1. Family BALANIDÆ, (or Sessile Cirripedes). (continued 3)

On the internal viscera I have nothing to add. The cement-duct is represented in Pl. 30, fig. 2, t′, on the near side, running into the antennæ; and I repeatedly traced it, for the duct is very strong, as far as the disc segment; at the other end it joins the cement-gland (t) on the same side of the body. This cement-gland is proved, by the clearest series of facts, to be converted into the incipient ovaria and ovarian cæca. The cement-glands in the older pupæ could be traced as far as the cæca of the stomach, exactly where the ovaria lie in the mature animal; but in some young pupæ, they extended further posteriorly, past the mouth, between the outer and inner membranes of the overlapping carapace. I have faintly shown the course of the stomach, with its two cæca at the upper end; the anus lies between the caudal appendages, at the extremity (above b′) of the abdomen. At this age there is no trace of the vesiculæ seminales, so conspicuous in the mature Cirripede.

Young Cirripede, whilst within the pupa.—I repeatedly succeeded in dissecting the young Lepas australis out of the pupa; and by the previous action of boiling potash, and by a strong light, I was enabled to make a camera sketch (Pl. 30, fig. 2) of the relative positions of their several parts. The young Cirripede is drawn very faintly, and is best seen by holding the plate in the same position with the mature animal, of which a section is given in my volume on the Lepadidæ, Pl. 9, fig. 4. I may just notice how complicated are the membranes in a longitudinal section taken at this period: we have, 1st, beginning at the back, the horny tissue of the carapace or bivalve shell of the pupa; 2d, the primordial valve (z, in fig. 3) of the young Cirripede; 3d and 4th, two folds of corium; 5th, the membrane of the sack of the Cirripede; 6th, the membrane of the sack of the pupa; 7th, the outer tunic of the thorax of the pupa; 8th, the outer tunic of the thorax of the young Cirripede; 9th, the corium lining the latter membrane; and these nine membranes would be repeated on the opposite side of the section, if it were taken through either side of the shell or carapace, bordering the orifice.

After the exuviation of the outer membranes of the pupa, certain pre-existing coloured marks in the corium, such as those round the eyes and round the acoustic orifices, along the ridge of the back and on the borders of the orifice, &c., are still retained by the young Cirripede, either temporarily or permanently; so that the correspondence of part with part of the external surface admits of no doubt. Moreover, the three terminal segments of the antennæ are invariably retained by the young Cirripede, though in a functionless condition, and into them the outer membrane of the body, and an important organ, viz., the cement-ducts are still prolonged; hence these prolongations must be considered as aborted antennæ. Again, we have seen that the mouth of the young Cirripede is formed under the rudimentary mouth of the pupa, with the new œsophagus, round the old œsophagus, leading into the same alimentary canal. The twenty-four extreme tips, likewise, of the six pairs of biramous cirri of the Cirripede are formed within the twenty-four extremities of the six pairs of biramous, natatory legs of the pupa. Consequently, in the Cirripede and pupa, thus far, part corresponds with part, notwithstanding that new eyes are formed posteriorly to the old eyes, and new acoustic organs in a quite different position from the old ones; but now we come to a most important diversity in the metamorphosis, or rather to follow Professor Owen,[66] in the metagenesis, of the young Cirripede. Although, as just stated, the extremities of the cirri are formed within the legs of the pupa, yet, from the great length of the cirri, they occupy more than the whole of the thorax of the pupa; so that the thorax of the young Cirripede is not formed within the pre-existing thorax of the pupa, but within that part of the pupa, (homologically a portion of the first three cephalic segments), which lies anteriorly to the thorax. As a consequence of this, the pedicels and lower portions of the cirri, the segments of the thorax and its dorsal surface, all come to occupy a position at nearly right angles to that of the corresponding parts in the pupa: this is shown in Pl. 30, fig. 2. And as a further consequence, (and this is the more important point), the sack, which both in the young Cirripede and pupa is formed by the overhanging and produced portion of the carapace, and which is internally lined by a reduplication of the membrane of the thorax, is necessarily, owing to the changed position of the thorax, altered in extent and carried much further; namely, from extending merely parallel to the longitudinal axis of the pupa (from b to b′), it is now in the young Cirripede, in addition, carried (to s′) almost quite across the inside of the animal. Hence it comes that the young Cirripede is, as I have said in my former volume, internally almost intersected; and its body remains attached only by a small space, (see the broken line, round a and b in Pl. 25, fig. 1, of a Balanus with the shell, &c., removed from one side), to the sternal or ventral, inner surface of the carapace,—the carapace being modified either into the capitulum and peduncle, or into the shell with its operculum and basis. As a still further consequence of this change of position of the body of the young Cirripede within the body of the pupa, the alimentary canal becomes shortened to fully half its former length. At the same time, the interspace between the mouth and first pair of legs of the pupa, (consisting of the seventh and eighth segments of the archetype), is quite lost in the Cirripede by coalescence. The final cause of the thorax of the young Cirripede not being developed within the thorax of the pupa, probably is, that the cirri may be formed of considerable length, so as to be immediately enabled to seize prey; and that the thorax, which supports the cirri (and this probably is even more important) should be as free as possible within the sack, so as to aid the captorial action of the cirri.

After these remarks, more especially with regard to the formation of the sack, if any one will look at the sectional drawing of a pedunculated Cirripede in my former volume, or of a sessile Cirripede (Pl. 25, fig. 1) in this present volume, in which latter the shell adds to the complexity, he will perceive the cause of the extreme difficulty in understanding the relative position of the parts throughout the whole class. Even after I had discovered that the prehensile antennæ of the pupa might always be found in the centre of the basis or surface of attachment, and which fact, it might have been thought, should have convinced me that this was the anterior end of the whole animal, yet still I fancied that the prominent mouth represented the entire head, and that the shell was something quite distinct. It is clear that others have been equally perplexed; for that which is the anterior end in the eyes of one naturalist, is the posterior end in the eyes of another; so with the dorsal and ventral surfaces: one naturalist considers the peduncle of the Lepas as the abdomen; another considers it as a pair of metamorphosed, thoracic limbs, &c.! The probable position of the segments of the body of a mature Cirripede, in relation to the three anterior cephalic segments, or carapace, is shown in the diagram (Pl. 25, fig. 6) of the supposed position of the mature Proteolepas within its pupal envelopes. Here, in the diagram, the two segments immediately succeeding the mouth (c), which are the seventh and eighth of the archetype, (for the mouth consists of three segments, and all in front of the mouth of three other segments), have come to adhere by their dorsal surfaces to the internal surface of the carapace,—that is, of the first three segments, which ought of course to have stood quite in advance of these two segments, and these two segments again ought to have stood in advance of the mouth. The mouth is directed posteriorly, instead of from the body; and the three segments of which it is formed (closed at their anterior end by the labrum), and are very small compared to the relatively monstrously great, three anterior cephalic segments, composing the carapace. To place the segments of the body of Proteolepas in proper sequence, in respect to those of the carapace, and in accordance with the sequence of the archetype Crustacean, it would be necessary, by seizing the extremity of the abdomen (a), to tear the two segments succeeding the mouth from their dorsal attachment, as far back as the basal margin of the labrum; and then pull them till they stood posteriorly to (or in the diagram, above) the mouth; which latter part would, by the same movement, be made to project out at right angles to the ventral surface, and would then be preceded only by the first three, great, confluent segments of the head, which being produced backwards, form the carapace. All that has just been said on the position, in Proteolepas, of the segments of the body in relation to those forming the carapace, I believe to be applicable to all ordinary Cirripedes, with this difference, that in the latter, after the metamorphosis, the two segments succeeding the mouth quite disappear on the ventral surface, and dorsally are either aborted or have coalesced with the adjoining segments.

Act of Metamorphosis.

When the due time for the act of metamorphosis has arrived, the pupal carapace splits along the dorsal ridge, and is cast off, together with the acoustic sacks, the basal segments of the two antennæ, and the great, black, compound eyes, hanging to the UU-like apodemes. The three terminal segments of the antennæ invariably remain cemented to the surface of attachment. The exuviæ usually continue for a time united to the cemented antennæ, but are finally washed away. Besides the split along the dorsal ridge, the carapace separates, all round the orifice, from the delicate tunic lining the sack and investing the thorax and natatory legs of the pupa; for these membranes are not moulted for some considerable time afterwards. Hence all these inner parts retain for a period the appearance and structure of the natatory pupa, whilst the exterior resembles, in every respect, a fixed and perfect Cirripede.

In my former volume, I have insisted on the important and curious results which ensue from the eye-apodemes penetrating so deeply into the body (see Pl. 30, fig. 7, in which the proportions are more correct than in fig. 2), with the eyes attached exteriorly to their outer arms; for as these apodemes have to be ejected, the external membrane of the young Cirripede (Pl. 30, fig. 2) has to be formed in a deep fold or arch over them, and consequently the membrane on the sternal surface is formed considerably longer than on the dorsal surface. From this it follows, when all the membranes are free and are stretched fully out after the moult, that the whole animal, posteriorly to the cemented-down surface, turns vertically up, and assumes its normal position at right angles to the surface of attachment, and to that which it held in its pupal condition; for the pupa always adheres with its sternal surface parallel to the surface of attachment. A young Lepas, which has just moulted its pupal carapace, and has assumed its proper vertical position, with the cemented antennæ and the surface of attachment remaining as before, is shown at fig. 3, but is drawn on a smaller scale than the pupa fig. 2, out of which it may be supposed to have been excluded. In this fig. 3, it may be observed that the natatory legs and caudal appendages of the pupa have not as yet been moulted. The fact of the stretching out, in the young Cirripede, of the fold of membrane, which in pupa, just before the metamorphosis passes over the apodemes and eyes, is well shown by three darkly-coloured bands in the corium, which in the pupa are curled, but after the moult, are stretched straight out on the peduncle of the young Lepas.

The pupa, and consequently the young Cirripede, from being attached at first by the antennæ, does not adhere by the actual anterior extremity, but by the sternal surface near it; the anterior extremity, however, soon becomes cemented down, and afterwards, in ordinary cases, ceases to grow. In Cryptophialus, however, and in certain genera of the Lepadidæ, as Alcippe, Lithotrya, and Anelasma, the anterior or basal extremity does continue to grow, and is not cemented down, and therefore comes to be prolonged beyond the original point of attachment; in order to allow of this, the surface to which the Cirripede is attached has to yield, apparently simply to the pressure exerted in the case of Anelasma, but in the three other genera, to the rasping action of the roughened surface of the extremity of the peduncle.

When after a period the pupal membranes of the sack, thorax, and natatory legs are moulted, the cirri of the young Cirripede are curled up, and its thorax is raised towards the orifice, and we have the animal in its ordinary position, and perfect with the exception of a few parts to be further developed or modified. For, instead of calcareous valves, we have at this period only the so-called primordial valves, composed of chitine; and in the case of Lepas australis, some minute spines and some coloured marks on the peduncle, which soon disappear. The muscles, which enter the three terminal segments of the antennæ in the pupa, have to be absorbed and converted into ligamentous threads. In Lepas, the labrum has to become bullate; and the cæca have to increase in number round the upper end of the stomach, and their dark colour and that of the whole alimentary canal has to disappear or be much weakened. The filamentary appendages at the bases of the cirri, which generally contain some of the testes, have to be developed. The probosciformed penis, which at first equals only the pedicels of the posterior cirri in length, and is apparently imperforate, has to increase greatly in length. The testes and vesiculæ seminales have to be formed. And lastly, and this is a more important point, the two gut-formed cement-glands (or incipient ovaria, t, fig. 2, Pl. 30) which, at the period of the moulting of the carapace and eye-apodemes, and when the whole animal was upturned, came to occupy, together with the cement-ducts (t′), their normal position, i. e. nearly parallel to the sternal surface, now undergo further changes. Their upper and posterior ends lying near the cæca of the stomach, increase in size, but retain nearly the same character, and thus form the two true ovaria; their middle parts become emptied of their cellular contents, and are converted into the two simple ovarian tubes; and their lower ends branch out, inosculate, and form the inextricable mass of ovarian tubes and cæca. The points of junction on each side between the two cement-ducts and the newly branched ovarian tubes, become now developed into the two cement-glands. The cement-ducts, which continue throughout life growing, either still enter the old antennæ and there pour out the cement-tissue, or they pour it out through special orifices formed for this purpose in the lower part of the peduncle. The changes, supervening during the metamorphosis, in the ovaria and in the cementing apparatus, here described, I have no doubt are general throughout the Order.

I have above alluded to the primordial valves; these are beautiful objects when viewed under a high power: they are composed of chitine without a trace of calcareous matter, but prefigure in shape, size, and direction of growth, the shelly valves soon to be formed under and round them. They are composed of an outer membrane, with its margins separated by yellow thickened rims from the membrane uniting the several primordial valves together; and this outer membrane is underlaid by a single layer of generally hexagonal, thickish cells (Pl. 30, fig. 3 a), varying from 1 to 2/6000th of an inch in diameter. These cells seem to contain a nucleus; and they are at first separated from each other by clear interspaces. If a specimen be taken, only a little before the formation of the calcareous valves, one or more layers of membrane, marked by an hexagonal reticulation, can be separated from the lower surface of the main hexagonal network. It is a singular fact, that in those genera in which there are several valves, the primordial valves occur only on five, namely, on the two scuta, two terga, and the carina; and these are the most persistent valves in the several genera. The other valves are prefigured only by brownish membrane, without the hexagonal tissue. In the mature Lepas, the membrane connecting the several shelly valves is not moulted, but disintegrates; in the primordial valves, however, which stand far separate from each other, this membrane is moulted; and immediately after the first moult, the first layer of shell appears under and a little way beyond each primordial valve; shelly matter likewise appears, at least in some cases, between the cells of the hexagonal tissue. The young shelly valves are connected together, at each successive moult, by narrower strips of membrane, till, in the case of Lepas, the valves when mature come to touch each other (Lepadidæ, Pl. 1, fig. 5). The primordial valves are often preserved for a long time on the umbones, or centres of growth of the five valves, on which they occur, in the same manner as the larva-shell is sometimes preserved on the apex of certain spiral molluscs. Had not Cirripedes gone through so many and such complicated metamorphoses, this last state, when furnished only with primordial valves and with several internal organs only partially or not at all developed, would have deserved to have ranked as a special stage, and not as merely subordinate to the last or pupal condition.

In the Balanidæ, or sessile Cirripedes, the young animal, immediately after the metamorphosis, or still better if dissected out of the pupal carapace, as I succeeded in doing with Balanus balanoides, may be said to be pedunculated; for it is attached by a little disc of cement closely surrounding the antennæ, the rest of the membranous basis forming an almost semi-globular, flexible peduncle. The valves, at this the earliest period, are all membranous, and do not overlap each other. In the Balaninæ they do not present the peculiar structure of the primordial valves of the Lepadidæ; but in the Chthamalinæ, in Chthamalus, I saw traces of this structure. Calcareous valves are soon formed under the membranous valves. The opercular valves, at this early period, are much larger than the valves or compartments of the shell, which are only four in number, for the carino-lateral compartments are not yet formed. The compartments from the first are surprisingly strong, and have their alæ already formed and overlapped by the adjoining compartments; but of the radii there is as yet no trace. The four compartments form a narrow but nearly circular hoop, which, from its relatively large diameter, tends to draw down the upper or posterior end of the animal, now forming the opercular valves; and as the basis soon becomes throughout cemented to the surface of attachment, the young Cirripede is much depressed. Soon the opercular valves are drawn a little way down within the shell, becoming attached to the sheath, instead of, as at first, to the very summits of the compartments. In regard to the changes which take place in the shell, in the number of the segments in the cirri, and in the number of spines borne on these segments, &c., during the continued growth of the animal, as they are chiefly important for the identification of the species, I will here refer to a discussion on this subject under the head of the Genus Balanus.

On the Homologies of the Carapace and Shelly Valves.

In the pupa, the carapace is produced, not only posteriorly, but anteriorly, so as to cover the entire animal, with the exception of a narrow sternal surface (Pl. 30, fig. 4): in front it is notched, where the sternal surface terminates, and from this notch a faint line runs along the dorsal surface, separating its tergal elements. In the young Cirripede, after the metamorphosis, there is no trace of this medial dorsal suture, or of the wider sternal surface. Looking at the several genera of the Lepadidæ, the external covering of the whole peduncle and capitulum is so continuous and of so uniform a nature, that I think we must consider the whole as a carapace, of which the sternal borders have become completely confluent; formerly I was inclined to look at the capitulum alone as formed by the carapace, and at the peduncle as being composed of the two or three anterior cephalic segments, cased only by their own integuments. As far as can be discerned, the carapace in the pupa, and consequently in the Cirripede, consists only of the tergal elements of the segments; and this seems likewise to be the case with the carapace of the Podophthalmia. Until lately,[67] Prof. Milne Edwards doubted whether the carapace in the higher Crustaceans (to which I believe the carapace of Cirripedes must be compared) was formed by the backward production of the third segment, which bears the second pair of antennæ, or of the fourth, i. e. the mandibular segment; but from the distribution of the nerves, he now argues that it must mainly belong to the third segment. In Cirripedes, the course of the nerves leads to the same conclusion; for the whole shell, sack, and peduncle are supplied with nerves proceeding from the compounded ganglion, which belongs to the second and third cephalic segments.[68]

The whole of the head in front of the mouth, together with the carapace, is, as we know, formed of three segments; and each of these segments, homologically, ought to consist of eight elements; I recall to mind these facts, inasmuch as the transverse separation between the peduncle and capitulum in the Lepadidæ, and between the basis, the shell, and the opercular valves in the Balanidæ, might be thought to be connected with the separation of the three cephalic segments. So again, as in the Balanidæ the shell normally consists of eight compartments, these might be thought to be related to the eight elements of one or other of the three segments. But I see no reason for admitting this view; and in the case of the carina and rostrum, two of the most persistent and important of the compartments, they exactly cover the sutures which ought to separate the two tergal and two sternal elements of the segment. The valves, moreover, often form many more whorls than three, or the number of the true cephalic segments in front of the mouth; and in each whorl the valves tend to stand in tile-like or alternate order, with respect to those in the whorls both above and below, which would not be the case, if they were the eight elements of the segments.

For the true homologies of the sclerodermic plates, or Shelly valves, with which the external covering of Cirripedes is so generally strengthened, we must, I believe, look to the carapace of the Podophthalmia. In these latter, we find the carapace composed of sclerodermic plates, which, though closely joined and only occasionally separated by sutures, yet in their origin are distinct;[69] they tend, also, to be arranged in alternate or tile-like order. As the animal grows, the old sclerodermic plates, all joined together, are moulted, and new ones, also all joined together, of a larger size, are formed beneath. Now let us imagine the growth to be more gradual but yet periodical, and the new and larger sclerodermic plates, when formed under the old ones, to adhere firmly to them; the older plates would thus be prevented from becoming confluent, and instead of being all moulted together, as is now the case, they would be almost continually separated from each other, owing to the almost continuous increase in size of the new underlying plates. Consequently lines of splitting would run between the several plates, however numerous they might be, instead of there being, as now, a single line of splitting extending down the back. In fact, we should have the identical manner of growth of the shell or carapace, which occurs in Cirripedes. It is on this ground, and from the several points of homological resemblance incidentally mentioned in the last few paragraphs, that, in the early part of this Introduction (p. 13), when discussing the whole class, I stated that I believed that the carapace of Cirripedes presented more real resemblance with the carapaces of the Podophthalmia, or higher Crustacea, than with those of the lower Crustacea, though in mere shape they more nearly resembled the latter.

Cementing Apparatus. (Plate 28.)

I have already (p. 128) given an account of the manner in which, in the pupa of Lepas, the cement-tissue escapes from the prehensile antennæ, and of the structure of the cement-ducts, and of the cement-glands or incipient ovaria; and likewise of the changes by which these organs assume their ultimate form in the mature Cirripede. In my former volume, on the Lepadidæ, I described the cement-glands and the cement-tissue in several genera, and I have there shown (singular as the fact is) that the two cement-glands, with their contents, actually consist of ovarian tubes with their contents (for there seemed to be a relation in the state of fulness in both) in a modified condition. In the Balanidæ, I am not able, from the difficulty of the dissection, to confirm these conclusions, excepting in so far that the tubes on which the cement-glands are formed, run into the mass of ovarian cæca; but, I may add, that in the abnormal Proteolepas, belonging to another Order (see the section, Pl. 24, fig. 1), nothing could be plainer than that the membrane of the ovarian sack (b) formed the cement-ducts, and that their cellular contents, which within the sack (a) were in process of conversion into ova, within the ducts were converted into the cement-tissue. This cement, by some unknown power, travels down the ducts, and debouches at the antennæ.

In the Lepadidæ, there are only two cement-glands, which are situated high up in the midst of the ovarian cæca, with one cement-duct proceeding from each: both the glands and ducts increase in size with the age of the animal:[70] the cement issues either permanently from the prehensile antenna, or, after a short period, through apertures in the peduncle, arranged irregularly or in straight lines,—the last formed apertures being furthest from the central and basal point of the peduncle. In the Balaninæ, on the other hand, at each period of growth, a pair of new cement-glands is developed, larger than those last formed, and making, with the older glands, a chain, connected together by what I have called the cement-trunk. The cement-trunk consists of a tube, which generally becomes enlarged just before entering each gland. The glands, the cement-trunk, and cement-ducts, all adhere to the basal membrane or basal shelly plate. Each gland gives rise to two cement-ducts, these often bifurcate, and sometimes repeatedly bifurcate and inosculate before pouring out their contents round the circumference of the basis; and sometimes they all first enter a circumferential cement-duct. The probable cause of the greater complexity of the cementing apparatus and of the greater number of the excretory orifices in the Balanidæ, compared with the Lepadidæ, is that the entire surface of the broad basis, which answers to the whole peduncle in the Lepadidæ, is firmly cemented down to the supporting object, instead of merely the basal end of the peduncle. The cement issues either in a cellular condition, or more commonly as a fine network, which, at a short distance from the orifices (Pl. 28, fig. 4 a, z), becomes so fine as to form a sheet or layer: I may here recall the fact, that in the cement proceeding from the disc of the antennæ, in some species of Lepas, a similar structure was observed. The cement itself presents the same transparent, brown, laminated, structureless appearance, and the same chemical reaction, as described in my former volume. The cement has the capacity of occupying and filling up all inequalities in the supporting surface; I have seen it, when spread over an encrusting Flustra, present an exact model of every cell; in the case of Coronula, it seems, as we shall immediately see, to have the power of penetrating into, and even almost blending with the epidermis of the supporting Cetacean. The last-formed cement-glands and cement-ducts present a delicate and transparent appearance, and contain cellular matter; whereas the old cement-glands, and sometimes the old cement-ducts, are filled with brownish cement, not acted on by boiling potash. The foregoing remarks are confined to the sub-family Balaninæ, for I have not been able to examine thoroughly the Chthamalinæ, and can only affirm, that in Chthamalus and Pachylasma the cement-ducts repeatedly bifurcate and inosculate, in the same manner as in the Balaninæ. I will now proceed to describe, in some detail, the cementing apparatus in the several following genera.

Coronula.—The cementing apparatus is here more simple than in any other genus of the Balaninæ, and I have studied it more carefully. The basal membrane of Coronula balænaris is figured in Pl. 28, fig. 1 a, and must first be described; its relation to the shell will hardly be understood without looking at the outline of the folded walls of this species, in Pl. 16, fig. 5. The basal membrane closes the central circular hollow, and is continuous with rays (not represented in Pl. 28) extending under the doubled walls and terminal transverse loops. It has eighteen concave sides, corresponding with the inner ends of the folded walls, for each of the six compartments is trebly folded. The membrane consists of successive, conformable slips (c′, c′), bordered exteriorly by thickened yellowish rims, and internally overlapping (when viewed from the inner side) the few last-formed slips, and then thinning out. The membrane forming each slip is itself laminated. The middle portion, about 1/50th of an inch in diameter, is rather opaque, owing to the slips being so close together. Beyond this central part, the slips suddenly increase in size, but yet have a different shape from the 18-sided outline, which they ultimately assume: this difference is owing to the great changes in shape, as explained under the genus Coronula, which the shell undergoes, when the walls at first assume their folded structure. The walls are invested by longitudinally striated membrane (p, p, p, fig. 1 a), which turns in under their basal edges; and this membrane is united with the basal membrane, by what I shall call the circumferential slip (b), and which is shaded in fig. 1 a, simply for the sake of catching the eye. It is the circumferential slip of membrane which sends rays under the spoke-like folded walls: thin as it is, this slip is yet laminated, but is not bordered by thickened edges. The membrane investing the walls is, like the basal membrane, formed of successive slips with thickened edges, which overlapping (viewed from the inside) the last-formed slips, project beyond them, and so face the edges of the slips in the basal membrane; they are only obscurely indicated in fig. 1 a. The circumferential slip (b) lies over (as viewed from within) both the basal and wall membrane. This whole structure will, perhaps, be best understood by the sectional diagram (fig. 1 b), in which the letters (c′, c′) show the slips of basal membrane; (p) the parietal membrane, coating the outside surface of the walls of the shell, not here represented; (b) the circumferential slip overlying both; and (z, z) the layers of cement, which may for the present be disregarded. In order to allow, of the growth of the shell, the circumferential slip (b) periodically splits in the middle, all round, in a line exactly conformable to the edge of the last-formed slip of basal membrane; and likewise in straight, medial lines under the spoke-like (cut off in fig. 1 a) doubled walls. I have seen, under a high power, the line of splitting, very shortly after its formation, with the two edges ragged and near together, with an extremely narrow, new circumferential slip just formed, between and over the edges of the previously formed slip. What causes the circumferential slip to split so symmetrically, I cannot say: the opposed edges, after a time, become thickened, apparently from adhering to the underlying layer of cement, as will presently be described. The circumferential slip continues increasing in breadth till the period of its splitting arrives, by which time it has become much broader than the last-formed slip of basal membrane; and after the splitting takes place, the interior half towards the basal membrane, forms a new basal slip all round the basis, and the exterior half adds a new slip to the membrane investing the walls. This latter membrane being inflected under the basal edges of the walls, is, during the growth of their edges, drawn straight down, the newly-formed portion taking the inflected position.

In the sectional diagram, (1 b) the circumferential slip is not yet broad enough to split; when it has become so, it will split under the letter (b). The slips of basal membrane are, as may be seen in fig. 1 a, narrower towards the circumference; but the two or three last-formed slips, are out of proportion narrower than the others; and it is certain, from the comparison of the basal membranes of specimens of different ages, that these will afterwards increase in width.[71] I have seen no other instance, in Cirripedes, of growth in membranes, except at their extreme margins: I suspect that these last-formed slips are pulled, during the downward and outward growth of the shell, a little from over the last-formed slips, new and larger laminæ being all the time thrown down, so as to prevent any fissure being formed. I also suspect that the gradual increase in width of the circumferential slip itself, is due to the opposed edges of the underlying and last-formed circumferential slip being dragged further apart from each other, new and wider laminæ of membrane being continually thrown down; the new circumferential slip being thus, also, all the time thickened, as well as rendered broader.

The central slip or rather disc of membrane, is 3/400ths of an inch in diameter; and this shows the basal diameter of the shell immediately after the metamorphosis. In the middle of this little disc I saw, in several specimens, the prehensile, pupal antennæ; I made out distinctly the ultimate segment with its bristles, and, as I believe, the disc-segment, which was 7/2000ths of an inch in diameter; but this portion was much obscured by the quantity of cement. When the corium is removed from the inner side of the basal membrane, the two chains of glands, extending from exactly over the antennæ in the centre about half way towards the circumference, are conspicuous. The cement-trunk, connecting the glands, is thin, and at the further end is always broken off, by the removal of the corium and overlying layer of ovarian cæca, into which the two cement-trunks enter; and without which removal, nothing could be seen. The two chains of glands form a very large angle, open towards the rostral end of the shell, as represented at fig. 1 c, much enlarged; by a mistake in fig. 1 a, the two have been drawn in a straight line. The cement-trunk increases in diameter in proceeding from the centre to the circumference, and the glands likewise increase in size, at the same time altering somewhat in shape. From near the lower side (the basal membrane being viewed from within) of each gland, two cement-ducts proceed, which pour out their contents beneath the basal membrane. The orifices of the ducts always exactly face the middle folds of the two lateral, and two carino-lateral compartments. In a full-sized specimen, there are from thirty-five to forty cement-glands on each side, always corresponding exactly with the number of slips of basal membrane, including the circumferential slip, to which the last-formed pair of glands and cement-ducts belong. In correspondence with the great number and narrowness of the central slips of membrane, so are the cement-glands towards the centre numerous and very small. All the glands, in the more central parts, consist of a mere transverse enlargement of the cement-trunk; but the exterior and larger glands, which are more closely packed together, are more globular or pear-shaped; and the two ducts (of which the one on the rostral side is considerably enlarged at its base) do not come out of the gland exactly at the same level. The trunk, connecting the glands, runs straight from one to the other. The ducts proceeding from the outer and older glands, on the carinal side, are much curved (Pl. 28, fig. 1 c). To give an idea of the dimensions of the several parts, I may state that the largest ducts were 3/1000ths of an inch in diameter, and the glands belonging to them nearly thrice as much, measured in the direction of the cement-trunk; on the other hand, some of the ducts from the small central glands had a diameter eighteen times less than that of the largest ducts. Towards the circumference, the ducts that proceed from the older and larger glands are piled one exactly over the other—the last formed being the topmost, and all are imbedded in the corium which overlies the basal membrane: this position of the ducts, one over the other (which could not be well represented in figs. 1 a and 1 c), is owing to their all debouching at the same exact point. But the ducts form the smaller and younger glands, when the shell had a different shape, are spread out, and are all attached to the basal membrane. Altogether, the basal membrane of Coronula, when well cleaned, and examined under a moderately high power, presents the most singular and elegant appearance.

We now come to the cement-tissue: this lies on the under or outer side of the basal membrane; it is not represented in figs. 1 a or 1 c, but only in the sectional diagram, 1 b, by the letters z, z: it presents its usual character and appearance, like solid glue or brown gum, but is obliquely laminated and sub-laminated: it forms a layer, much thicker than the basal membrane itself, being as much as .004 or .005 of an inch in thickness. It is, however, difficult to ascertain its thickness, from the singular manner in which it penetrates into and almost blends with the epidermis of the whale’s skin; so much so, that for a considerable time I thought (not then knowing anything about the cement of Cirripedes) that this transparent horny substance probably answered to a corn on the human foot produced by pressure. But I soon observed that this horny substance certainly extended into and up the cement-ducts; and this fact first led me to the examination of the whole subject in the several genera of Lepadidæ and Balanidæ. It was not difficult to remove the cement-ducts, leaving small portions of the contained cement projecting freely up as points from the general surface of cement. The cement adheres slightly to the whole basal membrane, but chiefly to the yellowish rims or edges of the successive slips; and it is indeed this adhesion which, I believe, produces the rims; for the circumferential slip, when first split, had very thin ragged edges. The cement also extends under the spoke-like prolongations of the circumferential slip, and likewise some way up the sides of the walls.

The cement-glands, the trunk, and the ducts, except the two, three, or even four last-formed ones, are all filled with an apparently solid thread of transparent, brownish cement, differing in no respect from the cement under the central parts of the basal membrane. In one instance, in which the last-formed pair of glands and ducts had apparently been only just developed, they were so perfectly transparent that I could distinguish them only under certain lights, and I could not perceive any contents. The last-formed glands and ducts always appear very delicate, and include a tube of very delicate tissue, containing more or less of granular matter. The next succeeding pair of glands and ducts are always more opaque, and contain much more granular matter; which, in the next, or next but one, may be seen passing into the state of pale brown, transparent chitine. I have seen some most distinct instances, in which, in the same duct, the part towards the centre of the basis was filled with homogeneous cement, and the part towards the circumference with still plainly granular matter. In the successive circular slips of cement-tissue, lying attached under the circumferential slip and under the two or three succeeding slips of basal membrane, an exactly analogous series of changes is presented; as indeed might have been expected, as the slips of cement are absolutely continuous with the contents of the ducts. Moreover, if a vertical section be made across one of the last-formed slips of cement, it may sometimes be seen to be apparently in the act of separating into layers, with the lower layers more pulpy, elastic, and white than the upper layers, which are less coherent, and show as yet even still less the character of cement. The cement under the circumferential and adjoining slips, often presents a peculiar wrinkled appearance, in lines conformable with the outline of the basal membrane; but I do not believe that these are real wrinkles, though so like them; they seem to consist of sinuous threads, longer or shorter, sometimes slightly branched, crossing and inter-joined, and composed apparently of faintly coloured cement. I suspect that these threads are formed by the union and subsequent drawing out of aggregations of that matter, which within the ducts is first granular, and then changes into cement; for at the orifices of the ducts these wrinkled threads sweep outwards in curved lines on both sides. The cement in these early stages adheres, with very little force, to the basal membrane; and with only a little more force to the underlying layers of cement; in fact, till it assumes the brown translucent appearance, like solid glue, it hardly seems to act as cement.

How the cement reaches the skin of the whale, will be best understood by referring to the sectional diagram (Pl. 28, fig. 1 b). When the circumferential slip of membrane (b) splits, a new circumferential slip will be formed over it, together with new cement-glands and ducts, and cement (z, z) will issue from four new orifices, and will extend on both sides of these orifices, till the ends meet and become united, thus forming a narrow, 18-sided, continuous, new slip of cement, with 18 spokes proceeding from it. I have not noticed lines of union in the cement of any one slip; but the matter forming each slip, certainly has proceeded from four distinct orifices. Seeing how perfectly successive layers of cement often become blended together, lines of union or junction, could hardly be expected to be preserved in the same individual layer. When the circumferential slip of basal membrane splits, the underlying slip of cement, which we will call A, does not split, but becomes stretched, so that the newly formed slip of cement, which we will call B, does not reach the skin of the whale. As the new circumferential slip of basal membranes goes on increasing in width, A continues to be stretched, but does not split, till at least another circumferential slip of basal membrane has been formed and has been split, and till B has been also stretched. By this time, the cement-tissue A has assumed its normal structure, and has the power of adhering to the whale’s skin, which it has now reached, owing to the splitting of the under and older slips of cement. At the next period of growth, A itself will split, and B will touch the whale’s skin and adhere to it; and this, also, will ultimately split. It results from this action, that the cement has a stretched, and sometimes even a fibrous appearance, with the lower edges of the layers, of which each slip of cement is formed, thinning out. I have before stated, that the two or three last-formed slips of basal membrane are formed at first too narrow, and apparently have to be dragged outwards, over each other; and it is perhaps owing to this circumstance, and to globules of cement having first adhered to the under surface of the slips of basal membrane, that these slips are studded beneath with parallel little vermiform bodies, sometimes of considerable length, and furnished with heads, all directed outwards. These tapering, vermiform bodies have a considerable resemblance to the threads before mentioned, which give the wrinkled, concentric appearance to the newly-formed layers of cement, and have probably a closely analogous origin: in one case, indeed, it appeared as if some of these concentric threads were in process of being drawn out at right angles to their original course. Lastly, it should be observed, that as the exterior half of the membrane of the circumferential slip, after each splitting, is dragged down, and thus comes to invest the outer surface of the wall of the shell (the wall not being represented in the diagram, but standing where the letter (p) stands), so it must be with the cement, which thus likewise comes, in an unusual manner, to invest the outer surface of the folded walls of the shell, and attaches them to the skin of the whale,—which latter is always growing upwards, and tending to bury the shell.

Platylepas decorata.—This genus is closely allied to Coronula, and the cementing apparatus is essentially similar. In one specimen, I counted no less than forty-nine slips of basal membrane, each of which, of course, had its pair of cement-glands, and each of the latter its two ducts. The glands consists of a transverse enlargement of the trunk, as in the early stages of Coronula. Neither the glands nor the duct, when old, become filled up with cement, but only the main-trunk. The ducts are very delicate and thin; the larger ones being only 3/10,000 of an inch in diameter. The glands stand some way apart on the two cement trunks; and the latter, instead of being straight as in Coronula, are deeply serpentine; the glands are formed on each bend, so that, though all on one side are connected on the same trunk, they form a double row on each side of the basal membrane. The basal membrane (in the centre of which I distinctly saw the antennæ of the pupa) has six deep bays or excisions, corresponding with the midribs (see Pl. 17, 1 a, 1 d) of the six compartments; and the two ducts from each gland, on the right and left sides, debouch at the heads of the four lateral excisions, exactly opposite the midribs of the lateral and carino-lateral compartments. The later-formed glands, owing to all of them being situated some way apart from each other on the two cement-trunks, lie further from the centre of the basis than do the orifices of their ducts; hence the later-formed ducts are directed a little backwards, or from near the circumference towards the heads of the deep excisions.

Tubicinella.—The cementing apparatus is here less symmetrical; but this, I believe, is chiefly owing to the basal membrane being formed of successively larger discs (not slips) of membrane, thrown down not quite concentrically one over the other; each new disc of membrane seems to cover the last-formed cement-glands and ducts; and there are as many ducts and glands as there are discs of membrane, all adhering together. In one specimen, it appeared that normally there were four sets of cement-ducts, as in the allied genera of Coronula and Platylepas; but in other specimens, the ducts were distributed very irregularly. In one case the two cement-trunks extended parallel and close together, one of them terminating long before the other. I have given a figure (Pl. 28, fig. 3) of three of the cement-glands, removed from the basal membrane, together with their ducts. The cement-trunk (f f) seemed to be a little enlarged, and to be crossed by septa, as it entered the glands (h), but I could not make out this structure clearly enough to be represented. Whilst young, the cement-glands stand some little way apart from each other; and in the figure given of some of the last-formed glands, they are hardly separate enough. Each gland gives out obliquely, on one side, a cement-duct (c) which I traced in several cases to the margin of one of the discs of basal membrane, where cement issued from it; and on the opposite side, a tapering spur (b), varying in length, which may be called, and I believe really is, a rudimentary duct. Of these spurs we shall meet many instances in other genera. The duct (c) and the spur (b), close to where they entered the gland, in some specimens gave off, at about right angles, short blunt points, or rudimentary branches. This duct and spur correspond, I believe, with the two ducts in Coronula; but besides these, a duct (a) is given off from one end of the gland, from the surface opposite to that at which the cement-trunk enters. This duct (a) is very singular, from always forming a loop (a′), with two spurs projecting from it: these two spurs occasionally spring from a common point: I have seen nothing like this structure in any other Cirripede. This duct (a) runs, like the duct (c), to the margin of its own disc of basal membrane, where it debouches. Besides these ducts, in the best specimen which I examined, there were two other sets of ducts, which were slightly zig-zag, and at each angular bend, a mere knob or point, or at most a short branch, was given off; but this branch seemed never to run to the margin of the basal membrane or to give out cement; whereas the main branch did give out cement. I was not able to trace these ducts to their glands. In these zig-zag ducts, and in the rudimentary points sometimes observed at the base of the duct (c), and likewise at the base of the spur (b), we see the first indication of that tendency to bifurcation, so strongly characteristic of the cement-ducts in all the genera, excepting those already described, which are allied to Coronula.

Chelonobia patula.—The cementing apparatus is here chiefly remarkable for the thinness and straightness of the main trunk, (f f, Pl. 28, fig. 2), and from the great distance at which the glands stand apart; had another gland been drawn, it would, on the scale here used, have stood exactly under the two upper, (c′ c′) in fig. 1 c. We here see that the trunk (f), before entering the gland (h), has an enlarged portion (g); this, I suspect, is a very general structure. Each gland gives out, on opposite sides, two ducts (a a, b b), larger even than the main trunk; and these ducts bifurcate repeatedly, and inosculate. By this inosculation it is not improbable that all four ducts, proceeding from the two glands of the same age, may be connected together; certainly the bifurcating branches from the same duct thus become repeatedly connected. For the first two or three bifurcations the ducts decrease very little or not at all in diameter; but nearer the circumference they become smaller. The ducts, also, proceeding from the younger and smaller glands, are, of course, proportionably smaller. In one case I was able to count four bifurcations in the duct between the gland and the edge of the basal membrane. It follows from this structure, that the basal membrane, at each period of growth, is cemented down by cement issuing from several orifices; but we shall presently find that in other genera the cement proceeds from many more orifices. In fig. 2 there is represented, by the aid of the camera, a small portion (from the outer (a) to the outer (b) being 12/100ths of an inch in length) of the basal membrane, with all the several cement-ducts adhering to it, which I could distinguish, and drawn of their proper relative sizes; this figure also shows some of the bifurcations, but no inosculation happened to be included in the space here given; the basal membrane itself has not been represented. In taking a view of a considerable portion of the basal membrane, especially towards the circumference, some parallelism in the branches could be perceived; one set of branches tending to run in the direction of the ray of the circle, and the other set in the line of the circumference.

Elminius Kingii.—The cement-glands here resemble those of Chelonobia, but the trunk does not seem to be enlarged before entering the gland. The glands are situated rather far apart; and the chief peculiarity is, that the trunk connecting the glands is as tortuous as the track of a worm. Each gland gives out two ducts, which bifurcate repeatedly, and often inosculate, making, in parts, an hexagonal mesh-work: some of the branches do not debouch on the basal membrane, but terminate in blunt points. So numerous are the ducts, that the basal membrane may be compared to pieces of paper with the fine fibrous branching roots of some plant dried and heaped on it. Some of these ducts are very regularly jointed, and resemble a conferva,—an appearance which I believe is owing to divisions in the contained cement; other ducts are partially marked by little wrinkles, as presently to be described under Balanus. The cement, instead of, as heretofore, invariably forming a slip round and beneath the circumference of the basal membrane, here often forms little, independent, circular, and irregularly-shaped discs, each of which has issued from a single orifice. I may here add that in two species of Tetraclita I saw the cement-ducts repeatedly bifurcating, with some of the branches inosculating, as in Elminius and Chelonobia.

In Balanus balanoides, which, like all the Cirripedes hitherto mentioned, has a membranous basis, I could only make out an amazing number of bifurcating and inosculating cement-ducts, of very various diameters. The cement-tissue, on the under side of the basal membrane, generally consisted of little circular discs, on an average from 2 to 5/1000ths of an inch in diameter; but there were also globules and short tortuous threads of cement. In some very much elongated and crowded specimens—in which, during the downward growth of the walls, the basal membrane had ceased to reach the surface of attachment, and being thus suspended had become, as viewed from the outside, deeply concave—the cement had apparently continued to try to reach the rock, and now hung down in the form of two thickish roots, some tenths of an inch in length. These roots were round, and tapered either to a fine or blunt point; one was doubled on itself, and so had become united; in the other, I could perceive five layers or sheaths of the cement-tissue, one within the other; the innermost of these layers, which once, no doubt, formed the outside surface of the root, was only a quarter of its present length.

In Balanus tintinnabulum, the basis is calcareous: when its upper surface is cleaned, dried, and examined under a good light, the numerous larger cement-ducts can be seen, even by the naked eye, or under a weak lens, and present an elegant appearance. These larger ducts run in parallel lines from the two chains of glands towards the circumference. They are all encrusted with calcareous matter, and in the more central parts are hidden under it; at each period of growth, when the basis is added to round the circumference, it would appear that a layer of excessive tenuity of shell is thrown down over the whole surface, just in the same way as in Tubicinella, at each period, a new and larger disc of membrane was thrown down over the pre-existing membranes with their cementing apparatus. The cement-glands, in the middle of the basal plate, seem often to give rise to small abnormal depositions of calcareous matter. When the basis (it is best to take a young specimen) is slowly dissolved in acid, all the cementing apparatus is left uninjured, adhering to the delicate tissue which before existed in a calcified condition. Near the middle I saw the two antennæ of the pupa; and from them the two cement-trunks extended about half-way towards the circumference. These two chains of glands are often placed very irregularly, but they tend to form, as in Coronula, a large angle, open towards the rostral end of the shell. The glands, close to the old antennæ, commence abruptly, of rather large size: the later-formed glands, with their ducts, are in regular order larger than the younger ones, and stand much closer together. After immersion in acid all the glands and ducts appeared empty, instead of the older ones being, as in Coronula, filled with cement. In one case I counted on each trunk twenty-five glands, besides some smaller obscure ones close to the centre.

In Pl. 28, fig. 4 b, I have given a drawing of two of the cement glands: the cement-trunk (f f) is smooth and apparently cylindrical: it becomes enlarged (at g) before entering the gland: it seems even to be prolonged across the gland under the form of a narrow bar (not represented), which apparently serves to keep the two ends of the trunk, on the two sides of the gland, in their proper relative places and distances. The gland itself is an elongated bag (h), which properly lies exactly over the enlarged portion (g) of the trunk, but in the drawing has been purposely displaced: it gives rise, in the later-formed glands, to a sort of neck (see the upper gland), which is either so long as to deserve rather to be called a duct and which soon bifurcates, or is quite short (see the lower gland) and gives rise to two separate ducts. On the opposite side of these glands, there is a spur (m), of greater or shorter length, which is evidently a rudimentary duct, for in the younger glands it existed as a perfect duct. Moreover, the first-mentioned duct often gives off branches (t′), having an exactly similar appearance with the spur (m). The membrane of which the cement-trunk (f), with the enlargements (g), is composed, is smooth, but that of the glands and of all the ducts, presents a very peculiar appearance, which at first would be called scaled, but more properly perhaps notched,—each notch being apparently formed by a line of thickened membrane, extending obliquely round only a short portion of the tube, and indenting it. The ducts, which I measured, were between 1/3000th and 4/3000ths of an inch in diameter.

In fig. 4 a, I have given a drawing of the two chains of glands, but with only those ducts figured which proceeded from the last-formed pair of glands. The specimen here drawn was old; and it is rare to find the structure of the ducts so simple. From both glands[72] a neck or thick duct arises, which soon bifurcates; one branch runs direct into the circumferential duct, and the other (t) bifurcates again; of the latter, one branch unites with its fellow from the opposite gland, and then forming a single duct (t′) enters, as do the two other branches, the circumferential duct. Thus, into the latter, five main ducts enter: the position of their points of entrance, with respect to the shell, varies considerably; but I think the five points tend to face the middle of the rostrum, and middle of the two lateral compartments on each side. In some other specimens, in which the ducts were nearly as simple, I observed that the neck or main duct at once divided into three branches, instead of into two, with one soon bifurcating; and on one side a rudimentary branch or spur was given off (above t), indicating a tendency to an additional bifurcation. In the later-formed glands, the ducts proceed only from the outer sides and form the ends of the glands furthest from the centre; but in the earlier-formed and smaller glands of the same individual, other ducts proceed from the inner sides, where in the older glands the spurs (m) are situated: moreover, in the younger glands, all the ducts bifurcate much oftener (how often I was not able to ascertain), before entering the circumferential duct; many of the branches, however, terminating in spur-like points. Now if we imagine twenty or thirty repetitions of the ducts given in fig. 4 a, (independently of the greater complication of the ducts of the younger glands), each a very little smaller than the other, and placed, with the main branches parallel, one over and within the other, we shall gain some insight into the wonderfully complicated structure of the cementing apparatus in this and many other species of Balanus.

I have as yet only alluded to the circumferential duct (i, i, Pl. 28, fig. 4 a): we have not hitherto met with this duct, but I suspect that the branches which in Chelonobia inosculate, and which seem to run nearly parallel to the circumference of the basal membrane, answer the same purpose of connecting the ducts together, and are, perhaps, strictly homologous. In this, and some other species of Balanus, the last-formed circumferential duct runs round the margin of the upper lamina of the basal plate, close to the basal edges of the walls; and as these latter have projecting longitudinal ribs, the duct curves a little round each rib; so that the whole duct is formed by as many short inwardly curved portions as the walls have ribs, or longitudinal septa. Between the basal extremities of these parietal, longitudinal septa, the extremities of the radiating septa of the basis project and enter; and along the crests of the latter, little branch-ducts (i′), proceeding from the circumferential duct, extend. In the basis, beneath the tubes formed by the just-mentioned radiating septa, there is a cancellated shelly mass (which, in fig. 4 a, was of unusual thickness), and along the crests of the branching ridges forming this cancellated mass, the sub-branches of the above branch-ducts (i′) run; these soon become so minute as not to be distinguished by the highest powers, and thus form a sheet of cement, which attaches the last-formed zone of the shelly basis to the supporting surface. At what point the membrane forming any one duct ceases, the cement-tissue being alone left, I was not able to ascertain; but the lower parts of the reticulated slip (z, z, fig. 4 a) closely resembled the cement-tissue which surrounds the disc-segment of the pupal antennæ in Lepas australis. The circumferential duct, here and there, forms little loops, as may be seen in fig. 4 a: and often two branches, running along the crests of two adjoining basal septa, proceed from a common point of the circumferential duct. The cement itself, under different parts of the basis, appears as little separate discs, as threads, globules, and as a fine network, but most commonly as simple layers. As each thick zone of shelly matter is added round the basis, the exterior branches of the ducts, between the circumferential duct and the new layer of cement beneath, are fairly imbedded in shell, and are for ever hidden, without, indeed, acid be used for the dissolution of the calcareous matter: so, also, the pre-existing ducts and glands, and the main trunk, would all have been hidden, if the layer of calcareous matter, which, I believe, is thrown down at each period of growth over the entire surface, had not been of excessive tenuity.

I cursorily examined the cementing apparatus in Balanus galeatus, improvisus and crenatus, which have all calcareous bases, and belong to different sections of the genus; and the structure seemed to be essentially the same. In Bal. galeatus, I found the cement-ducts varying in diameter from 1/4000th to 1/10,000th of an inch in diameter. In B. improvisus, the cement-glands do not differ much from those of B. tintinnabulum; but the cement-ducts bifurcate often before entering the circumferential duct; and the little branches, which proceed from the latter, are very short, and almost immediately, owing to the thinness of the basis, blend into a slip of cement.

I hope to be excused for describing at such length, the apparatus by which sessile cirripedes are permanently attached to a supporting surface; for this is the great leading character of the sub-class, not hitherto observed in any other Crustacean.[73] It is not easy to overstate the singularity and complexity of the appearance of the basal membrane of a Balanus or Coronula: and when we consider the homological nature of the apparatus, the subject becomes still more curious: I feel an entire conviction, from what I have repeatedly seen in several genera of the Lepadidæ, both in their mature and pupal condition, and from what I have seen in Proteolepas, that the cement-glands and ducts are continuous with and actually a part of an ovarian tube, in a modified condition; and that the cellular matter which, in one part, goes to the formation of ova or new beings, in the other and modified part, goes to the formation of the cementing tissue. To conclude with an hypothesis,—those naturalists who believe that all gaps in the chain of nature would be filled up, if the structure of every extinct and existing creature were known, will readily admit, that Cirripedes were once separated by scarcely sensible intervals from some other, now unknown, Crustaceans. Should these intervening forms ever be discovered, I imagine they would prove to be Crustaceans, of not very low rank, with their oviducts opening at or near their second pair of antennæ, and that their ova escaped, at a period of exuviation, invested with an adhesive substance or tissue, which served to cement them, together, probably, with the exuviæ of the parent, to a supporting surface. In Cirripedes, we may suppose the cementing apparatus to have been retained; the parent herself, instead of the exuviæ, being cemented down, whereas the ova have come to escape by a new and anomalous course.

Affinities, Classification, Variation.

Under the Order it has been stated that the Balanidæ, are, on the cirripedial type, the highest in the class; that is, they are the most complicated, but not (to use Professor Owen’s term) by mere vegetative repetition. Amongst the Balanidæ, the first section of the genus Balanus may be taken as typical; here we have the structure of the shell extremely complicated, yet beautifully adapted for strength, and for the protection of the included body. The cementing apparatus is here, also, most complicated. I have divided the Balanidæ into two natural sub-families, the Balaninæ and Chthamalinæ, in accordance with certain differences in the structure of the shell and of the animal’s body: that this division is natural, might almost be inferred from the one fact, that all the characters by which the Chthamalinæ differ from the Balaninæ, are those by which the former approaches the family of the Lepadidæ; moreover, certain anomalous characters in the Chthamalinæ, as the supplemental whorls of compartments in Catophragmus, and the presence of caudal appendages in this same genus and in Pachylasma, reveal this same affinity. The only objection which I can see to the separation of the sixteen genera into the above two sub-families, may be drawn from the degree to which they blend together; thus, as far as the shell is concerned, Chelonobia, in one important internal point of structure, tends to assume the character of the Chthamalinæ; and on the other hand, Pachylasma, a member of the Chthamalinæ, has a shell, which if not examined during its earliest growth, would be placed without doubt amongst the Balaninæ. But it fortunately so happens, that in no one character of the body does Pachylasma approach the Balaninæ more nearly, than do the other members of its sub-family; or Chelonobia approach, in the same respects, the Chthamalinæ. It is only in Chthamalus, of which the shell clearly places it in the sub-family bearing its name, that in some of the species, the less bullate labrum,—the larger palpi,—the lower teeth of the mandibles being laterally double,—and the lower segments of the third pair of cirri being thickly clothed, like the lower segments of the second pair, with bristles—all show that these species make an approach in structure to the Balaninæ.

It will be seen that I have divided the Balaninæ into two little groups, according as whether the branchiæ consist of one or of two plicated folds of membrane, and as whether or not the scutum and tergum are articulated together. I have been greatly tempted to follow Drs. Leach and Gray, who have separated the second of these groups, containing the genera Coronula, Tubicinella, Xenobalanus, and Platylepas, into the sub-family of the Coronulinæ. Certainly these genera have a peculiar aspect in common, and agree in being parasitic and imbedded in the skin of Cetaceans, as is the case with the first three genera, or in that of turtles, manatee, and sea-snakes, as in Platylepas. Though these genera possess several peculiar characters, yet I can find none common to all four, excepting their imbedment in the skin of Vertebrata, their double branchiæ, and their non-articulated opercular valves; and these I do not think of sufficient importance to serve for the separation of a sub-family; for in Chthamalus, one species has double branchiæ, one species has no branchiæ at all, and the other species have single branchiæ; so again in Chelonobia, the scutum having only a horny articular ridge, makes an approach to Coronula and its allies. I may further specify that the folded walls, a singular character common to Coronula, Platylepas and Xenobalanus, fails in Tubicinella; the open tubes and the imperfect outer lamina of the parietes towards their bases, are characters which fail in one species of Platylepas; the muscles running to the opercular valves being thinly spread out, and partially without transverse striæ, is also a character which fails in Platylepas; the simplicity of the cement-ducts partially fails in Tubicinella; and lastly, the existence of small intermediate teeth on the mandibles, fails in Xenobalanus: hence, I repeat, I have not thought it prudent to admit the sub-family of the Coronulinæ though in many respects a very natural group.

The genera in the Balaninæ and Chthamalinæ are founded chiefly on the number of the compartments (the number being apparently due, as previously explained, to the fusion or abortion of certain of the eight typical compartments); and secondarily, on the nature and even form of the basis, and on the porosity of the walls. In Coronula and its allies, the non-articulated opercular valves and deeply folded walls come into play. As a justification for using these characters in distinguishing the genera, and even to a certain extent in separating the two sub-families, I must call to mind that the shell, with the basis, is not merely a dermal envelope, as amongst Molluscs, but actually consists of the first three segments of the head. The parts of the mouth and the cirri are of very little service in distinguishing the genera,—a singular fact, considering that most of the genera amongst the Lepadidæ could be distinguished by these organs,—though trifling details in their structure sometimes come in useful as specific characters. Balanus, with the sub-genus Acasta; Pyrgoma, with the sub-genus Creusia; Tetraclita, and Elminius, are genera of about equal value; though perhaps the two latter are rather more nearly related together than to the others. Chelonobia is more distinct; it shows some little affinity to the Chthamalinæ, and likewise to the four following genera. Coronula, Platylepas, Tubicinella, and Xenobalanus, are genera quite distinct from the foregoing, and from each other; yet, as we have just seen, palpably allied together. Amongst the Chthamalinæ, Pachylasma, Octomeris, and Catophragmus, are more closely related to each other than to the other two genera of the sub-family; yet Pachylasma, as far as the shell is concerned, leads into the Balaninæ, and Catophragmus into the Lepadidæ; Octomeris leads towards Chthamalus, and Chthamalus towards Chamæsipho.

Variation.—The discrimination of the species in most of the genera, offers very great difficulties. I cannot too strongly impress on any one intending to study this class, not to trust to external characters: he must separate and clean, and carefully examine the internal structure and form of the compartments, and more especially of the opercular valves. After considerable experience, when numerous varieties of a species have been carefully examined, the eye acquires a sort of instinctive knowledge, by which it can recognise the species, though the character cannot be defined by language; but I have found that no amount of experience with some of the commonest species, will save frequent and grave errors, as long as external characters alone are trusted to. Not only does every external character vary greatly in most of the species, but the internal parts very often vary to a surprising degree; and to add to the difficulty, groups of specimens not rarely vary in the same manner. After having given up several years to the study of this class, I must express my deliberate conviction that it is hopeless to find in any species, which has a wide range, and of which numerous specimens from different districts are presented for examination, any one part or organ,—which from differing in the different species is fitted for offering specific characters,—absolutely invariable in form or structure. I may in one respect even go further, and affirm, that, if in a species, any part or organ differs remarkably from the same part in its congeners, then if many specimens are examined, especially when collected from different districts, such part or organ will be found eminently variable. I may instance the antenniformed third pair of cirri in Chthamalus antennatus, the teeth on the posterior cirri in Acasta sulcata, the terga in Pyrgoma dentatum, the adductor ridge of the scuta in Pyrgoma cancellatum and in Creusia, and other such cases: hence it will not do to found a species on a slight, or sometimes even on a considerable difference, in any single point or organ. On the other hand, I am far from asserting, that if only half-a-dozen specimens of some rare species of Cirripede be brought from some one quarter of the world, characters beautifully defined may not be readily discovered. In determining what forms to call varieties, I have followed one common rule; namely, the discovery of such closely-allied, intermediate forms, that the application of a specific name to any one step in the series, was obviously impossible; or, when such intermediate forms have not actually been found, the knowledge that the differences of structure in question were such as, in several allied forms, certainly arose from variation,—for instance, in the case of two shells otherwise identical, one being longitudinally ribbed and the other smooth, a character which we know to vary,—but I have always used this argument from analogy with great caution. Finally, as in the large genus Balanus, there is an especial amount of variation, I have there entered in detail on this subject; and I hope that those interested in it, will refer to that discussion, which is almost verbally applicable to some other genera of the family, as Tetraclita and Chthamalus.

Rate of Growth, Exuviation, Powers of Repairing Injuries.

In my former volume I have shown that the pedunculated cirripedes grow rapidly; this is likewise the case with the Balanidæ. Mr. Stutchbury informs me that he has seen numerous specimens of Balanus tintinnabulum from 2 to 3 inches in height and from 5 to 6 inches in circumference (and this is nearly the full size which the species attains), on a vessel which had been to sea only during one year. At Coquimbo, in Chile, I have seen a specimen of B. psittacus 1.3 of an inch in basal diameter, and .8 in height, adhering to a chain that had been only six months under water. Poli, also, gives the case of some Balani (probably B. perforatus), which, in about four months, had attained a basal diameter of 1 inch, and a height of 1-1/6th of an inch. Balanus balanoides is a smaller species, and of slower growth; for the late Mr. W. Thompson, of Belfast, found that in three months from July 3d, certain marked specimens had increased from 2-1/2-3 lines to 4-1/2 lines, which is the usual maximum size attained in that locality. From other observations, Mr. Thompson believes that the extreme duration of life of this species is about two years: whether the other and apparently quicker-growing species, are shorter-lived, I have no means of judging.

In accordance with this rapid growth is the frequency of the periods of exuviation. Mr. Thompson kept twenty specimens of B. balanoides alive, and on the twelfth day he found the twenty-first cast-off integument, showing that all had moulted once, and one individual twice, within this period.[74] This frequency of exuviation, together with the durability of the cast-off integuments, explains the astonishing masses of exuviæ, which Mr. Peach assures me he annually has observed off the coast of Cornwall; they are most abundant in April and May, but he has seen quantities also in September; he could easily, as he tells me, have filled several quart-measures with them. The specimens sent to me consisted chiefly of Balanus balanoides, perforatus, and Chthamalus stellatus. The opercular membrane, with a narrow strip from between the two scuta, and another narrow strip from between the two terga, are moulted together, in connection with the more delicate membrane lining the sack, and investing the plicated branchiæ. This membrane, likewise, is continuously connected with that covering the whole body and its appendages. As I have stated under the Lepadidæ, the inner tunic of the œsophagus, of the rectum, of the olfactory pouches, and that which enters a little way into the acoustic meatus, and the apodemes of the maxillæ, are all moulted. On the cirri and jaws, new spines are formed with their upper ends enclosed within the old spines, but with their lower ends projecting inwards, beyond the bases of the old spines, and inverted like the fingers of a glove hastily pulled off. The membranes of the body, in the act of exuviation, split, I believe, only over the prosoma. How the neat separation of the opercular membrane, from all round the sheath and opercular valves, is effected, I do not fully understand; but it is, probably, analogous to the splitting of the thick carapace of the common crab. I suspect in Coronula, in which genus and its allies the opercular membrane is not periodically moulted, that the membrane lining the sack is not always thrown off at the same exact time with that of the body. In Chthamalus stellatus, in the act of moulting, the opercular membrane is the last part that separates from the new underlying membranes: I find that this species can moult when kept in a damp box out of water. The new membranes of the body, immediately after the exuviation, are not lax in any extreme degree. The exuviæ of the genus Chthamalus, and of some other genera amongst the Chthamalinæ, can at once be recognised by the divergence of the posterior four pairs of cirri: in the case of Chthamalus stellatus I have also noticed that the animal generally dies with these cirri in the same divergent position. Finally, I cannot doubt[75] that the Triton described by Linnæus was only the exuviæ of some Balanus (probably B. porcatus); Linnæus mistaking the probosciformed penis for the mouth of his imagined distinct animal.

I have seen a few specimens showing that when the shell has been broken it can be repaired; and this I believe is effected by the growth of a crest of corium between the broken edges, and the subsequent calcification of this crest. Mr. Stutchbury possesses a monstrous specimen of Chelonobia testudinaria, in which one of the lateral compartments on one side has not been developed. The cirri not rarely get cut off, but are, as it appears, soon repaired. I have observed a singular number of examples of the act of reparation in a group of the Australian Balanus vestitus. The manner in which the cirri are repaired seems to me curious: the cut-off end is closed by a rounded scale of yellowish chitine, and then the corium, in the four or five subjacent segments, separates from the external articulated membrane, which now serves only as a case or capsule. The tube of corium thus enclosed, with its contained muscles, shrinks a little, and soon can be perceived to be in process of dividing into new and smaller segments, in one instance ten in number,—which at the next exuviation would, no doubt, be invested with an external membrane, and be freely exposed. In another instance, the pedicel of a posterior cirrus had been cut off and subsequently closed; in this instance, a whole, immature, miniature cirrus, with the two rami, each having fifteen minute segments, was thus enclosed in what had been the single lower segment of the pedicel. I have seen several specimens of Balanus balanoides, as described under that species, with several of the cirri and the penis truncated; but I believe this was owing to monstrosity, which seemed particularly to affect the male organs of generation; for no reparation seemed to be in progress. In a specimen of Coronula, however, the penis appeared to have been really cut off by accident; it had been closed, by a scab, with concentric lines, like the articular rings on the penis itself; and within the case thus formed, the corium had healed, and had become pointed, but inverted; I presume that the point would, after another exuviation, have been everted, and its length thus increased.

Geographical Range and Habits.

With respect to range, the results arrived at have no particular interest, for the species are not sufficiently numerous; and, what is still more adverse, the genera, with unimportant exceptions, range over the world; so that there is no scale of differences, and it cannot be said that these two regions differ in their genera, and these two only in their species. In all the following remarks, I have trusted exclusively to my own specific identification; and have rejected all assigned localities which appeared from any cause to be doubtful. Sessile cirripedes are found in every sea, from lat. 74° 18′ north to Cape Horn. The area included between the north point of the Philippine Archipelago and the south point of Australia, extending on the right hand to New Zealand, and on the left to Sumatra,—an area, which, though including two distinct Cirripedial regions, is small compared with the surface of the globe, yet includes a greater number of species, especially of peculiar species, than the whole rest of the known world. This is, probably, in chief part due to the broken nature of the land, affording diversified habitats, and to much of the coast being rocky. Cirripedes, from requiring to be attached, cannot live where the shores and adjoining bottom are sandy or muddy or formed of moving shingle; hence, no doubt, it arises, that there is such a remarkable contrast in the great number of the species inhabiting the bold rocky western coast of South America, and the few species living on the sloping, and often sandy or muddy or shingly, eastern shores of this continent. Hence, also, I believe, it is that not many species have been brought from India. Coral-reefs are not favorable to Cirripedes, consequently but few are known to inhabit the islands of the Pacific Ocean. Where Cirripedes can live, though the species in no district are numerous, the individuals abound in infinite numbers: I have walked over the coast-rocks of the Falkland and Chonos Islands, of Chile and Van Diemen’s Land, fairly encrusted over wide spaces with a continuous layer of Cirripedes, consisting of only two or three species; in the same manner as may be observed on many parts of the shores of Great Britain, and, I believe, of North America.

With respect to the effects of temperature on the range of Cirripedes, no genus (having more than one species) is confined to the torrid zones. Pyrgoma, from being always attached to corals, is, of course, ordinarily found in the hotter seas; but one species ranges from the Cape de Verde Islands in the torrid zone to the southern shores of England and Ireland. Tetraclita is not found in the colder seas, but is numerous in species and in individuals, on the southern shores of Australia and at the Cape of Good Hope. I may here add, that the two genera with the most confined ranges, are Chamæsipho and Elminius; the former has only two species, one inhabiting Australia, and the other the East Indian Archipelago; Elminius has four species, confined to the southern hemisphere, and inhabiting Australia, New Zealand, and South America. To return to the effects of temperature; in Mr. Dana’s great work on Crustacea, an excellent chart is given, in which the isocrymal lines, or those exhibiting the mean temperature of the waters along their course, for the coldest thirty consecutive days in any season of the year, are given; and which lines Mr. Dana has shown are the most influential in governing the distribution of marine animals. At the isocryme of 68°, Mr. Dana divides the torrid and sub-torrid zones from the several temperate zones; and at 44°, the temperate from the sub-frigid and frigid zones; but as no Cirripedes are exclusively confined to these frigid zones, we may here disregard them. From Mr. Dana’s[76] table of the areas of the torrid and temperate ocean-zones, on both sides of the equator, it seems that they are nearly as 337 to 278, in relative area; and consequently, he remarks, that the marine species in any class, if distributed equally over the two, would be one fifth more numerous in the torrid than in the temperate zones. Now of Cirripedes, taking all the orders, there are at present known 147 species; of these, 7 have doubtful habitats, leaving 140 for comparison. Of these 140, nearly one quarter, or 37, inhabit both the torrid and temperate zones, as above defined; 46 are found exclusively in the torrid, and 57 exclusively in the temperate zones; so that the temperate zones, though less in area, and having, proportionally, even a considerably lesser length of coastline, nevertheless have a preponderance in the number of species. But it should be borne in mind, that there are two great temperate districts, separated from each other by one great torrid district; and, inasmuch as the number of species in any region seems to depend in some degree on the isolation of the sub-regions, we might have expected (the other conditions now being, and the past conditions having been alike), that the two great temperate areas would have contained more species, perhaps doubly more, than the single great torrid area.

The proportional numbers, above given, are not very widely different, whether we take separately the Balanidæ, the Lepadidæ, or all together. Mr. Dana has shown[77] at length, that generally amongst the Crustacea, the species which he considers of highest rank, belong to the extra-torrid zones: there seems to me in all such cases to be some degree of vagueness in the attempt to determine which are highest or lowest, but I have already elsewhere stated that Balanus is, probably, the most eminently Cirripedial form, and exhibits in the strongest manner all the characters by which Cirripedes differ from other Crustacea; as this genus is the largest, containing 36 species, of which the habitats are known, I may state that of these, exactly one third, or 12, inhabit both zones; 9 exclusively inhabiting the torrid, and 15 exclusively the temperate zones. According to the proportions of the whole class, the numbers should have been 9 torrid, to only 11.11 temperate; so that evidently the genus Balanus (in one sense typical) inclines towards the temperate regions more strongly than does either the family or the sub-class to which it belongs.

With respect to the relation between the size acquired by the different species of sessile cirripedes, and the temperature of the localities inhabited by them, the genera Chthamalus, Tetraclita, and Balanus, alone can serve for comparison: in Chthamalus much the largest species is found in the temperate zone: on the other hand, the two largest species of Tetraclita are from the torrid zone, though one of them also sometimes ranges into the temperate seas: in Balanus, the largest species, and six other species having a basal diameter sometimes over two inches, inhabit the temperate regions; and two out of these seven species, flourish even in the Arctic seas; whereas, within the torrid zone, there are only three species with a diameter sometimes exceeding two inches, but two of these frequently become very large and massive; so that Balanus, judging from the size of the species, as well as from their range, does not require for its highest development the temperature of the torrid zones.

The greater number of the species of the Balanidæ have wide ranges, as might be inferred from the fact of between one third and one fourth of the total number inhabiting both the torrid and temperate zones; but it should not be overlooked, that those species, as Balanus tintinnabulum, amphitrite, improvisus, and, in a lesser degree, B. trigonus and Tetraclita radiata, which seem to range over nearly the whole world (excepting the colder seas), are species which are habitually attached to ships, and which could hardly fail to be widely transported. Indeed, it appears to me surprising, that such species as Balanus psittacus and eburneus, which often become attached to vessels, should still be confined, the one to Southern, and the other to Northern America. But some other Cirripedes, which I have never seen attached to vessels, have likewise immense ranges: thus Tetraclita porosa is found in every tropical and warmer sea, and Chthamalus stellatus ranges round the world in the northern hemisphere, and, along the eastern side of America, far south of the equator: Balanus spongicola, and Acasta spongites, extend from the shores of Britain to the Cape of Good Hope: Balanus lævis ranges from Tierra del Fuego to California. I may further remark, that the only two other species of Balanus, and the one Chthamalus, inhabiting Tierra del Fuego, are, also, found on the shores of Peru. But to show the powers of endurance in some species, I may specify the case of Balanus improvisus, which flourishes on the coast of Nova Scotia, amongst the West Indian Islands, in Southern Patagonia, and near Guayaquil. Even more striking is the case of B. crenatus, of which I have seen specimens from latitude 74° 48′ north, from the West Indies, and the Cape of Good Hope! In these two latter localities, however, it seems to be rare, and may have been first transported to them from the shores of Europe, on the bottoms of vessels, to which it sometimes adheres.

The several species of Balanidæ live attached either to coast-rocks, or to objects at various depths, down to, as in the case of Balanus crenatus, 50 fathoms. Balanus balanoides sometimes adheres to rocks or wood so high up, that it is not covered by water during the neap tides. Mr. Thompson has informed me, that he once accidentally kept some specimens of this species out of water for seven days in a warm room, and that they were then quite lively. This species, is very easily killed by brackish water, as are some other species, whilst B. improvisus and eburneus can flourish in it; and at the Falkland Islands, I saw Elminius Kingii attached to rocks at the mouth of a fresh-water brook, so as to be covered by pure water during the ebb of each tide. Sessile cirripedes adhere to all sorts of objects, floating and fixed, animal and vegetable, living and dead, organic and inorganic. Chthamalus is, perhaps, more commonly attached to rocks than are the other genera. Living Mollusca are, I think, the most frequent objects of attachment: Mr. Cuming has remarked to me, that shells covered by an epidermis, as Patella, Haliotis, and Mytilus, are the greatest favorites. Acasta is always imbedded in sponges, or in the sponge-like bark of Isis; Pyrgoma and Creusia in corals; Chelonobia is attached to turtles, and one species to crabs or very smooth shells; Coronula, Tubicinella, and Xenobalanus, are imbedded in the skin of Cetaceans; and Platylepas in that of manatee, turtles, or sea-snakes.

If we attempt, with our present not very imperfect materials, to divide the globe into provinces, according to the amount of difference in their Cirripedial inhabitants, including all orders and families, and disregarding entirely, as I think we ought, all probabilities or conclusions deduced from the distribution of other tribes of animals, we find that the globe may be divided into the four following great provinces and one sub-province. I should premise, that in the following remarks and tables,[78] the species of Lepas, Conchoderma, Chelonobia, Coronula, Platylepas, and Tubicinella, are excluded, owing to their being attached to floating or swimming objects, and being consequently widely and irregularly distributed.

The first, or North Atlantic province, is that of Europe and the eastern shores of North America, from the arctic regions to lat. 30°: the island of Madeira, part of the north-west coast of Africa, and the whole Mediterranean being included. In this province (the above-named genera being excluded) we have 31 species, of which 22 are not found in any other distant quarter of the globe. As some few of these species range into the West Indies, I have not, on this account, excluded them from the 22 peculiar forms. Had I included the West Indies[79] in my first province, the total number of species would have been 42, of which 28 would have been peculiar. The coast of Brazil, even as far south as the Rio Plata, might, also, have been included, for I have not seen from it a single species not included in the above 42 West Indian species. So also, by adding a single species, might the west coast of equatorial Africa. The two coasts of South America and Africa, which face each other within the torrid zone, seem to be remarkably barren in Cirripedes. Europe has several more species than the United States, which is inhabited by only ten species, including even the probably imported Balanus tintinnabulum and amphitrite. Of these ten United States species only two are not found in Europe; and both these two range into the West Indies, and as far as the northern shores of South America, and therefore cannot be considered as peculiar to the United States.