club[1544]; of nine in Oenas; and ten in Cerapterus[1545]. All the above, you will observe, are beetles. In the other orders there are eleven joints in the knob of some butterflies; twelve in that of Ascalaphus[1546] and Myrmeleon; and lastly, fourteen in Trachelus[1547].
Under structure also, the number of joints of which antennæ in general consist, should be considered. If you examine the insects belonging to the different orders, you will find remarkable variations in this respect. Let us run through them:—In the Coleoptera the natural number of joints is eleven; but this rule is not without many exceptions. Thus, many have fewer than the prescribed number: Paussus has only two[1548], Claviger and Platypus five, Dorcatoma and Calandra eight[1549], Geniates K. and Phanæus MacLeay nine[1550], and lastly Melolontha ten[1551]. Others, again, have more than eleven joints: Cebrio grandis, Chrysomela stolida, some Saperdæ, and several others, have twelve. In Prionus imbricornis the female has nineteen, and the male twenty[1552]. Rhipicera marginata has thirty-two; and in a New Holland species of this genus I counted thirty-eight. In the Orthoptera I can trace no general law in this respect. In Locusta Leach in some species you may count fourteen joints, in others sixteen, and in others twenty-five. In one, which appears to be a pupa, I found only thirteen. In Mantis they exceed thirty; but in Blatta, from between thirty and forty, they reach nearly one hundred and fifty; often varying in number in different individuals of the same species. The order Hemiptera exhibits two peculiar types of antennæ, which, with some exceptions, distinguish the two natural sections into which M. Latreille has judiciously divided it. In the Heteropterous section they are without a bristle at their end; and in the Homopterous one, with the exception of Aphis, Thrips, &c. they have one. In the genera of both these tribes, the number of joints varies in these organs. Thus, exclusive of the seta, in Flata and Cixius there are only two joints; in Galgulus, Fulgora, and Cercopis, there are three; in Lygæus, Coreus, &c. there are four; in Tetyra, Pentatoma, Tettigonia, there are five[1554]; in Aleyrodes there are six; in Aphis seven; in Thrips eight; in Psylla ten, the last of which is terminated by two bristles[1555]; and in Coccus eleven. The Neuroptera order, as it stands at present, is regulated by no general rule with regard to the number of joints in the antennæ of the insects that compose it. Several types of form in these organs distinguish its discordant tribes. The first is that of the Ephemeræ, in which the antennæ consist of two short joints, crowned by a short, tapering, unjointed bristle. The second is that of the Libellulina, similar to the above, but with a jointed bristle. The third is that of Psocus, in which the antenna has two short thick joints at the base, terminated by a long filiform bristle, consisting of seven or eight joints, and finer than a hair. Perhaps these three may be regarded as belonging to a common type. The fourth type is presented by the short filiform antennæ of Termes; the fifth by the setaceous ones of Corydalis, Hemerobius, &c.; and the sixth and last by the clavate and capitate ones of Myrmeleon and Ascalaphus. In the Lepidoptera and Trichoptera orders the antennæ, though varying in their general form in the three tribes of which Linné formed his genera Papilio, Sphinx, and Phalæna, with the exception of Hepialus, in which the joints are few, are always multiarticulate:—we will therefore, without further delay, proceed to the Hymenoptera. In Latreille's tribe Aculeata the general rule is, that the females shall have twelve joints and the males thirteen. In his Ichneumonides the law seems to be, that the antennæ shall be multiarticulate and setaceous; but in most of the other tribes of the order, even those that in other respects are most nearly related,—as in his Tenthredinetæ,—the number of joints of these organs varies without end. Thus in Hylotoma there are only three joints[1556]; in Cimbex læta[1557] five; in C. axillaris and Perga Leach[1558], six: and so on to twenty-five or more[1559]. The same fluctuation in this respect runs throughout the rest of the order. In the Diptera there are two general types of antennæ:—those of the Tipulariæ Latr., consisting usually of from fourteen to sixteen joints, in the males often resembling beautiful plumes; and those of the remainder of the order, in which they do not exceed three joints[1560]: though the last, or patella, is often further divided into obsolete or indistinct ones[1561]. These antennæ may be further subdivided into filatæ and aristatæ, or those without and those with a bristle, either lateral or terminal.
The clothing of antennæ also merits attention, since it is often not a little remarkable. By clothing I understand the down or hairs of every kind with which they are either generally or partially covered. A great number of filiform and setaceous antennæ of Predaceous beetles (Cicindela L., Carabus L.) have the first two, three, or four joints naked, and the rest covered with a fine down. In insects that have a knob at the end of these organs, whether lamellated or perfoliate, this down is often confined to it, or to its intermediate joints, and seems intermixed with nervous papillæ. These are particularly visible in the flabellate antennæ of Rhipicera, Lampyris Latreillii[1562], Elater flabellicornis[1563], &c. covering both surfaces of the processes of the joints. In some male bees these papillæ are inclosed in hexagonal spaces into which the antennæ are marked out[1564]. It is to be observed, that in many antennæ the joints of the clavolet have one or two bristles or more at their apex, one above perhaps, and one below; the lower angle in those of the serrated antennæ of Elater is usually so furnished, and sometimes the upper. In many Capricorn-beetles and various insects the antennæ are clothed, instead of down, with stiffish hairs or short bristles. Other insects have these organs, at least the clavolet, beset with longer hairs standing out from them on all sides: of this kind are those of a singular beetle (Sarrotrium muticum) sometimes found in this country[1565]. Again, there are some that have only their underside bearded with longer hairs; as Lamia curculionoides, speculifera K., and other Capricorns[1566]. In another of this tribe, Saperda hirsuticornis, the three intermediate joints are ornamented with branches of long black hairs, which give them an elegant and feathery appearance[1567]. In Callichroma alpina the apex of the slate-coloured joints of its antennæ is bearded with black hairs. In Lamia reticulata, and Saperda fasciculata and plumigera, all also Capricorns, a single bunch of hairs, resembling the brush of a bottle-cleaner, signalizes the middle of the antenna[1568]: in Saperda scopulicornis K. this is star-shaped[1569]. Sometimes the scape is externally bearded, as in Trox, a beetle found in horns and bones; and in many other Lamellicorns[1570]. In this last tribe the two exterior leaves of the knob of the antennæ are often set with short bristles[1571]; and in a minute beetle called by De Geer Dermestes atomarius, the hairs of this part are said to form a brush[1572].