Plagiacantha arachnoides, var., Claparède, 1858, Études sur les Infusoires et les Rhizopodes, p. 462, Taf. xxii. fig. 9.
Spines straight, cylindrical, trifid, with one pair of lateral branches. The nine diverging, straight, and smooth branches are connected in the distal part by slender concave bows, so that the whole skeleton exhibits nine wide meshes, three larger pentagonal and six smaller triangular.
Dimensions.—Length of the spines 0.14, of their branches 0.07.
Habitat.—North Atlantic, west coast of Norway (Claparède), surface.
3. Plectophora novena, n. sp.
Spines slightly curved, three-sided prismatic, thorny, with two pairs of lateral branches. The six smaller distal branches end freely, whilst the six larger basal branches are connected by slender convex bows. There are therefore nine wide meshes, as in the preceding closely allied species.
Dimensions.—Length of the spines 0.18, of the basal branches 0.08.
Habitat.—North Atlantic, Færöe Channel (Gulf Stream) (John Murray), surface.
4. Plectophora pyramidalis, n. sp.
Spines straight, three-sided prismatic, with three to four verticils of short lateral branches. The branches of the basal verticils are again ramified, and form by connecting bows a delicate loose framework, covering the three sides of a flat pyramid, the three edges of which are the three radial spines.
Dimensions.—Length of the spines 0.2, base of the pyramid 0.16.
Habitat.—Central Pacific, Station 267, surface.
Subfamily 2. Tetraplectida, Haeckel, 1881, Prodromus, p. 424.
Definition.—Plectanida with four radial spines.
Genus 395. Tetraplecta,[14] Haeckel, 1881, Prodromus, p. 424.
Definition.—Plectanida with four equal radial spines, arising from one common central point and corresponding to the four axes of a tetrahedron.
The genus Tetraplecta has been derived from Tetraplagia by union of the neighbouring branches of the four radial spines, diverging from a common point in different directions. In some forms of this genus the four rods seem to correspond exactly to the four axes, which are directed from the centre of a tetrahedron towards its four corners; whilst in other forms the four rods and the angles between them are perhaps not perfectly equal.
1. Tetraplecta tetrahedra, n. sp.
Spines straight, equal, three-sided prismatic, pinnate, each with three or four pairs of opposite straight slender pinnulæ or lateral branches; the pinnulæ of each side are correspondingly parallel. All pinnulæ connected by a few slender bridges, thus producing a delicate network with irregular rhomboidal meshes.
Dimensions.—Length of the spines 0.15, of the basal branches 0.05.
Habitat.—Central Pacific, Station 272, depth 2600 fathoms.
2. Tetraplecta quadricornis, n. sp.
Spines in the basal half straight, in the distal half slightly curved and irregularly branched, with two to three pairs of unequal alternate lateral branches, the distal ends of which are connected by a few slender bows, marking the six edges of an irregular tetrahedron.
Dimensions.—Length of the spines 0.22, of the basal branches 0.14.
Habitat.—Central Pacific, Station 266, depth 2750 fathoms.
3. Tetraplecta pinigera, n. sp. (Pl. 91, fig. 8).
Spines three-sided prismatic, straight, or in the basal half slightly curved, verticillate; each spine with six to eight three-branched regular verticils, tapering gradually towards the distal end. All branches of each spine lie parallel in three equidistant meridian planes, and are connected by delicate parallel threads, perpendicular to the branches. Therefore the skeleton consists of four pine-shaped trees and twelve delicate triangular wings with rectangular meshes.
Dimensions.—Length of the spines 0.25 to 0.3, of the basal branches 0.1 to 0.15.
Habitat.—Central Pacific, Station 271, surface.
Genus 396. Plectaniscus,[15] nov. gen.
Definition.—Plectanida with four unequal radial spines, arising from one common central point; one vertical or apical spine opposed to three divergent or basal spines.
The genus Plectaniscus has probably been derived from Plagoniscus by reticular union of the neighbouring branches of the four radial spines, and exhibits therefore to it the same relation that Tetraplecta bears to Tetraplagia. Whilst in these two latter genera the four spines are equal, in the two former genera there is an important difference between a vertical spine (or apical horn) and three divergent (commonly larger) spines, corresponding to the three basal feet of the triradiate Nassellaria. Perhaps, therefore, Plectaniscus is one of the remotest ancestors of the latter; but differs in the absence of a complete lattice-shell.
1. Plectaniscus cortiniscus, n. sp. (Pl. 91, fig. 9).
Spines straight, three-sided prismatic, with three to four verticils of short perpendicular branches; the branches of the distal verticils are simple and free, those of the proximal verticils again ramified and connected by an arachnoidal network of delicate threads. The vertical apical spine (or horn) is scarcely one-third or half as long as the three divergent basal spines (or feet), and the angle between it and the latter is smaller.
Dimensions.—Length of the apical spine 0.05 to 0.1, of the three basal spines 0.18 to 0.2.
Habitat.—North Pacific, Station 244, surface.
2. Plectaniscus tripodiscus, n. sp.
Spines curved, cylindrical, with a variable number of irregular branches, which in the distal half of the spines are free, in the basal half again ramified and connected by a loose spongy framework. The apical spine is shorter and simpler, nearly straight, less ramified.
Dimensions.—Length of the apical spine 0.1, of the three basal spines 0.15 to 0.18.
Habitat.—North Pacific, Station 238, surface.
3. Plectaniscus archiscenium, n. sp.
Spines slightly curved, three-sided prismatic, with two to three verticils of short curved branches. The vertical apical spine is about twice as long as the three basal spines, and connected with them by an irregular loose framework. The three basal spines are connected together only by a simple ring, so that between them remain three large collar holes, like those of the similar and closely allied Archiscenium quadrispinum (Pl. 53, fig. 11).
Dimensions.—Length of the apical spine 0.32, of the three basal spines 0.18.
Habitat.—South Pacific, Station 295, depth 1500 fathoms.
4. Plectaniscus cladoscenium, n. sp.
Spines curved, three-sided prismatic. The three basal spines pinnate, with four to five pairs of opposite curved pinnulæ; the distal pairs are simple and free, the basal pairs branched and connected by an irregular delicate framework, which together with the apical spine forms a spongy cap-shaped shell, similar to Cladoscenium. Apical spine about half as long, simple.
Dimensions.—Length of the apical spine 0.1, of the three basal spines 0.18.
Habitat.—Central Pacific, Station 274, surface.
5. Plectaniscus clathrocorys, n. sp.
Spines straight, three-sided prismatic, with seven or eight verticils of ramified branches, which in the basal half are connected by an irregular loose framework. In the distal half each of the three divergent basal spines (or feet) is connected with the longer apical spine (or horn) by a fenestrated triangular lattice wing. Therefore the skeleton becomes very similar to Clathrocorys (Pl. 64, figs. 8-10), but wants the regular central cephalis.
Dimensions.—Length of the apical spine 0.2, of the basal spines 0.14.
Habitat.—Central Pacific, Station 271, depth 2425 fathoms.
Genus 397. Periplecta,[16] Haeckel, 1881, Prodromus, p. 424.
Definition.—Plectanida with four unequal radial spines, arising in pairs from the two poles of a common central rod; one ascending apical spine opposed to three descending basal spines.
The genus Periplecta has probably been derived from Plagiocarpa (by concrescence of the meeting spine branches), and has the same important relation to a part of the triradial Stephoidea (Cortina, &c.) and Cyrtoidea (Pteroscenium, &c.). It differs from the latter only in the fact that the loose irregular framework connecting the bases of the four cortinar spines is not a regular lattice-shell.
1. Periplecta cortina, n. sp. (Pl. 91, fig. 10).
Spines of very different size and shape, three-sided prismatic. The apical spine (or horn) is larger, nearly straight, and bears a verticil of three large divergent branches, which are again ramified. The three basal spines (or feet) are curved, pinnate, each with three to five pairs of opposite simple curved branches. A small common central rod separates the two united pectoral feet from the two other spines (the caudal foot and the apical horn). The three basal feet are connected by three convex bows composing a horizontal collar ring, and since the central rod is prolonged between the pectoral feet to the ring, a very remarkable collar septum is formed with four cortinar meshes. Above this septum an irregular spongy shell, including the central capsule, is formed by a delicate framework, interwoven between the apical horn and the three basal feet.
Dimensions.—Length of the apical spine 0.25 to 0.3, of the three basal spines 0.15 to 0.2.
Habitat.—Central Pacific, Station 271, surface.
2. Periplecta pteroscenium, n. sp.
Spines of very different size and shape, three-sided prismatic, verticillate. The apical spine (or horn) with ten to twelve verticils, is one and a half times as long as the three basal spines (or feet), which bear six to eight verticils only; each verticil with three branches which by communicating ramules form a loose irregular wickerwork. Similar to Pteroscenium pinnatum, Pl. 53, figs. 14-16, but without a regular lattice-shell. A short basal central rod separates two equal (pectoral) spines from two unequal spines, the larger of the latter is the apical, the shorter the caudal spine.
Dimensions.—Length of the apical spine 0.26, of the three basal spines 0.17.
Habitat.—Central Pacific, Station 274, surface.
3. Periplecta monocyrtis, n. sp.
Spines of nearly equal size and shape, cylindrical, irregularly branched. The apical spine is straight, more branched and nearly twice as long as the three basal spines. All four spines in the basal third connected by a loose spongy framework, approaching the form of some Monocyrtida.
Dimensions.—Length of the apical spine 0.24, of the basal spines 0.13.
Habitat.—Central Pacific, Station 274, surface.
Subfamily 3. Hexaplectica, Haeckel.
Definition.—Plectanida with six radial spines.
Genus 398. Hexaplecta,[17] Haeckel, 1881, Prodromus, p. 425.
Definition.—Plectanida with six radial spines, arising from one common central point.
The genus Hexaplecta has probably been derived from Hexaplagia by concrescence of the meeting branches of the six spines, and therefore has to it the same relation that Plectophora bears to Plagiacantha. The two species here described represent perhaps two different genera, since in the first all six spines are equal, opposed in pairs, in the second different in pairs (three upper smaller and three lower larger).
1. Hexaplecta triaxonia, n. sp.
Spines equal, opposite in three crossed pairs, straight, three-sided prismatic, thorny, in the distal half with three leaf-shaped, dentate edges, in the basal half with three verticils of slender branches, which are connected by parallel threads, and so form an arachnoidal network with rectangular or rhomboidal meshes.
Dimensions.—Length of the spines 0.22, of their basal branches 0.06.
Habitat.—South Pacific, Station 285, depth 2375 fathoms.
2. Hexaplecta tricladonia, n. sp.
Spines unequal, three larger spines (in the lower half of the body) being opposite to three smaller spines (in the upper half). The latter are nearly horizontally expanded, each bearing three straight divergent branches. The larger are stronger, twice as long and armed with three verticils, each of three branches. All six spines are slender, three-sided prismatic. Their branches are very thin, thread-shaped, and form by their union a loose network with irregular polygonal meshes.
Dimensions.—Length of the larger spines 0.25, of the smaller 0.12.
Habitat.—South Pacific, Station 291, surface.
Genus 399. Plectanium,[18] Haeckel, 1881, Prodromus, p. 424.
Definition.—Plectanida with six radial spines, arising in two opposite divergent groups from the two poles of a common central rod.
The genus Plectanium has been derived from Plagonium by concrescence of the meeting branches, and bears therefore to it the same relation that the preceding Hexaplecta exhibits to Hexaplagia. Whilst in these two latter genera the six spines arise from a common central point, they arise here in two divergent groups from the two poles of a horizontal common middle rod, similar to the spicula of many Beloidea.
1. Plectanium trigeminum, n. sp. (Pl. 91, fig. 11).
Spines straight and stout, six to eight times as long as the common middle rod, three-sided prismatic; in the distal half thickened, with three divergent terminal thorns; in the basal half with three to four verticils of thin lateral branches, which are again ramified, and by their united threads produce an irregular loose framework.
Dimensions.—Length of the spines 0.25, basal breadth 0.005, terminal breadth 0.02; length of the middle rod 0.03.
Habitat.—North Pacific, Station 244, surface.
2. Plectanium ovodimare, n. sp.
Spines straight and stout, three-sided prismatic, about four times as long as the common middle rod, gradually thinned towards the distal end; in the basal half with two to three verticils of forked branches, which are dichotomously ramified, and by their united threads form a loose ovate framework.
Dimensions.—Length of the spines 0.17, of the middle rod 0.04.
Habitat.—North Pacific, Station 236, surface.
3. Plectanium sphærozoum, n. sp.
Spines straight, cylindrical, ten to twelve times as long as the common middle rod, with six to eight verticils of thorny branches, tapering towards the distal end; the branches of the verticils are simple in the distal half, again ramified in the basal half, and here connected together by irregularly branched threads forming a loose framework. (Resembles one spiculum of Sphærozoum verticillatum, Pl. 4, fig. 7, but is more richly branched, with united ramules.)
Dimensions.—Length of the spines 0.2, of the middle rod 0.01.
Habitat.—Central Pacific, Station 266, surface.
Subfamily 4. Polyplectida, Haeckel, 1881, Prodromus, p. 424.
Definition.—Plectanida with numerous (seven to nine or more) radial spines.
Genus 400. Polyplecta,[19] Haeckel.
Definition.—Plectanida with numerous (seven to nine or more) radial spines, arising from a common centre and lying in different planes.
The genus Polyplecta comprises provisionally all Plectanida possessing seven or more radial spines, united in the centre of the framework. It may be derived from Polyplagia by meeting and concrescence of the free branches. But as in the latter genus, here also the few observed species are very different, perhaps of different origin, and may be afterwards better separated as representatives of diverse genera.
1. Polyplecta heptacantha, n. sp. (Pl. 91, fig. 12).
Heptaplegma heptacantha, Haeckel, 1882, MS.
Seven unequal spines divergent from one common central point, slender, slightly curved, three-sided prismatic. Four spines are much larger, twice to three times as long and as thick as the three smaller spines. One of the four larger spines is directed upwards (as "apical horn"), whilst the three others diverge downwards (as "basal feet"). Each of the four larger spines bears five to seven equidistant verticils of three divergent branches, the proximal of which are branched, the distal simple. The three smaller spines lie nearly horizontally, midway between the odd apical spine and the three basal spines, and in the same meridian planes with them; each bears one verticil of three divergent branches. All the branches are united by arachnoidal threads, composing a loose wickerwork with irregular, generally quadrangular meshes.
Dimensions.—Length of the four larger spines 0.27 to 0.33, of the four smaller 0.1 to 0.14.
Habitat.—Central Pacific, Station 271, surface.
2. Polyplecta enneacantha, n. sp.
Enneaplegma enneacantha, Haeckel, 1881, Prodromus, p. 425.
Nine equal and equidistant, straight, cylindrical radial spines, lying nearly in one plane, arise from an irregular spongy central framework; six of them seem to be secondary, intercalated between three equidistant primary spines, which are united in the centre. (This species resembles in the nine-radial structure the remarkable Enneaphormis rotula, Pl. 57, fig. 9, and may perhaps be derived from a similar species; but it has no regular latticed shell.)
Dimensions.—Length of the spines 0.22, diameter of the framework 0.17.
Habitat.—Central Pacific, Station 266, depth 2750 fathoms.
3. Polyplecta decacantha, n. sp.
Pentaplegma decacantha, Haeckel, Prodromus, p. 425.
Ten radial spines, curved, cylindrical, irregularly branched, diverge in different directions and seem to arise in pairs from an irregular central framework, in the centre of which five primary spines are united; the latter correspond probably to the five spines of Pentaspyris, &c. The density of the spongy central framework did not allow of an accurate investigation, and makes it doubtful whether this species is not a Spongiomma.
Dimensions.—Length of the spines 0.3 to 0.4, diameter of the framework 0.18.
Habitat.—Tropical Atlantic, Station 338, depth 1990 fathoms.
4. Polyplecta polybrocha, Haeckel.
? Acanthodesmia polybrocha, Haeckel, 1865, Zeitschr. f. wiss. Zool., Bd. xv. p. 368, Taf. xxvi. fig. 3.
Plegmosphæra polybrocha, Haeckel, 1881, Prodromus, p. 455.
Numerous (twenty to thirty or more) radial spines, thin, cylindrical, curved and irregularly branched, arising from an irregular central spongy framework, are connected by numerous slender arches. The specimen observed by me in 1880 in Portofino was a true Polyplecta, with three primary spines centrally united, between which numerous other spines were intercalated. The similar specimen, however, observed in 1864 in Villafranca, and figured, loc. cit., was perhaps a Plegmosphæra.
Dimensions.—Length of the radial spines 0.05 to 0.1, diameter of the framework 0.16.
Habitat.—Mediterranean (Villafranca, Portofino), surface.
5. Polyplecta dumetum, Haeckel.
Acanthodesmia dumetum, J. Müller, 1868, Abhandl. d. k. Akad. d. Wiss. Berlin, p. 30, Taf. i. Fig. 3.
Numerous (ten to twelve or more) radial spines, thin and straight, with a few straight lateral branches, diverge in different directions and are connected by a few slender curved arches. Some similar forms, but more developed, with numerous branches and curved rods, are found in the Pacific Radiolarian ooze, and represent probably different species.
Dimensions.—Length of the radial spines 0.05 to 0.08, of the branches 0.02 to 0.03.
Habitat.—Mediterranean, French shore (Cette and Saint Tropez), Johannes Müller, surface.
Suborder III. STEPHOIDEA, Haeckel.
Stephoidea vel Stephida, Haeckel, 1881, Prodromus, p. 444.
Acanthodesmida (sensu ampliori), Bütschli, 1882, Zeitschr. f. wiss. Zool., vol. xxxvi. p. 495.
Definition.—Nassellaria without complete lattice-shell, with a skeleton composed of one or more simple rings, which may be united by a loose framework and are separated by large openings or gates. One primary or sagittal ring, determining the sagittal or median plane of the bilateral body, encloses the monaxonian central capsule.
The suborder Stephoidea, hitherto known by a few species only of "Acanthodesmida," comprises a large number of interesting Nassellaria (now more than two hundred species), which possess peculiar interest for the morphology and phylogeny of this legion. The monaxonian central capsule of the Stephoidea is surrounded either by one simple ring or by a complex system of several loosely connected rings; these may be united by a loose framework of connected branches, but never produce a complete lattice-shell, as is constantly the case in the Spyroidea, Botryodea, and Cyrtoidea. Therefore there remain between the parts of the connected rings a few large openings which we call "gates," separating them from the numerous small "pores" of the complete lattice-shells. In the most simple case, if only one ring be formed, there is also present only one "gate," the aperture of this simple ring.
The first known species of Stephoidea were observed in the Mediterranean by Johannes Müller in 1856, and described and figured in his last treatise (1858) under the names Lithocircus annularis (loc. cit., Taf. i. fig. 1) and Acanthodesmia vinculata (loc. cit., Taf. i. figs. 4-7). In the following year I myself observed two other living species in the Mediterranean, and described them in my Monograph (1862, pp. 268, 270) as Zygostephanus mülleri (Taf. xii. fig. 2) and Prismatium tripleurum (Taf. iv. fig. 6). For these four longest known Stephoidea I founded the new family of Acanthodesmida (loc. cit., p. 265), but united with them two other similar genera which I afterwards separated:—Plagiacantha (belonging to the Plectoidea) and Dictyocha (belonging to the Phæodaria).
When, in 1876, I received the rich material of the Challenger collection, I was astonished to find in it an enormous number of new, similar, and partly very interesting "Acanthodesmida," which were afterwards arranged in my Prodromus (1881, p. 444) in thirty-eight different genera and four "subfamilies," all united in one single large family, "Stephida or Monopylaria cricoidea." I retain here this natural group in the same sense, but give to it the rank of a "suborder," separating at the same time its four subfamilies as substantial "families." Since the names of the latter, proposed in the Prodromus, were not quite suitable, I replace them here by the following more convenient names:—
1. Family Stephanida (= Monostephida, 1881, p. 447). Skeleton entirely formed by one simple ring (the primary vertical sagittal ring). The only "gate" is the simple aperture of the ring.
2. Family Semantida (= Dyostephida, 1881, p. 446). Skeleton composed of two rings, perpendicular to one another; the primary vertical sagittal ring bears at the base a horizontal basal ring; between the two rings two or more "basal gates remain."
3. Family Coronida (= Triostephida, 1881, p. 445). Skeleton composed of two crossed vertical or meridional rings, perpendicular to one another—the primary sagittal ring and the secondary frontal ring. Commonly the two vertical rings are united at the base by a horizontal basal ring, and between them remain two or more "basal gates."
4. Family Tympanida (= Parastephida, 1881, p. 446). Skeleton composed of two parallel horizontal rings, and upper mitral ring, and a lower basal ring, both connected by vertical or divergent columellæ, which are parts of vertical rings (primary sagittal, and secondary frontal ring).
The peculiar structure of the central capsule of the Stephoidea, and their character as true Monopylea, were first recognised by Richard Hertwig, who in 1879, in his Organismus der Radiolarien, gave an excellent detailed description of it (loc. cit., pp. 68-72, Taf. vii. figs. 4, 5). He also pointed out the near affinity of these "Acanthodesmida" with the Spyroidea or Zygocyrtida, uniting the latter with the former family.
A fuller explanation of this affinity, and of the morphological and phylogenetic importance of the "Acanthodesmida," as ancestral forms of the Spyroidea and Cyrtoidea, was given in 1882 by Bütschli (Zeitschr. f. wiss. Zool., vol. xxxvi. pp. 495-501). He described in detail some important fossil forms of Acanthodesmida as different species of "Stephanolithis," a name which Ehrenberg had employed for various ring-like fragments of Radiolarian shells, sponges, and other fossil bodies. The four fossil species which Bütschli described represent four different genera of Stephoidea, viz., Semantis (spinescens), Semantrum (mülleri), Semantidium (haeckelii), and Tristephanium (hertwigii). On the phylogenetic conclusions, derived from these accurate observations, compare above, p. 893, &c. The topographical signification of the parts, employed by Bütschli, is contrary to mine; he calls my dorsal side the "anterior," and my ventral side the "posterior."
The geometrical fundamental form of the body in nearly all Stephoidea (with few exceptions) is distinctly bilateral or "dipleuric," so that we can easily distinguish the three different dimensive axes: the principal axis with different apical and basal poles, the sagittal axis with different dorsal and ventral poles, and the lateral axis with equivalent right and left poles. In only a few genera this bilateral symmetry is not expressed, and a simpler, more regular fundamental form appears. The latter may be either primary (in the monaxonian Archicircus and Lithocircus) or secondary, afterwards acquired (in the octahedral Trissocircus and Trissocyclus, the cubical Lithocubus, the prismatic Eutympanium, and some other forms).
The most important element of the skeleton, with which the formation begins, in all Stephoidea is the simple primary or sagittal ring, lying vertically in the sagittal or median plane of the body and surrounding the monaxonian central capsule. This sagittal ring is the only essential element of the skeleton in all Stephanida, and is completely preserved in all Semantida, also in the greater part of the Coronida and Tympanida. It is partially reduced in the small groups of the true Acanthodesmida (subfamily of Coronida) and the Dystympanida and Eutympanida (subfamilies of Tympanida). Here only the vertical parts of it are preserved (dorsal and ventral rod), whilst the horizontal parts are lost (mitral and basal rod).
The sagittal ring lies constantly in the vertical median plane of the body, and therefore divides the enclosed central capsule into a right and a left half. It is rarely regular or subregular, commonly dipleuric or distinctly bilateral, so that we may easily distinguish its dorsal and ventral, apical and basal parts. The most important of these four parts or "rods" is the "basal rod" or the inferior part, because here the ring is in closer connection with the central capsule and its "porous area"; here peculiar spines or branches are commonly developed, which even on the isolated ring immediately determine the basal pole. The opposite upper part, or the "mitral rod," is also often distinguished by the peculiar appendages. The posterior part, or the "dorsal rod" (the anterior rod, a, in the description of Bütschli), is commonly more straight, often quite vertical. The opposite anterior part, or the "ventral rod" (the posterior rod, b, of Bütschli), is usually more convex, and often strongly curved or semicircular.
Whilst this dipleuric or bilateral (commonly obliquely ovate or nearly triangular) form of the sagittal ring is distinctly preserved in by far the greater number of Stephoidea, it is replaced in some few genera by a more regular, amphithect, diphragmatic, or biradial form. In this case we may often suppose a primary regularity to exist, the dorsal and ventral parts being not yet differentiated, as in Archicircus and Lithocircus, Zygostephanus and Protympanium, and perhaps also in some other forms. But in other cases the regularity is, on the contrary, secondary, being derived from original bilateral forms.
The rod of the sagittal ring is either cylindrical (with circular transverse section) or angular (commonly with triangular transverse section). In nearly all Stephoidea (with very few exceptions) branches or apophyses are developed from the ring, regularly disposed and often of great morphological importance. Commonly these apophyses are developed in pairs, growing symmetrically on both sides of the ring. The most important of these apophyses are:—(1) Basal apophyses, arising from the basal hole of the ring; (2) mitral apophyses, arising from the apical pole; (3) dorsal apophyses, arising from the middle of the dorsal rod; and (4) ventral apophyses, arising from the middle of the ventral rod. The two former arise in the principal axis, the two latter in the sagittal axis of the body. Very frequently the latter pair is replaced by two pairs of transverse branches, one inferior (mandibular) and one superior (orbital). The apophyses of the ring are either simple or branched, often very large, richly ramified, and give origin to a number of further products.
Whilst in the Stephanida the primary sagittal ring alone represents the whole skeleton, it produces in all other Stephoidea one or more secondary rings. The most important of these is the horizontal basal ring, appearing first in the Semantida (Pl. 92). From the base of the sagittal ring there arise in the horizontal basal plane two pairs of lateral branches or "basal apophyses." The curved opposite branches of the corresponding pairs become united on each side of the primary ring (right and left), and so produce a second, horizontal ring, perpendicular to the former. This basal ring encloses two paired basal gates, which are enclosed on the medial side by the basal rod of the sagittal ring, and on the lateral side by two united apophyses (Semantis, Pl. 92, figs. 1, 2). These two primary basal gates are of the greatest morphological importance; we call them the "jugular gates or jugular pores" (in the description of Bütschli, the pores I, loc. cit., p. 498). The dorsal pair of basal apophyses (on their posterior edge) are the coracal rods, e (rods e of Bütschli); the opposite ventral pair (on their anterior edge) are the clavicular or furcular rods, f (rods e1 of Bütschli); compare Pls. 92-95, and their explanation.
The skeleton of Semantis, the prototype of the Semantida, thus assumes the characteristic form of a signet-ring. The basal ring enclosing the two jugular pores corresponds to the seal-plate. It is commonly more or less horizontal; but often the apophyses descend obliquely (Pl. 92, figs. 1, 2, 13, &c.), more rarely laterally (Pl. 29, fig. 11). The further development of this typical form is essentially effected by the production of new basal pores in the horizontal seal-plate. In Semantrum (Pl. 92, figs. 3, 4, 5) we find already four basal gates. Behind the jugular gates is formed a second pair, the "cardinal gates" (pores II of Bütschli). These are enclosed on the anterior margin by the coracal rods (e), on the posterior margin by the scapular rods (d), a third pair of basal apophyses, arising behind the former from the sagittal ring and uniting with them (the rods e2 of Bütschli). Commonly the two posterior, or cardinal gates are much larger than the two anterior, jugular gates. This characteristic basal plate of Semantrum, with two pairs of basal pores, is of the greatest morphological importance, as it is inherited in by far the greater number of the Nassellaria, though not so generally as Bütschli supposes. The basal ring of Semantrum is either more circular or elliptical, or more polygonal, and is connected with the basal rod of the sagittal ring by three pairs of radial apophyses, the anterior furcular, the middle coracal, and the posterior scapular rods.
A third important form of Semantida is Semantidium (Pl. 92, figs. 6, 7). Here we find three pairs of basal pores in the seal-plate; the third pair, newly formed, consists of the cervical gates (e), bounded in front by the scapular rods (d), behind by a fourth pair of basal apophyses, the cervical rods.
The basal apophyses of the sagittal ring are not only of great morphological importance, because they produce by their union three typical pairs of basal gates or "collar pores," but also because their prolongations often appear as typical basal feet. The distal prolongations of the coracal rods appear in Semantiscus (Pl. 92, figs. 16-18) as two pectoral feet, those of the scapular rods as two tergal feet, whilst the opposite prolongations of the basal rod of the sagittal ring appear as two "sagittal feet" (in front an anterior or sternal, and behind a posterior or caudal foot). In the typical Cortiniscus (Pl. 92, figs. 11-13) only three feet are developed; an odd caudal and two paired pectoral feet (compare above, p. 891). The typical basal ring of the Semantida, with its paired basal gates (Semantis), reappears in the majority of the Coronida, differing from the former in the development of a second vertical ring, which lies in the frontal plane (perpendicular to the sagittal ring), and which we therefore call the frontal ring. In only one small group of the Coronida the basal ring is absent, namely, in the Zygostephanida, and here the frontal ring appears in the simplest form, as a complete elliptical meridian ring, crossing the sagittal ring perpendicularly on the two poles of the main axis (Zygostephanus, Pl. 93, figs. 1-4), Four large lateral gates between the two rings remain open. This form may be derived directly from the Stephanida in the following way; from both poles of a simple sagittal ring there arise two opposite lateral apophyses, which in the frontal plane become curved one towards the other, and united in the poles of the transverse axis. The basal apophyses would be the coracal rods. But it is also possible that Zygostephanus was derived from Semantis by the loss of the furcular rods.
The three typical rings (or the "dimensive rings") of the Stephoidea appear in their most complete form in the subfamily Trissocyclida (Pl. 93, figs. 7, 13). Here all three rings are undivided and completely developed in the three dimensive planes, perpendicular one to another. Between them there remain eight large open gates; the four superior are the four "lateral gates" of Zygostephanus, the four inferior are the four basal gates of Semantrum. The four latter are originally much smaller than the four former; but in Trissocircus and Trissocyclus (Pl. 93, figs. 10-12) they reach the same size. Therefore all eight gates are here of equal form and similar size, and the basal ring, now a true equatorial ring, divides the two meridional rings into two equal halves.
In the Eucoronida, a third subfamily of Coronida, the sagittal and the basal rings are complete, but the frontal ring is incomplete, its basal part being wanting (Pl. 82, figs. 4-6). Therefore we find here six large gates between the three rings; four upper lateral gates (between the two crossed vertical rings) and two lower basal gates (between the basal rod of the sagittal ring and the two halves of the basal ring). Eucoronis, the type of this subfamily, may be derived either directly from Semantis by development of a frontal ring, or from Tristephanium by loss of the basal part of the frontal ring.
A quite simple basal ring with a single gate, distinguishes the fourth subfamily of Coronida, the Acanthodesmida (sensu restricto), the genera Coronidium (Pl. 82, figs. 1, 2, 7, 8) and Acanthodesmia (Pl. 93, fig. 5). The horizontal basal ring alone is here complete, whilst both vertical rings (the sagittal and frontal rings) are incomplete, their basal parts being wanting. Therefore there are here five gates, four lateral and one basal. These forms may be derived from Eucoronis by loss of the basal rod of the primary sagittal ring.
The Tympanida, the fourth family of Stephoidea, exhibit another type of ring structure. Here two parallel horizontal rings are constantly developed, one on the apical pole, the other on the basal pole of the sagittal ring. The latter is the same basal ring as in the Semantida and Coronida. The former is a "mitral ring," developed in the same manner, by union of two pairs of horizontal lateral branches, which arise on both sides from the apical rod (or mitral rod) of the sagittal ring. In the simplest case these two parallel horizontal rings are connected only by the sagittal ring, which is either complete (Protympanium, Pl. 93, fig. 14) or incomplete (Parastephanus, Pl. 93, fig. 21). But commonly also an incomplete frontal ring is present, so that the two horizontal rings are connected by four vertical or subvertical rods; two of these "columellæ" are the dorsal and ventral rods of the sagittal ring, the two others are the lateral rods of the frontal ring. Between the former and the latter are sometimes developed two, four, or more accessory columellæ (probably halves of accessory incomplete diagonal meridian rings). In this way arise the characteristic "drum-forms" of many Tympanida, in which the two parallel horizontal rings correspond to the upper and lower rings of a drum, whilst the connecting vertical columellæ correspond to its parallel lateral rods (Pl. 83, figs. 1, 2, &c.).
Originally the two horizontal rings of these "drum-shells" are both bisected by the complete sagittal ring, each provided with two lateral gates (Protympanida). But in the Eutympanida both rings exhibit one simple gate only, the apical and the basal rod of the sagittal ring being lost. In the Paratympanida both rings are closed by a secondary lattice-plate whilst in the Dystympanida the upper (mitral) ring alone is closed by such a plate, the lower (basal) ring is open. In some Eutympanida the shell assumes the strange form of a regular geometrical cube, the twelve edges of which are represented by thin rods of silex (Lithocubus)—its four upper edges represent the mitral ring, the four lower the basal ring; two opposite of the four vertical cube-edges are the lateral halves of the frontal ring, the two other alternate ones are the remaining halves of the reduced sagittal ring (dorsal and ventral rod) (Pl. 82, fig. 12).
In many Tympanida and Coronida a loose irregular lattice or framework is developed, which partly closes the large open gates. But this never reaches the completeness of a true lattice-shell, such as we find in the Spyroidea, Botryodea, and Cyrtoidea. In by far the greater number of Stephoidea the corners, and partly also the rods, of the shell are armed with numerous irregular spines, often forked or richly branched. Among these spines the descending "basal apophyses" possess a peculiar importance, since by their regular number and disposition they correspond to the radial rods of the Plectoidea, and to the typical "feet" of the Spyroidea and Cyrtoidea. The most important of them are the three cortinar feet (one caudal and two pectoral) of Cortina, Cortiniscus, &c. (compare above, p. 891).
The Central Capsule exhibits in the Stephoidea the same characteristic structure as in all other Monopylea, first exactly pointed out by Richard Hertwig in 1879 (Organismus der Radiol., p. 71, Taf. vii. figs. 4, 5). Its form is commonly ovate or ellipsoidal, sometimes also lentelliptical or nearly spherical. It exhibits constantly on the basal pole the porochora or porous area, and in the basal half the podoconus or pseudopodial cone. From the surrounding sagittal ring it is separated by a thick jelly-like calymma, which commonly exhibits numerous zooxanthellæ. The numerous pseudopodia are commonly branched, with rather rare anastomoses. The membrane of the central capsule is thick.
Synopsis of the Families of Stephoidea.
| I. Skeleton composed of the simple vertical sagittal ring only, without secondary rings, | 1. Stephanida. |
| II. Skeleton composed of two crossed rings, a vertical sagittal and a horizontal basal ring, | 2. Semantida. |
| III. Skeleton composed of two crossed vertical meridional rings (a primary sagittal and a secondary frontal ring), commonly also with a horizontal basal ring, | 3. Coronida. |
| IV. Skeleton composed of two parallel horizontal rings (upper mitral and lower basal ring), both connected by a vertical sagittal ring (and often by a vertical frontal ring), | 4. Tympanida. |