The foregoing remarks on the two threads by which [Proteolepas] is attached, are, independently of their relation to this individual animal, of considerable interest. In my volume on the Lepadidæ, I have stated, after repeated and rigorous examinations (for I was well aware how singular the facts were), that in Conchoderma aurita and in some other genera, the cement-ducts, which entered the pupal antennæ, could be traced till they joined a gland, the coat of which gland was absolutely continuous with the coats of the adjoining and continuous ovarian tubes, of which it was only a modified portion; and what was still more remarkable, that the matter within the gland was continuous with, and differed only from, the cellular matter within the ovarian tubes and cæca (from which ova were in the act of formation), by being more homogeneous and more coherent. Furthermore, I have shown, that in Ibla an ovarian tube, becomes by a very small change, namely, by a double flexure and slight thickening of its coat, converted into a gland, and thus acquires the power of affecting the cellular ovarian matter and changing it into cement. Now, in [Proteolepas], the great ovarian sack replaces the ovarian tubes and cæca; and we here see the very same relations even still more plainly; for the coat of the ovarian sack is indisputably continuous with that investing or forming the two cement-ducts within the two threads; and immediately that the coarse cellular matter, which within the ovarian sack is being converted into ova, enters the upper contracted end of the cement-duct, by some power, we must suppose, inherent in its coat, it is converted into cement, which debouches with all its usual properties through the pupal antennæ. I may venture to reaffirm that nothing could be plainer than this structure, or be in more striking conformity with my previous observations, given in the introduction to the Lepadidæ.
I can hardly express the perplexity which I felt when I first examined [Proteolepas], and when I naturally mistook the mouth for the entire head, for I saw, as I thought, the antennæ in direct connection with the second segment of the body, posteriorly to the mouth! It was quite as monstrous and incredible an inversion of the laws of nature, as those fabulous half-human monsters, with an eye seated in the middle of their stomachs. After a time, I perceived that the following considerations removed all difficulty, and brought [Proteolepas] into the type of other cirripedes.
Firstly: in ordinary cirripedes, the two cement-ducts can be traced up from the cemented antennæ to the glands, formed by a part of the ovarian branching cæca; and the latter can be traced to where they enter, as two simple tubes, the body of the animal, at a medio-dorsal point, a little anteriorly to the prosoma, or second thoracic segment.[155] From what is actually seen in the complemental male of Scalpellum Peronii, and from what may be inferred from the structure of these parts in the pupæ of all cirripedes, there can be no doubt that if the ovarian cæca were in any case not developed, the cement-ducts would enter the body at the spot where the two simple ovarian tubes, which serve to unite the ovarian cæca with the true ovaria, do enter. Now if we look at the drawing (Pl. [25], fig. [7]) of [Proteolepas], we shall see that the cement-ducts enter the body at a medio-dorsal point, a little anteriorly to the second thoracic segment, and therefore in the normal position.
[155] This may be partially seen in the section, fig. [1], of [Balanus], on the same plate ([25]) with the figure of [Proteolepas]; here (bearing in mind that [Balanus] is a much modified form) (z) shows the pupal antennæ, within which, whilst young, the cement-ducts are included, and are directly continuous with the layer of branching ovarian cæca (g), which are prolonged up to the ovaria as a pair of simple tubes (only one being here represented), entering the body above the upper margin of the prosoma (c). The prosoma of [Balanus], I may add, answers to the segment 2 t in fig. [7] of [Proteolepas]; (e) the mouth in [Balanus], of course corresponding with (m) the mouth of [Proteolepas]; the segment 1 c and 2 t of the latter, are in [Balanus] aborted or confluent, at least on the ventral surface; and, lastly, the whole great shell of [Balanus], the sack with its muscles and the branchiæ, and the opercular valves with their muscles, are all represented in [Proteolepas] merely by the outer membrane of the two threads (g), which enter the pupal antennæ!
Secondly: the external membrane of the two threads, investing the two cement-ducts, it should be remembered, is not moulted, and is added to during growth (being lined internally by corium), only round the upper, collar-like edge.
Thirdly: the external covering or carapace of every young cirripede, at the period of its metamorphosis, enters, at its lower end, the cemented antennæ, in the form of two short tubular prolongations, by which alone, at first, the cirripede adheres to the surface of attachment; within these prolongations the cement-ducts are included. I have, moreover, seen instances, as in Conchoderma aurita and in the male of Ibla and [Alcippe], in which these tubular prolongations, lined internally by corium, were increased a little in length, so as to form a trouser-like termination to the peduncle. That the forked extremity should be a little more developed, and so be converted into a pair of short tubular threads, cannot be considered as very improbable.
Fourthly: in the male Ibla the capitulum is so much atrophied that it does not enclose the thorax or mouth, but still an elongated support or peduncle is left. But it would be no very violent assumption to imagine the peduncle, which does not essentially differ from the capitulum, to become likewise rudimental,—to grow smaller and smaller, and shorter and shorter, till the merest remnant was left at the spot where it entered the cemented antennæ. And in the last paragraph it has been shown that it would be no violent assumption to imagine this lower end of the peduncle, where it enters the antennæ, developed into two short thread-like prolongations.
Lastly: it is certain, from the existence of the prehensile antennæ, that [Proteolepas] was developed within a pupa, probably differing in no very essential respect from the pupæ of other cirripedes. Therefore, in accordance with all analogy, we may believe that the position[156] of the young [Proteolepas] (probably much coiled up, with a deep fold close under the mouth) within the pupa, the general form and structure of the latter, and the course of the cement-ducts, did not essentially differ from the imaginary figure given, Pl. [25], fig. [6]. Now, at the period of the metamorphosis, let us imagine that no general covering or carapace was formed, except a small portion on the ventral surface, round the cemented antennæ. Let us further suppose this remnant to be specially developed (as in the case of some cirripedes) into a short trouser-like prolongation, entering the antennæ; and subsequently, in accordance with the almost universal laws of growth in cirripedes, that this portion was never moulted, but continued to be added to, during growth, only at its upper end. By this means we should produce every leading peculiarity of the [Proteolepas bivincta]. As this parasite lives within the sack of another cirripede, and is protected by the capitulum of the latter, we can understand, in accordance with the usual admirable economy of nature, the absence of any general covering for its body. We can now, also, understand the structure and manner of growth of the two threads by which it is bound to its prey; and the connection, at first so strange and perplexing, between the old pupal antennæ and the second segment of the thorax. I am convinced that no other explanation than that here given, will accord with the relations of the several parts and organs of [Proteolepas]. Consequently, I fully believe that we here see an articulate animal in which the whole of the three anterior segments of the head have been, during the act of metamorphosis, absolutely aborted, with the exception of a mere rudiment on the ventral surface, near the anterior end, round the old antennæ, and which rudiment has been specially developed as a covering for the two cement-ducts. As the pupal antennæ are, homologically, the second pair of antennæ, we may further infer that this modified remnant of the carapace, investing the two threads, belongs to the third cephalic segment.