Origin of Vertebrates/Chapter V

I have now given my reasons why I consider that the glossopharyngeal and vagus nerves were originally the nerves belonging to a series of mesosomatic branchial appendages, each of which is still traceable in the respiratory chamber of Ammocœtes, and gives the type-form from which to search for other serially homologous, although it may be specially modified, segments.

As long as the branchial unit consisted of the gill-pouch the segments of the head-region were always referred to such units, hence we find Dohrn and Marshall picturing to themselves the ancestor of vertebrates as possessing a series of branchial pouches right up to the anterior end of the body. Marshall speaks of olfactory organs as branchial sense-organs; Dohrn of the mouth as formed by the coalescence of gill-slits, of the trigeminal nerve as supplying modified branchial segments, etc.; thus a picture of an animal is formed such as never lived on this earth, or could be reasonably imagined to have lived on it. Yet Dohrn's conceptions of the segmentation were sound, his interpretation only was in fault, because he was obliged to express his segments in terms of the gill-pouch unit. Once abandon that point of view and take as the unit a branchial appendage, then immediately we see that in the region in front of the branchiæ we may still have segments homologous to the branchial segments, originally characterized by the presence of appendages, but that such appendages need never have carried branchiæ. The new mouth may have been formed by such appendages, which would express Dohrn's suggestion of its formation by coalesced gill-slits; the olfactory organ may have been the sense-organ belonging to an antennal appendage, which would be what Marshall really meant in calling it a branchial sense-organ.

This simple alteration of the branchiomeric unit from a gill-pouch to an appendage, which may or may not bear branchiæ, immediately sheds a flood of light on the segmentation of the head-region, and brings to harmony the chaos previously existing. Let us, then, follow out its further teachings. Next anteriorly to the glossopharyngeal and vagus nerves comes the facial nerve; a nerve which supplies the hyoid segment, or, rather, according to van Wijhe the two hyoid segments, for embryologically there is evidence of two segments. As already mentioned, the facial nerve is usually included in the trigeminal or pro-otic group of nerves, the opisthotic group being confined to the glossopharyngeal and vagus. This inclusion of the facial nerve into the pro-otic group of nerves forms one of the main reasons why this group has been supposed to have originally supplied gill-pouch segments, for the hyoid segment is clearly associated with branchiæ.

When, however, we examine Ammocœtes (cf. Figs. 63 and 64) it is clear that the foremost of the segments forming the respiratory chamber, which must be classed with the rest of the mesosomatic or opisthotic segments, is that supplied by the facial nerves.

An examination of this respiratory chamber shows clearly that there are six pairs of branchial appendages or diaphragms, which are all exactly similar to each other. These are those already considered, the foremost of which are supplied by the IXth or glossopharyngeal nerves. Immediately anterior to this glossopharyngeal segment is seen in the figures the segment supplied by the VIIth or facial nerves. It is so much like the segments belonging to the glossopharyngeal and vagus nerves as to make it certain that we are dealing here with a branchial segment, composed of a pair of branchial appendages similar to those in the other cases, except that the cartilaginous bar is here replaced by a bar of muco-cartilage and the branchiæ are confined to the posterior part of each appendage. The anterior portion is, as is seen in Fig. 74, largely occupied by blood-spaces, but in addition carries the ciliated groove (ps. br.) called by Dohrn 'pseudo-branchiale Rinne.' This groove leads directly into the thyroid gland, which is a large bilateral organ situated in the middle line, as seen in Fig. 80 and Fig. 85. As shown by Miss Alcock, the facial nerve supplies this thyroid gland, as well as the posterior hyoid branchial segment, and, as pointed out by Dohrn, there is every reason to consider this thyroid gland as indicative of a separate segment, especially when van Wijhe's statement that the hyoid segment is in reality double is taken into account.



Somatic muscles coloured red. Branchial and visceral muscles coloured blue. Tubular constrictor muscles distinguished from striated constrictor muscles by simple hatching. Tent., tentacles; ''Tent. m.c., muco-cartilage of tentacles; Vel. m.c., muco-cartilage of the velum; Hy. m.c., muco-cartilage of the hyoid segment; Ps. br., pseudo-branchial groove; Br. cart., branchial cartilages; Sp., space between somatic and splanchnic muscles; Th. op., orifice of thyroid; H.'', heart.

The evidence, then, of Ammocœtes points directly to this conclusion: The facial nerves represent the foremost of the mesosomatic group of nerves, and supply two segments, which have amalgamated with each other. The most posterior of these, the hyoid segment, is a branchial segment of the same character as those supplied by the vagus and glossopharyngeal nerves; represents, therefore, the foremost pair of branchial appendages. The anterior or thyroid segment, on the other hand, differs from the rest in that, instead of branchiæ, it carries the thyroid gland with its two ciliated grooves. If this segment, which is the foremost of the mesosomatic segments, also indicates a pair of appendages which carry the thyroid gland instead of branchiæ, then it follows that this pair of appendages has joined together in the mid-line ventrally and thus formed a single median organ—the thyroid gland. If, then, we find that the foremost of the mesosomatic appendages in the Palæostraca was really composed of two pairs of appendages, of which the most posterior carried branchiæ, while the anterior pair had amalgamated in the mid-line ventrally, and carried some special organ instead of branchiæ, then the accumulation of coincidences is becoming so strong as to amount to proof of the correctness of our line of investigation.

What, then, is the nature of the foremost pair of mesosomatic appendages in Limulus. They differ from the rest of the mesosomatic appendages in that they do not carry branchiæ, and instead of being separate are joined together in the mid-line ventrally to form a single terminal plate-like appendage known as the operculum. On its posterior surface the operculum carries the genital duct on each side.

So also in the scorpion group, the operculum is always found and always carries the genital ducts.

A survey of the nature of the opercular appendage demonstrates the existence of three different types—

1. That of Limulus, in which the operculum is free, and carries only the terminations of the genital ducts. In this type the duct on each side opens to the exterior separately (Fig. 75).

2. The type of Scorpio, Androctonus, Buthus, etc., in which the operculum is not free, but forms part of the ventral surface of the body-wall, but, like Limulus, carries only the terminations of the genital ducts. In this type the duct on each side terminates in a common chamber (vagina or uterus), which communicates with the exterior by a single external median opening. This common chamber, or uterus (Ut.), extends the whole breadth of the operculum (as seen in Fig. 76), and is limited to that segment.

3. The type of Thelyphonus, Hypoctonus, Phrynus, and other members of the Pedipalpi, in which the operculum forms a part of the ventral surface of the body wall, but no longer covers only the termination of the genital apparatus. It really consists of two parts, a median anterior, which covers the terminal genital apparatus, and a lateral posterior, which covers the first pair of gills, or lung-books, as they are called. In this type (Fig. 77) the genital ducts terminate in a common chamber or uterus, the nature of which will be further considered.

As has been pointed out by Blanchard, the terminal genital organs of the scorpions and the Pedipalpi vary considerably in the different genera, especially the male genital organs. The general type of structure is the same, and consists in both male and female of vasa deferentia, which come together to form a common chamber before the actual opening to the exterior. This common chamber has been called in the female scorpion the vagina, or in Thelyphonus the uterus. I shall use the latter term, in accordance with Tarnani's work, and the corresponding chamber in the male will be the uterus masculinus.

A considerable discussion has taken place about the method of action of the external genital organs in the members of the scorpion tribe, into which it is hardly necessary to enter here. The evidence points to the conclusion that in all these forms the operculum covers a median single chamber or uterus, into which the genital ducts open on each side, the main channels of emission being provided with a massive chitinous internal framework. We may feel certain that in the old extinct sea-scorpions, Eurypterus, etc., a similar arrangement existed, and that therefore in them also the median portion of the operculum covered a median chamber or uterus composed of the amalgamation of the terminations of the two genital ducts, which were originally separate, as in Limulus.



Opercular segment is marked out by thick black line. Ut. Masc., uterus masculinus; ''Int. Op., internal opening of uterus into genital chamber; Ext. Op., common external opening to genital chamber (Gen. Ch.'') and pulmonary chamber.

The observations of Schmidt, Zittel, and others show that the operculum in the old extinct sea-scorpions, Eurypterus, Pterygotus, etc., belonged to the type of Thelyphonus, rather than to that of Limulus or Scorpio. In Fig. 78 I give a picture from Schmidt of the ventral aspect of Eurypterus, and by the side of it a picture of the isolated operculum. Schmidt considers that there were five branchiæ-bearing segments constituting the mesosoma, the foremost of which formed the operculum. Such operculum is often found isolated, and is clearly composed of two lateral appendages fused together in the middle line, of such a nature as to form a median elongated tongue, which lies between and separates the first three pairs of branchial segments. This median tongue, together with the anterior and median portion of the operculum, concealed, in all probability, according to Schmidt, the terminal parts of the genital organs, just as the median part of the operculum in Phrynus and Thelyphonus conceals the complicated terminal portions of the genital organs. The posterior part of the operculum, like that of Phrynus and Thelyphonus, carried the first pair of branchiæ, so Schmidt thinks from the evidence of markings on some specimens.



The segments and appendages on the right are numbered in correspondence with the cranial system of lateral nerve-roots as found in vertebrates. M., metastoma. The surface ornamentation is represented on the first segment posterior to the branchial segments. The opercular appendage is marked out by dots.

Apparently an opercular appendage of this kind is in reality the result of a fusion of the genital operculum with the first branchial appendage in forms such as the scorpion; for, in order that the tergal plates may correspond in number with the sternal in Eurypterus, etc., it is necessary to consider that the operculum is composed of two sternites joined together. Similarly in Thelyphonus, Phrynus, etc., this numerical correspondence is only observed if the operculum is looked upon as double.

A restoration of the mesosomatic region of Eurypterus, viewed from the internal surface, might be represented by Fig. 79, in which the thick line represents the outline of the opercular segment, and the fainter lines the succeeding branchial segments. The middle and anterior part of the opercular segment carried the terminations of the genital organs; these I have represented, in accordance with our knowledge of the nature of these organs in the present-day scorpions, as a median elongated uterus, bilaterally formed, from which the genital ducts passed, probably as in Limulus, towards a mass of generative gland in the cephalic region, and not as in Scorpio or Thelyphonus, tailwards to the abdominal region.



The opercular segment is marked out by the thick black line. The segments II.-VI. bear branchiæ, and segment I. is supposed in the male to carry the uterus masculinus (Ut. Masc.) and the genital ducts.

It is possible that in Holm's representation of Eurypterus, Fig. 104, the genital duct on each side is indicated.

If we compare this mesosomatic region of Eurypterus with that of Ammocœtes, the resemblance is most striking, and gives a meaning to the facial nerve which is in absolute accordance with the interpretation already given of the glossopharyngeal and vagus nerves. In both cases the foremost respiratory or mesosomatic segment is double, the posterior lateral part alone bearing the branchiæ, while the median and anterior part bore in the one animal the uterus and genital ducts, in the other the thyroid gland and ciliated grooves. We are driven, therefore, to the conclusion that this extraordinary and unique organ, the so-called thyroid gland of Ammocœtes, which exists only in the larval condition and is got rid of as soon as the adult sexual organs are formed, shows the very form and position of the uterus of this invertebrate ancestor of Ammocœtes. What, then, is the nature of the thyroid gland in Ammocœtes?

Throughout the vertebrate kingdom it is possible to compare the thyroid gland of one group of animals with that of another without coming across any very marked difference of structure right down to and including Petromyzon. When, however, we examine Ammocœtes, we find that the thyroid has suddenly become an organ of much more complicated structure, covering a much larger space, and bearing no resemblance to the thyroid glands of the higher forms. At transformation the thyroid of Ammocœtes is largely destroyed, and what remains of the gland in Petromyzon becomes limited to a few follicles resembling those of other fishes. The structure and position of this gland in Ammocœtes is so well known that it is unnecessary to describe it in detail. For the purpose, however, of making my points clear, I give in Fig. 80 the position and appearance of the thyroid gland (Th.) when the skin and underlying laminated layer has been removed by the action of hypochlorite of soda. On the one side the ventral somatic muscles have been removed to show the branchial cartilaginous basket-work.



The series of transverse sections in Fig. 81 represents the nature of the organ at different levels in front of and behind the opening into the respiratory chamber; and in Fig. 82 I have sketched the appearance of the whole gland, viewed so as to show its opening into the respiratory chamber and its posterior curled-up termination.



Sections 1 and 2 are anterior to the thyroid opening, ''Th. o.''; sections 3, 4, and 5 are through the thyroid opening; and section 6 is posterior to the thyroid opening before the commencement of the curled portion.

The series of transverse sections (1-6, Fig. 81) show that we are dealing here with a central glandular chamber, C (Fig. 81 (6) and Fig. 82), which opens by the thyroid duct (Th. o.) into the pharyngeal chamber, and is curled upon itself in its more posterior part. This central chamber divides, anteriorly to the thyroid orifice, into two portions, A, A′ (Fig. 82), giving origin to two tubes, B, B′, which lie close alongside of, and extend further back than, the posterior limit of the curled portion of the central chamber, C. The structure of the central chamber, C, and, therefore, of the separate coils, is given in both Schneider's and Dohrn's pictures, and is represented in Fig. 81 (6), which shows the peculiar arrangement and character of the glandular cells typical of this organ, and also the nature of the central cavity, with the arrangement of the ciliated epithelium. The structure of each of the lateral tubes, B, is different from that of the central chamber, in that only half the central chamber is present in them, as is seen by the comparison of the tube B with the tube C in Fig. 81 (5 and 6), so that we may look upon the central chamber, C, as formed of two tubes, similar in structure to the tubes B, which have come together to form a single chamber by the partial absorption of their walls, the remains of the wall being still visible as the septum, which partially divides the chamber, C, into halves.

In the walls of each of these tubes is situated a continuous glandular line, the structure of the glandular elements being specially characterized by the length of the cells, by the large spherical nucleus situated at the very base of each cell, and by the way in which the cells form a wedge-shaped group, the thin points of all the wedge-shaped cells coming together so as to form a continuous line along the chamber wall. This free termination of the cells of the gland in the lumen of the chamber constitutes the whole method for the secretion of the gland; there is no duct, no alveolus, nothing but this free termination of the cells.

Moreover, sections through the portion A, A′ (Fig. 82) show that here, as in the central chamber, C, four of these glandular lines open into a common chamber, but they are not the same four as in the case of the central chamber, for if we name these glandular lines on the left side a b, a′ b′ (Fig. 81), and on the right side c d, c′ d′, then the central chamber has opening into it the glands a b, c d, while the chambers of A and A′ have opening into them respectively a b, a′ b′, and c d, c′ d′. Further, the same series of sections shows that the glands a and b are continuous with the glands a′ and b′ respectively across the apex of A, and similarly on the other side, so that the two glandular rows a b are continuous with the two glandular rows a′ b′, and we see that the cavity of the portion A or A′ is formed by the bending over of the tube or horn, B or B′, with the partial absorption of the septum so formed between the tube and its bent-over part. If, then, we uncoil the curled-up part of C, and separate the portion, B, on each side from the chamber, C, we see that the so-called thyroid of Ammocœtes may be represented as in Fig. 83, i.e. it consists of a long, common chamber, C, which, for reasons apparent afterwards, I will call the palæo-hysteron, which opens, by means of a large orifice, into the respiratory or pharyngeal chamber. The anterior end of this chamber terminates in two tubes, or horns, B, B′, the structure of which shows that the median chamber, C, is the result of the amalgamation of two such tubes, and consequently in this chamber, or palæo-hysteron, the glandular lines are symmetrically situated on each side.



C, central chamber; A, A′, anterior extremity; B, B′, posterior extremity; ''Th. o., thyroid opening into respiratory chamber; Ps. br., Ps. br'.'', ciliated grooves, Dohrn's pseudo-branchial grooves.

B, B′,

Any explanation, then, of the thyroid gland of Ammocœtes, must take into account the clear evidence that it is composed of two tubes, which have in part fused together to form an elongated central chamber, in part remain as horns to that chamber, and that in its walls there exist lines of gland-cells of a striking and characteristic nature.

Further, this central chamber, with its horns, is not a closed chamber, but is in communication with the pharyngeal or respiratory chamber by three ways. In the first place, the central chamber, as is well known, opens into the respiratory chamber by a funnel-shaped opening—the so-called thyroid duct (Th. o.). In the second place, there exist two ciliated grooves (Ps. br., Ps. br′.), the pseudo-branchial grooves of Dohrn, which have direct communication with the thyroid chamber. The manner in which these grooves communicate with the thyroid chamber has never, to my knowledge, been described previously to my description in the Journal of Physiology and Anatomy; it is very instructive, for, as I have there shown, each groove enters into the corresponding lateral horn, so that, in reality, there are three openings into the thyroid chamber or palæo-hysteron—a median opening into the central chamber, and a separate opening into each lateral horn.

The system of ciliated grooves on the inner ventral surface of the respiratory chamber of Ammocœtes was originally described by Schneider as consisting of a single median groove, which extends from the opening of the thyroid to the posterior extremity of the branchial chamber, and a pair of grooves, or semi-canals, which, starting from the region of the thyroid orifice, run headwards and diverge from each other, becoming more and more lateral, and more and more dorsal, till they come together in the mid-dorsal pharyngeal line below the auditory capsules. The latter are the pseudo-branchial grooves of Dohrn, of which I have already spoken. Schneider looked upon the whole of this system as a single system, for he speaks of "a ciliated groove, which extends from the orifice of the stomach (i.e. anterior intestine) to the orifice of the thyroid, then divides into two, and runs forward right and left of the median ridge, etc." Dohrn rightly separates the median ciliated groove posterior to the thyroid orifice (seen in Fig. 81 (6)) from the paired pseudo-branchial grooves; the former is a shallow depression which opens into the rim of the thyroid orifice, while the latter has a much more intimate connection with the thyroid gland itself.

A series of sections, such as is given in Fig. 81, shows the relation of this pair of ciliated grooves to the thyroid better than any elaborate description. In the first place, it is clear that they remain separate up to their termination—they do not join in the middle line to open into the thyroid duct; in the second place, they are separate from the thyroid orifice—they do not terminate at the rim of the orifice, as is the case with the median groove just mentioned, but continue on each side on the wall of the thyroid duct (Fig. 81 (2)), gradually moving further and further away from the actual opening of the duct into the pharyngeal chamber. During the whole of their course on the wall of the funnel-shaped duct they retain the character of grooves, and are therefore open to the lumen of the duct. The direction of the groove (Ps. br.) shifts as it passes deeper and deeper towards the thyroid, until at last, as seen in Fig. 81 (3 and 4), it is continuous with the narrow diverticulum of the turned-down single part of the thyroid (B), or turned-down horn, as I have called it. In other words, the median chamber opens into the pharyngeal or respiratory chamber by a single large, funnel-shaped opening, and, in addition, the two ciliated grooves terminate in the lateral horns on each side, and only indirectly into the central chamber, owing to their being semi-canals, and not complete canals. If they were originally canals, and not grooves, then the thyroid of Ammocœtes would be derived from an organ composed of a large, common glandular chamber, which opened into the respiratory chamber by means of an extensive median orifice, and possessed anteriorly two horns, from each of which a canal or duct passed headwards to terminate somewhere in the region of the auditory capsule.

Dohrn has pointed out that a somewhat similar structure and topographical arrangement is found in Amphioxus and the Tunicata, the gland-cells being here arranged along the hypobranchial groove to form the endostyle and not shut off to form a closed organ, as in the thyroid of Ammocœtes. Dohrn concludes, in my opinion rightly, that the endostyle in the Tunicata and in Amphioxus represents the remnants of the more elaborate organ in Ammocœtes, and that, therefore, in order to explain the meaning of these organs in the former animals, we must first find out their meaning in Ammocœtes. Dohrn, however, goes further than this; for just as he considers Amphioxus and the Tunicata to have arisen by degeneration from an Ammocœtes-like form, so he considers Ammocœtes to have arisen from a degenerated Selachian; therefore, in order to be logical, he ought to show that the thyroid of Ammocœtes is an intermediate downward step between the thyroid of Selachians and that of Amphioxus and the Tunicates. Here, it seems to me, his argument utterly breaks down; it is so clear that the thyroid of Petromyzon links on to that of the higher fishes, and that the Ammocœtes thyroid is so immeasurably more complicated and elaborate a structure than is that of Petromyzon, as to make it impossible to believe that the Ammocœtes thyroid has been derived by a process of degeneration from that of the Selachian. On the contrary, the manner in which it is eaten up at transformation and absolutely disappears in its original form is, like the other instances mentioned, strong evidence that we are dealing here with an ancestral organ, which is confined to the larval form, and disappears when the change to the higher adult condition takes place. Dohrn's evidence, then, points strongly to the conclusion that the starting-point of the thyroid gland in the vertebrate series is to be found in the thyroid of Ammocœtes, which has given rise, on the one hand, to the endostyle of Amphioxus and the Tunicata, and on the other, to the thyroid gland of Petromyzon and the rest of the Vertebrata.

The evidence which I have just given of the intimate connection of the two pseudo-branchial grooves with the thyroid chamber shows, to my mind, clearly that Dohrn is right in supposing that morphologically these two grooves and the thyroid must be considered together. His explanation is that the whole system represents a modified pair of branchial segments distinct from those belonging to the VIIth and IXth nerves. The cavity of the thyroid and the pseudo-branchial grooves are, therefore, according to him, the remains of the gill-pouches of this fused pair of branchial segments, which no longer open to the surface, and the glandular tissue of the thyroid is derived from the modified gill-epithelium. This view of Dohrn's, which he has urged most strongly in various papers, is, I think, right in so far as the separateness of the thyroid segment is concerned, but is not right, and is not proven, in so far as concerns the view that the thyroid gland is a modified pair of gills.

We may distinctly, on my view, look upon the thyroid segment, with its ciliated grooves and its covering plate of muco-cartilage, as a distinct paired segment, homologous with the branchial segments, without any necessity of deriving the thyroid gland from a pair of gills,

The evidence that such a median segment has been interpolated ventrally between the foremost pairs of branchial segments is remarkably clear, for the limits ventrally of the branchial segments are marked out on each side by the ventral border of the cartilaginous basket-work; and it is well known, as seen in Fig. 80, that whereas this cartilaginous framework on the two sides meets together in the middle ventral line in the posterior branchial region, it diverges in the anterior region so as to form a tongue-shaped space between the branchial segments on the two sides. This space is covered over with a plate of muco-cartilage which bears on its inner surface the thyroid gland.



In addition to this evidence that we are dealing here with a ventral tongue-like segment belonging to the facial nerve which is interpolated between the foremost branchial segments, we find the most striking fact that at transformation the whole of this muco-cartilaginous plate disappears, the remarkable thyroid gland of the Ammocœtes is eaten up, and nothing is left except a small, totally different glandular mass; and now the cartilaginous basket-work meets together in the middle line in this region as well as in the more posterior region. In other words, the striking characteristic of transformation here is the destruction of this interpolated segment, and the resulting necessary drawing together ventrally of the branchial segments on each side.



VII. 1, thyroid part of segment; VII. 2, hyoid or branchial part; 3-9, succeeding branchial segments belonging to IXth and Xth nerves; V, the velar folds; Ps. br., Dohrn's pseudo-branchial groove; ''Th. o., thyroid opening; C'', curled portion of thyroid.

Moreover, another most instructive piece of evidence pointing in the same direction is afforded by the behaviour of the ventral epithelial pits, as determined by Miss Alcock. Although there is no indication on the ventral surface of the skin of any difference between the anterior and posterior portions of the respiratory region, yet when the ventral rows of the epithelial pits supplied by each branchial nerve are mapped out, we see how the most anterior ones diverge more and more from the mid-ventral line, following out exactly the limits of the underlying muco-cartilaginous thyroid plate (Fig. 84).

The whole evidence strongly leads to the conclusion that the thyroid portion of the facial segment was inserted as a median tongue between the foremost branchial segments on each side, and that, therefore, the whole facial segment, consisting as it does of a thyroid part and a hyoid or branchial part, may be represented as in Fig. 85, which is obtained by splitting an Ammocœtes longitudinally along the mid-dorsal line, so as to open out the pharyngeal chamber and expose the whole internal surface. The facial segment is marked out by shading lines, the glosso-pharyngeal and vagus segments and the last of the trigeminal segments being indicated faintly. The position of the thyroid gland is indicated by oblique lines, C being the curled portion.

Seeing how striking is the arrangement and the structure of the glandular tissue of this thyroid, how large the organ is and how absolutely it is confined to Ammocœtes, disappearing entirely as such at transformation, we may feel perfectly certain that a corresponding, probably very similar, organ existed in the invertebrate ancestor of the vertebrate; for the transformation process consists essentially of the discarding of invertebrate characteristics and the putting on of more vertebrate characters; also, so elaborate an organ cannot possibly have been evolved as a larval adaptation during the life of Ammocœtes. We may therefore assert with considerable confidence that the thyroid gland was the palæo-hysteron, and was derived from the uterus of the ancient palæostracan forms. If, then, it be found that a glandular organ of this very peculiar structure and arrangement is characteristic of the uterus of any living member of the scorpion group, then the confidence of this assertion is greatly increased.

In Limulus, as already stated, the genital ducts open separately on each side of the operculum, and do not combine to form a uterus; I have examined them and was unable to find any glandular structure at all resembling that of the thyroid gland of Ammocœtes. I then turned my attention to the organs of the scorpion, in which the two ducts have fused to form a single uterus.



I there found that both in the male and in the female the genital ducts on each side terminate in a common chamber or uterus, which underlies the whole length of the operculum, and opens to the exterior in the middle line, as shown in Fig. 76. In transverse section, this uterus has the appearance shown in Fig. 86, i.e. it is a large tube, evidently expansible, lined with a chitinous layer and epithelial cells belonging to the chitinogenous layer, except in two symmetrical places, where the uniformity of the uterine wall is interrupted by two large, remarkable glandular structures. The structure of these glands is better shown by means of sagittal sections. They are composed of very long, wedge-shaped cells, each of which possesses a large, round nucleus at the basal end of the cell (Fig. 87). These cells are arranged in bundles of about eight to ten, which are separated from each other by connective tissue, the apex of each conical bundle being directed into the cavity of the uterus; where this brush-like termination of the cells reaches the surface, the chitinous layer is absent, so that this layer is, on surface view, seen (Fig. 88 (b)) to be pitted with round holes over that part of the internal surface of the uterus where these glands are situated. Each of these holes represents the termination of one of these cone-shaped wedges of cells. If the section is cut across at right angles to the axis of these cones, then its appearance is represented in Fig. 88 (a), and shows well the arrangement of the blocks of cells, separated from each other by connective tissue. When the section passes through the basal part of the cones, and only in that case, then the nuclei of the cells appear, often in considerable numbers in one section, as is seen in Fig. 89. In Fig. 88 the section shows at b the holes in the chitin in which the cones terminate, and then a series of layers of sections through the cones further and further away from their apices.

These conical groups of long cells, represented in Fig. 87, form on each side of the uterus a gland, which is continuous along its whole length, and thus forms a line of secreting surface on each side, just as in the corresponding arrangement of the glandular structures in the thyroid of Ammocœtes. This uterus and glandular arrangement is found in both sexes; the gland is, however, more developed in the male than in the female scorpion.



The resemblance between the structure of the thyroid of Ammocœtes and the uterus of the scorpion is most striking, except in two respects, viz. the nature of the lining of the non-glandular part of the cavity—in the one case ciliated, in the other chitinous—and the place of exit of the cavity, the thyroid of Ammocœtes opening into the respiratory chamber, while the uterus of Scorpio opens direct to the exterior.



With respect to the first difference, the same difficulty is met with in the comparison of the ciliated lining of the tube in the central nervous system of vertebrates with the chitinous lining of the intestine in the arthropod. Such a difference does not seem to me either unlikely or unreasonable, seeing that cilia are found instead of chitin in the intestine of the primitive arthropod Peripatus. Also the worm-like ancestors of the arthropods almost certainly possessed a ciliated intestine. Finally, the researches of Hardy and McDougall on the intestine of Daphnia point directly to the presence of a ciliated rather than a chitinous epithelial lining of the intestine in this animal—all evidence pointing to the probability that in the ancient arthropod forms, derived as they were from the annelids, the intestine was originally ciliated and not chitinous. It is from such forms that I suppose vertebrates to have sprung, and not from forms like the living king-crabs, scorpions, Apus, Branchipus, etc. I only use them as illustrations, because they are the only living representatives of the great archaic group, from which the Crustacea, Arachnida, and Vertebrata all took origin.

The second difference is more important, and is at first sight fatal to any comparison between the two organs. How is it possible to compare the uterus of the scorpion, which opens on the surface by an external genital opening, with the thyroid of Ammocœtes, which opens by an internal opening into the respiratory chamber? However close may be the histological resemblance of structure in the two cases, surely such a difference is too great to be accounted for.

It is, however, to be remembered that the operculum of Scorpio covers only the terminal genital apparatus, and does not, therefore, resemble the operculum of the presumed ancestor of Ammocœtes, which, as already argued, must have resembled the operculum of Thelyphonus with its conjoint branchial and genital apparatus, rather than that of Scorpio. Before, therefore, making too sure of the insuperable character of this difficulty, we must examine the uterus of the Pedipalpi, and see the nature of its opening.

The nature of the terminal genital organs in Thelyphonus has been described to some extent by Blanchard, and more recently by Tarnani. The ducts of the generative organs terminate, according to the latter observer, in the large uterus, which is found both in the male and female; he describes the walls of the uterus in the female as formed of elongated glandular epithelium, with a strongly-developed porous, chitinized intima. In the male, he says that the epithelium of the uterus masculinus and its processes is extraordinarily elongated, the chitin covering being thick. In these animals, then, the common chamber or uterus into which the genital ducts empty, which, like the corresponding chamber in the scorpion, occupies the middle region of the operculum, is a large and conspicuous organ. Further, and this is a most striking fact, the uterus masculinus does not open direct to the exterior, but into the genital cavity, "which lies above the uterus, so that the latter is situated between the lower wall of the genital cavity and the outer integument." The opening, therefore, of the uterus is not external but internal, into the large internal space known as the genital cavity. The arrangement is shown in Fig. 91, taken from Tarnani's paper, which represents a diagrammatic sagittal section through the exit of the male genital duct. Yet another most striking fact is described by Tarnani. This genital cavity is continuous with the pulmonary or gill cavities on each side, so that instead of a single opening for the genital products and one on each side for each gill-pouch, as would be the case if the arrangement was of the same kind as in the scorpion, there is a single large chamber, the genital chamber, common to both respiratory and genital organs.

(From .)

The thick line is the operculum, composed of two segments, I. and II. Ut. Masc., uterus masculinus; Gen. Ch., genital chamber; ''Int. Op., internal opening; Ext. Op.'', external opening common to the genital and respiratory organs.

This genital chamber, according to Tarnani, opens to the exterior by a single median opening between the operculum and the succeeding segment; similarly, a communication from side to side exists between the second pair of gill-pouches. I have been able to examine Hypoctonus formosus and Thelyphonus caudatus, and in both cases, in both male and female, the opening to the exterior of the common chamber for respiration and for the genital products was not a single opening, as described by Tarnani in Thelyphonus asperatus, but on each side of the middle line, a round orifice closed by a lid, like the nest of the trapdoor spider, led into the common genital chamber (Gen. Ch.) into which both uterus and gills opened. In Fig. 77 I have endeavoured to represent the arrangement of the genital and respiratory organs in the male Thelyphonus according to Tarnani's and my own observations.

If we may take Thelyphonus as a sample of the arrangement in those scorpions in which the operculum was fused with the first branchial appendage, among which must be included the old sea-scorpions, then it is most significant that their uterus should open internally into a cavity which was continuous with the respiratory cavity. Thus not only the structure of the gland, but also the arrangement of the internal opening into the respiratory, or, as it became later, the pharyngeal cavity, is in accordance with the suggestion that the thyroid of Ammocœtes represents the uterus of the extinct Eurypterus-like ancestor.

Into this uterus the products of the generative organs were poured by means of the vasa deferentia, so that there was not a single median opening or duct in connection with it, but also two side openings, the terminations of the vasa deferentia. These are described by Tarnani in Thelyphonus as opening into the two horns of the uterus, which thus shows its bilateral character, although the body of the organ is median and single; these ducts then pass within the body of the animal, dorsal to the uterus, towards the testes or ovaries as the case may be, organs which are situated in these animals, as in other scorpions, in the abdomen, so that the direction of the ducts from the generative glands to the uterus is headwards. If, however, we examine the condition of affairs in Limulus, we find that the main mass of the generative material is cephalic, forming with the liver that dense glandular mass which is packed round the supraœsophageal and prosomatic ganglia, and round the stomach and muscles of the head-region. From this cephalic region the duct passes out on each side at the junction of the prosomatic and mesosomatic carapace to open separately on the posterior surface of the operculum, near the middle line, as is indicated in Fig. 75.

We have, therefore, two distinct possible positions for the genital ducts among the group of extinct scorpion-like animals, the one from the cephalic region to the operculum, and the other from the abdominal region to the operculum.

The whole argument, so far, has in every case ended with the conclusion that the original scorpion-like form with which I have been comparing Ammocœtes resembled in many respects Limulus rather than the present-day scorpions, and therefore in the case also of the generative organs, with which the thyroid gland or palæo-hysteron was in connection, it is more probable that they were cephalic in position rather than abdominal. If this were so, then the duct on each side, starting from the median ventral uterus, would take a lateral and dorsal course to reach the huge mass of generative gland lying within the prosomatic carapace, just as I have represented in the figure of Eurypterus (Fig. 79), a course which would take much the same direction as the ciliated groove in Ammocœtes.

We ought, therefore, on this supposition, to expect to find the remains of the invertebrate generative tissue, the ducts of which terminated in the thyroid, in the head-region, and not in the abdomen.

Upon removal of the prosomatic carapace of Limulus, a large brownish glandular-looking mass is seen, in which, if it happens to be a female, masses of ova are very conspicuous. This mass is composed of two separate glands, the generative glands and the hepatico-pancreatic glands—the so-called liver—and surrounds closely the central nervous system and the alimentary canal. From the generative glands proceed the genital ducts to terminate on the posterior surface of the operculum. From the liver ducts pass to the pyloric end of the cephalic stomach, and carry the fluid by means of which the food is digested, for, in all these animals, the active digesting juices are formed in the so-called liver, and not in the cells of the stomach or intestine.

It is a very striking fact that the brain of Ammocœtes is much too small for the brain-case, and that the space between brain and brain-case is filled up with a very peculiar glandular-looking tissue, which is found in Ammocœtes and not elsewhere. Further, it is also striking that in the brain of Ammocœtes there should still exist the remains of a tube extending from the IVth ventricle to the surface at the conus post-commissuralis, which can actually be traced right into this tissue on the outside of the brain (see Fig. 13, a-e, Pl. XXVI., in my paper in the Quarterly Journal of Microscopical Science). This, in my opinion, is the last remnant of one of the old liver-ducts which extended from the original stomach and intestine into the cephalic liver-mass. This glandular-looking material is shown surrounding the pineal eye and its nerve, in Fig. 31, also in Fig. 22, and separately in Fig. 92. It is composed of large cells, with a badly staining nucleus, closely packed together with lines of pigment here and there between the cells; this pigment is especially congregated at the spot where the so-called liver-duct loses itself in this tissue. The protoplasm in these large cells does not stain well, and with osmic acid gives no sign of fat, so that Ahlborn's description of this tissue as a peculiar arachnoideal fat-tissue is not true; peculiar it certainly is, but fatty it is not.



This tissue has been largely described as a peculiar kind of connective tissue, which is there as packing material, for the purpose of steadying a brain too small for its case. On the face of it such an explanation is unscientific; certainly for all those who really believe in evolution, it is out of the question to suppose that a brain-case has been laid down in the first instance too large for the brain, in order to provide room for a subsequent increase of brain; just as it is out of the question to suppose that the nervous system was laid down originally as an epithelial tube in order to provide for the further development of the nervous system by the conversion of more and more of that tube into nervous matter. Yet this latter proposition has been seriously put forward by professed believers in evolution and in natural selection.

This tissue bears no resemblance whatever to any form of connective tissue, either fatty or otherwise. By every test this tissue tells as plainly as possible that it is a vestige of some former organ, presumably glandular, which existed in that position; that it is not there as packing material because the brain happened to be too small for its case, but that, on the contrary, the brain is too small for its case, because the case, when it was formed, included this organ as well as the brain; in other words, this tissue is there because it is the remnant of the great glandular mass which so closely surrounds the brain and alimentary canal in animals such as Limulus. In my paper in the Quarterly Journal of Microscopical Science, in which I was comparing the tube of the vertebrate nervous system with the alimentary canal of the invertebrate, I spoke of this tissue as being the remnant of the invertebrate liver. At the same time the whole point of my argument was that the glandular material surrounding the brain of Limulus was made up of two glands—liver and generative gland—so that this tissue might be the remnant of either one or the other, or both. All I desired, at that time, was to point out the glandular appearance of this so-called packing tissue, which surrounded the brain-region of Ammocœtes, in connection with the fact that the brain and alimentary canal of Limulus were closely surrounded with a glandular mass composed partly of liver, partly of the generative gland. At present, I think these large cells found round the brain in Ammocœtes are much more likely to be the remnant of the generative gland than of the liver; the size of the cells and their arrangement recalls Owen's picture of the generative gland in Limulus, and seeing how important all generative glands are in their capacity of internal secreting glands, apart entirely from the extrusion of the ripe generative products, and how unimportant is an hepato-pancreas when the alimentary canal is closed, it is much more likely that of the two glands the former would persist longer than the latter. It may be that all that is left of the old hepato-pancreas consists of the pigment so markedly found in between these cells, especially at the place where the old liver-duct reaches the surface of the brain; just as the only remnant of the two pineal eyes in the higher vertebrates is the remains of the pigment, known as brain-sand, which still exists in the pineal gland of even the highest vertebrate. This, however, is a mere speculation of no importance. What is important is the recognition of this tissue round the brain as the remnant of the glandular mass round the brain of animals such as Limulus. Still further confirmation of the truth of this comparison will be given when the origin of the auditory organ comes up for discussion.

I conclude, therefore, from the evidence of Ammocœtes, that the generative glands in the ancestral form were situated largely in the cephalic region, and suggest that the course and direction of the ciliated pseudo-branchial grooves on each side indicate the direction of the original opercular ducts by which the generative products were conveyed to the uterine chamber, i.e. to the chamber of the thyroid gland, and thence to the common genital and respiratory cavity, and so to the exterior.

It is easy to picture the sequence of events. First, the generative glands, chiefly confined to the cephalic region, communicating with the exterior by separate ducts on the inner surface of the operculum as in Limulus. Then, in connection with the viviparous habit, these two oviducts fused together to form a single chamber, covered by the operculum, which opened out to the exterior by a single opening as in Scorpio: or, in forms such as Eurypterus, in which the operculum had amalgamated with the first branchial appendage and possessed a long, tongue-like ventral projection, the amalgamated ducts formed a long uterine chamber which opened internally into the genital chamber—a chamber which, as in Thelyphonus, was common with that of the two gill-chambers, while at the same time the genital ducts from the cephalic generative material opened into two uterine horns which arose from the anterior part of the uterus, as in Thelyphonus.

Such an arrangement would lead directly to the condition found in Ammocœtes, if the generative material around the brain lost its function, owing to a new exit for generative products being formed in the posterior part of the body. The connection of the genital duct with this cephalic gland being then closed and cut off by the brain-case, the position of the oviducts would still be shown by the ciliated grooves opening into the folded-down thyroid tube, i.e. the folded-down horns of the uterus; the uterus itself would remain as the main body of the thyroid and still open by a conspicuous orifice into the common respiratory chamber. Next, in the degeneration process, we may suppose that not only the oviducts opened out to form the ciliated groove, but that the uterine chamber itself also opened out, and thus formed the endostyle of Amphioxus and of the Tunicata.

It might seem at first sight improbable that a closed tube should become an open groove, although the reverse phenomenon is common enough; the difficulty, however, is clearly not considered great, for it is precisely what Dohrn imagines to have taken place in the conversion of the thyroid of Ammocœtes into the endostyle of Amphioxus and the Tunicata; it is only carrying on the same idea a stage further to see in the open, ciliated groove of Ammocœtes the remains of the closed genital duct of Limulus and its allies.

Such is the conclusion to which the study of the thyroid gland in Ammocœtes seems to me to lead, and one cannot help wondering why such an unused and rudimentary organ should have remained after its original function had gone. Is it possible to find out its function in Ammocœtes?

The thyroid gland has been supposed to secrete mucus into the respiratory chamber for the purpose of entangling the particles of food, and so aiding in digestion. I see no sign of any such function; neither by the thionin method, nor by any other test, have Miss Alcock and myself ever been able to see any trace of mucous secretion in the thyroid, and, indeed, the thyroid duct is always remarkably free from any sign of any secretion whatever. Not only is there no evidence of any mucous secretion in the thyroid of the fully developed Ammocœtes, but also no necessity for such secretion from Dohrn's point of view, for so copious a supply of mucus is poured out by the glands of the branchiæ, along the whole pharyngeal tract, especially from the cells of the foremost or hyoid gills, as to mix up with the food as thoroughly as can possibly be needed. Further, too, the ciliated pharyngeal bands described by Schneider are amply sufficient to move this mixed mass along in the way required by Dohrn. Finally, the evidence given by Miss Alcock is absolutely against the view that the thyroid takes any part in the process of digestion, while, on the other hand, her evidence directly favours the view that these glandular branchial mucus-secreting cells play a most important part in the digestive process.

In Fig. 93, A is a representation of the respiratory tissue of a normal gill; B is the corresponding portion of the first or hyoid gill, in which, as is seen, the whole of the respiratory epithelium is converted into gland-tissue of the nature of mucous cells.

To sum up, the evidence is clear and conclusive that the Ammocœtes possesses in its pharyngeal chamber mucus-secreting glands, which take an active part in the digestive process, which do not in the least resemble either in structure or arrangement the remarkable cells of the thyroid gland, and that the experimental evidence that the latter cells either secrete mucus or take any part in digestion is so far absolutely negative. It is, of course, possible, that they may contain mucin in the younger developmental stages, and therefore possible that they might at that stage secrete it; they certainly, however, show no sign of doing so in their more adult condition, and cannot be compared in the very faintest degree to the glandular cells of the pharyngeal region. It is also perfectly possible for gland-cells belonging to a retrograde organ to become mucus-secreting, and so to give rise to the cells of Amphioxus and the Tunicata.

Fig. 93.—

If, then, these cells were not retained for digestive purposes, what was their function? To answer this question we must first know the function of the corresponding gland-cells in the uterus of the scorpion, which undoubtedly secreted into the cavity of the uterus and took some part in connection with the generative act, and certainly not with digestion. What the function of these cells is or in what way they act I am unable at present to say. I can only suppose that the reason why the thyroid gland has persisted throughout the vertebrate kingdom, after the generative tissues had found a new outlet for their products in the body-cavity of the posterior region, is because it possessed some important function in addition to that connected with the exit of the products of the generative organs; a function which was essential to the well-being, or even to the life of the animal. We do not know its function in the scorpion, or the nature of its secretion in that animal. We know only that physiology at the present day has demonstrated clearly that the actual external secretion of a gland may be by no means its most important function; in addition, glands possess what is called an internal secretion, viz. a secretion into the blood and lymph, and this latter secretion may be of the most vital importance. Now, the striking fact forces itself prominently forward, that the thyroid gland of the higher vertebrates is the most conspicuous example of the importance of such internal secretion. Here, although ductless, we have a gland which cannot be removed without fatal consequences. Here, in the importance of its internal secretion, we have a reason for the continued existence of this organ; an organ which remains much the same throughout the Vertebrata down to and including Petromyzon, but, as is seen at transformation, is all that remains of the more elaborate, more extensive organ of Ammocœtes. Surely we may argue that it is this second function which has led to the persistence of the thyroid, and that its original form, without its original function, is seen in Ammocœtes, because that is a larval form, and not a fully-developed animal. As soon as the generative organs of Petromyzon are developed at transformation, all trace of its connection with a genital duct vanishes, and presumably its internal secretory function alone remains.

Yet, strange to say, a mysterious connection continues to exist between the thyroid gland and the generative organs, even up to the highest vertebrate. That the thyroid gland, situated as it is in the neck, should have any sympathy with sexual functions if it was originally a gland concerned with digestion is, to say the least of it, extremely unlikely, but, on the contrary, likely enough if it originated from a glandular organ in connection with the sexual organs of the palæostracan ancestor of the vertebrate.

Freund has shown, and shown conclusively, that there is an intimate connection between the condition of the thyroid gland and the state of the sexual organs, not only in human beings, but also in numerous animals, such as dogs, sheep, goats, pigs, and deer. He points out that the swelling of the gland, which occurs in consequence of sexual excitement (a fact mentioned both in folk-lore tales and in poetical literature), and also the swelling at the time of puberty, may both lead to a true goitrous enlargement; that most of the permanent goitres commence during a menstrual period; that during pregnancy swelling of the thyroid is almost universal, and may become so extreme as to threaten suffocation, or even cause death; that the period of puberty and the climacteric period are the two maximal periods for the onset of goitre, and that exophthalmic goitre especially is associated with a special disease connected with the uterus.