3. Lithocubus vinculatus, Haeckel.

The twelve rods of the cubical shell are curved and armed with scattered, simple, short spines. The figure 7 of Johannes Müller (loc. cit.) corresponds exactly to the Mediterranean form observed by me at Portofino, and is quite different from his true Acanthodesmia vinculata (loc. cit., figs. 4-6), so that I have no doubt he did observe these two different species (compare above, p. 975).

Dimensions.—Diameter of the cube 0.07; length of the spines 0.02.

Habitat.—Mediterranean (Nice, Portofino).

4. Lithocubus astragalus, n. sp. (Pl. 82, fig. 12).

The twelve rods of the cubical shell are stout and slightly curved, armed with numerous, simple, and irregularly branched spines. Eight larger spines, branched like a deer's antler, arise from the eight corners of the cube.

Dimensions.—Diameter of the cube 0.1 to 0.12; length of the corner spines 0.05 to 0.08.

Habitat.—Central Pacific, Station 271, depth 2425 fathoms.

Genus 439. Circotympanum,^[57] n. gen.

Definition.—Tympanida with two simple different horizontal rings, connected by numerous (six to eight or more) divergent columellæ.

The genus Circotympanum comprises those Tympanida in which two simple horizontal rings of different size (an upper smaller "mitral ring," and a lower larger "basal ring") are connected by a variable number of divergent rods or columellæ (at least six or eight). Circotympanum is probably derived from Tympaniscus or Tympanidium by partial reduction of the sagittal ring, the upper (apical) and lower (basal) part of which is lost. The geometrical fundamental form is a truncated pyramid with six to eight or more edges.

1. Circotympanum hexagonium, n. sp.

Mitral and basal ring hexagonal, the latter twice as broad as the former, each with a simple small spine on the six corners. Six columellæ nearly straight, as long as the radius of the basal ring. All rods with three prominent edges.

Dimensions.—Breadth of the mitral ring 0.06, of the basal ring 0.11.

Habitat.—Central Pacific, Station 268, depth 2900 fathoms.

2. Circotympanum octogonium, n. sp. (Pl. 94, fig. 17).

Mitral and basal rings octagonal, the latter one and a half times as broad as the former. Eight columellæ slightly curved, thorny. The mitral ring bears on each corner a short ascending spine, the basal ring a larger, thorny, obliquely descending spine, which has some recurved thorns on the upper edge. All rods more or less thorny.

Dimensions.—Breadth of the mitral ring 0.06, of the basal ring 0.09.

Habitat.—Central Pacific, Station 271, depth 2425 fathoms.

3. Circotympanum decagonium, n. sp.

Mitral and basal rings circular, the latter nearly twice as broad as the former, both armed with scattered irregular thorns. Ten columellæ smooth, curved.

Dimensions.—Breadth of the mitral ring 0.07, of the basal ring 0.12.

Habitat.—North Pacific, Station 244, depth 2900 fathoms.

Genus 440. Eutympanium,^[58] Haeckel, 1881, Prodromus, p. 447.

Definition.—Tympanida with two simple equal horizontal rings, connected by numerous (six to eight or more) parallel and vertical columellæ.

The genus Eutympanium contains those Tympanida in which the drum-shaped shell is composed of two simple, parallel and equal horizontal rings, and numerous vertical and parallel rods or columellæ connecting them. The number of the latter is at least six or eight. Two opposite of these must be regarded as remaining vertical parts of the sagittal ring, two others as parts of the frontal ring, whilst the other columellæ (between the former and the latter) are secondary and intercalated. Therefore Eutympanium may be derived from Tympanidium by reduction and loss of the apical and basal parts of the sagittal ring.

1. Eutympanium musicantum, n. sp. (Pl. 83, fig. 2).

Eight parallel and vertical columellæ, simple, straight and smooth. Both horizontal rings equal, circular, each armed with eight to twelve short conical thorns.

Dimensions.—Height of the shell 0.09, breadth 0.07.

Habitat.—Central Pacific, Station 266, depth 2750 fathoms.

2. Eutympanium octonarium, n. sp.

Eight vertical columellæ slightly curved, thorny. Both horizontal rings equal, octagonal, each armed with eight large, irregularly-branched spines, which on the mitral ring are directed obliquely upwards, on the basal ring downwards.

Dimensions.—Height of the shell 0.12, breadth 0.1.

Habitat.—Central Pacific, Station 268, depth 2900 fathoms.

3. Eutympanium coronarium, n. sp.

Eight thorny columellæ, obliquely descending and curved like a bow, divergent in the upper part, convergent in the lower. Basal ring circular, mitral ring octagonal. Both horizontal rings armed with eight stout thorny spines, which on the mitral ring are smaller and directed obliquely upwards, on the basal ring larger and directed downwards, each with a recurved hook.

Dimensions.—Height of the shell 0.08, breadth 0.06.

Habitat.—South Atlantic, Station 332, depth 2200 fathoms.

4. Eutympanium militare, n. sp. (Pl. 82, fig. 11).

Ten parallel and vertical columellæ straight and smooth, cylindrical. Both horizontal rings equal, circular, armed with numerous short conical spines, divergent in all directions. Ten to twenty spines on each ring larger than the others.

Dimensions.—Height of the shell 0.1, breadth, 0.08.

Habitat.—Western Tropical Pacific, Station 225, depth 4475 fathoms.

5. Eutympanium dodecarium, n. sp.

Twelve vertical columellæ, slightly curved, smooth. Mitral and basal rings circular, both horizontal rings armed with twelve stout, irregularly-branched spines, which are prolongations of the columellæ, and on the mitral ring directed upwards, on the basal ring downwards.

Dimensions.—Height of the shell 0.13, breadth 0.1.

Habitat.—South Pacific, Station 300, depth 1375 fathoms.

Order VI. CYRTELLARIA, Haeckel, 1881.

Definition.—Nassellaria with a complete lattice-shell enveloping the central capsule.

Suborder I. SPYROIDEA, Haeckel.

Definition.—Nassellaria with a complete lattice-shell, exhibiting constantly a bilocular cephalis with a sagittal constriction.

The suborder Spyroidea, and the two following closely allied suborders, Botryodea and Cyrtoidea, represent together that large group of Radiolaria which I first described in my Monograph (1862, pp. 272, 280) as the family Cyrtida, but afterwards as a separate order (or sublegion) under the name Cyrtellaria (1883, Jena. Sitzungsber., Feb. 16, p. 18). This group comprises all those Monopylea or Nassellaria which possess a complete lattice-shell, whilst the preceding Plectellaria never develop a perfect fenestrated shell enveloping the central capsule. The Spyroidea differ from the other Cyrtellaria (Botryodea and Cyrtoidea) in the bilocular shape of the cephalis, which is bisected by the sagittal ring and a corresponding longitudinal constriction into two symmetrical halves.

The Spyroidea appear in the first system of Polycystina of Ehrenberg (1847, loc. cit., pp. 53, 54) as the fourth of his seven families, under the name Spyridina, with the following definition:—"Testæ nucleo destitutæ (associatæ et coalitæ); cellulæ binæ clathratæ, nucis forma amplæ, strictura longitudinali levius discretæ." Ehrenberg united them with his "Polycystina composita or Spumellaria" and separated them from the closely allied "Polycystina solitaria or Nassellaria." He distinguished among them five genera, two of which have no external appendages (Dictyospyris and Pleurospyris), whilst the other three possess spiny or lamellar appendages (Ceratospyris, Cladospyris, and Petalospyris). These five genera and the accompanying definitions were also repeated in the same terms in the last system of Ehrenberg (1875, loc. cit., p. 157). In my Monograph (1862, pp. 280, 291) the Spyroidea are enumerated as a subfamily of the Cyrtida, under the name Zygocyrtida, with the following definition:—"Lattice-shell divided by an annular sagittal longitudinal constriction into two neighbouring equal joints or chambers." I adopted at that time the above five genera of Ehrenberg, only eight species of which were known.

Richard Hertwig gave in 1879 the first accurate description of the central capsule of the Spyroidea, and demonstrated that they possess the same characteristic structure of the capsule, with a porochora and a podoconus, that is common to all Monopylea. He separated the Zygocyrtida from the other Cyrtida, and united them with the Stephoidea (Lithocircus) in the family Acanthodesmida. In my Prodromus (1881, p. 449) I adopted that division and put the "Spyrida or Spyroidea" as a separate family between the Cyrtoidea and Stephoidea. I there gave the following definition:—"Monopylaria gemina, testa silicea clathrata gemina, cameris binis juxta compositis, annulo verticali sagittali contiguis, capsula centrali a testa inclusa. Promorpha dipleura vel bilaterali, dextro et sinistro antimero symmetricis, plano sagittali annuloque separatis. Polum superiorem axis verticalis vel principalis plerumque spina apicalis (vel occipitalis) occupat, polum inferiorem ostium clathratum (poris tribus aut quatuor aut pluribus) et spina caudalis (posterior). Ad dextram et sinistram duæ spinæ laterales distant. Ab hac forma tripoda (Monocyrtida triradiata acuta cum annulo mediano) diversæ Spyridum formæ derivandæ sunt." I there divided the Spyrida into seven subfamilies and forty-eight genera. Bütschli in 1882 published accurate researches on the Zygocyrtida in his Beiträge zur Kenntniss der Radiolarien-Skelete, insbesondere der der Cyrtida (Zeitschr. f. wiss. Zool., vol. xxxvi. p. 501, Taf. xxxii.). He gave a very exact anatomical description of the skeleton of several fossil Spyroidea of Barbados, and pointed out their close affinity to the Acanthodesmida on the one hand, and to the Cyrtida on the other. Regarding their phylogenetical relation, he supposed that the Zygocyrtida have descended from the Acanthodesmida (our Stephoidea) and are the ancestors of all other Cyrtida. The best part of this valuable paper of Bütschli is his exact description of the sagittal ring and its apophyses, producing the basal plate; and the general conception thus derived of the essential parts of the skeleton in all Monopylea, bearing a sagittal ring. But his general conclusions were partly erroneous, since the fossil Spyroidea and Cyrtoidea of Barbados, upon which they were founded, contain only the minority of genera of the large group; and many important and typical forms remained unknown to him.

The numerous new genera discovered in the rich collection of the Challenger, and mainly in the wonderful Radiolarian ooze of the Central Pacific (Stations 263 to 274) throw a new light upon the complicated affinities of the whole group, and manifest a far greater richness in developmental variations and admirable morphological productions than Bütschli could suppose. The following description, however, must remain very incomplete, since the exact and thorough study of all individual forms, in the same manner as Bütschli had done in the case of a small number of species, would require several years. I can therefore only give the basis of further researches.

In the majority of Spyroidea the three essential elements of the Nassellarian skeleton are easily recognisable, viz.:—A, the vertical sagittal ring of the Stephoidea; B, the basal tripod of the Plectoidea, and C, the latticed cephalis of the Cyrtellaria. These three elements are constantly so combined that the primary sagittal ring bisects the cavity of the lattice-shell and produces an external, longitudinal, more or less distinct constriction, separating the lateral inflated halves of the bilocular cephalis; while the three divergent feet of the basal tripod descend from the cortinar or basal plate of the cephalis. The number of the feet is often augmented; they are rarely wanting (probably reduced). The sagittal ring is constantly present, though its relation to the shell-wall exhibits many modifications. The bilocular form of the fenestrated cephalis, with its sagittal constriction, is characteristic of all true Spyroidea, and separates them from the Botryodea (with multilobate cephalis) on the one hand, and from the Cyrtoidea (with simple cephalis) on the other. There are, however, some intermediate forms between these three suborders of Cyrtellaria, which show a direct transition to one another. Almost constantly the length of the three dimensive axes in the cephalis of the Spyroidea is so different that the lateral or frontal axis is the longest, the sagittal or dorso-ventral axis the shortest, and the principal or longitudinal axis intermediate between them. As in all dipleuric or bilaterally symmetrical forms, the two poles of the frontal axis are equal, whilst the two poles of the sagittal and of the principal axis are more or less different. In only a few forms of Spyroidea these differences are difficult to make out; usually the dorsal side is distinctly different from the ventral, and the apical side from the basal. A perfect knowledge of the dipleuric shell requires therefore accurate observation from all six sides.

The suborder of Spyroidea is here divided into four different families. In the first and original family, the Zygospyrida, the shell is represented by the bilocular cephalis only. The second family, Tholospyrida, is distinguished by a fenestrated cupola or galea, a hemispherical or conical dome, arising from the upper or coryphal face of the cephalis. In the third family, Phormospyrida, a thorax is developed, or a fenestrated basal chamber, arising from the lower or basal face of the cephalis. The fourth family, Androspyrida, combines the characters of the second and the third family, a galea arising from the upper and a thorax from the lower face of the cephalis.

The bilocular cephalis, as the most important and constant part of the skeleton in all Spyroidea, requires a further general consideration. Its lattice-work, very variable in the numerous species, is usually more or less different on the six sides of the nut-shaped shell. Its most important part is the horizontal basal plate or lower face, and the sagittal ring arising from it in the vertical median plane. This part of the shell corresponds exactly to the skeleton of the Semantida, and exhibits the same modifications as in that family; it is therefore probable that the majority of the Spyroidea (if not the whole group) have been derived directly from the Semantida by further development of network from the fenestrated ring of Clathrocircus. Whilst in this latter Semantid the two lateral sides of the central capsule remain naked, in the Spyroidea they become enclosed and covered by lattice-work, which arises on the right and left from the ring. In some forms, however, as in the Ceratospyris acuminata of Hertwig, and in my Perispyrida, the frontal ring also of the Coronida and the mitral ring of the Tympanida appear as the initial parts of a complete latticed cephalis.

The basal plate of the cephalis, or the "cortinar plate," the "Basal-Scheibe" of Bütschli, has constantly, according to this author, two pairs of pores, an anterior smaller pair of jugular pores (the pores I of Bütschli) and a posterior larger pair of cardinal pores (the pores II in his description) (compare Pl. 95, figs. 3, 6). This applies to the majority of Spyroidea, but by no means to the whole group. We find basal plates with two pores only (Pl. 95, fig. 1), with three pores (figs. 2, 5), with six pores (figs. 4, 7), with nine pores (Pl. 87, fig. 2), &c. The forms with two basal pores may be derived directly from Semantis (Pl. 92, figs. 1, 2); the common forms with four pores from Semantrum (figs. 3-5); the rarer forms with six pores from Semantidium (figs. 6, 7); and the forms with three pores from Cortiniscus, &c. In this as well as in other respects the variety of different forms and of developmental variations is far greater than Bütschli (1882, loc. cit.) supposed. The bars between the pores of the basal plate possess the same value and the same names as in the Semantida (compare above, p. 954).

The coryphal plate of the cephalis, its upper or apical lattice-plate, does not exhibit such important differences as the opposite basal plate, is far less variable, and is usually pierced by numerous, smaller pores. Some larger pores lie, often in pairs, on the right and left side of the sagittal constriction, and have in some groups a regular form and disposition. Either in the middle of the coryphal plate, or (usually) nearer to its dorsal margin, there arises in majority of Spyroidea an apical horn, directed either vertically or more or less obliquely backwards. In many Spyroidea three horns are developed in the coryphal plate, the odd middle apical horn and two paired frontal or lateral horns arising on each side of the latter and directed more forwards (Pl. 84, figs. 9-12). Sometimes the apical horn disappears, while the two frontal horns remain (Pl. 95, fig. 12).

The anterior or ventral plate of the cephalis (the "Hinterseite" of Bütschli), and the posterior or dorsal side (the "Vorderseite" of that author), exhibit in the majority of Spyroidea more or less marked differences in the number, form, and disposition of their pores, which require a far more accurate description than is here possible. Usually the sagittal constriction of these two plates, produced by the primary vertical ring, is deeper and sharper in the middle of the ventral than in that of the dorsal plate. On both sides of the ring there appear in each plate usually two or three pairs of larger pores, whilst numerous smaller pores are situated towards the lateral sides. We may distinguish the large pores of the ventral plate as facial pores (upper orbital, middle nasal, and lower maxillary pores), and the opposite large pores of the dorsal plate as occipital pores (upper epoccipital, middle mesoccipital, and lower suboccipital pores). A closer comparison of these pores, and of the separating bars in the numerous Spyroidea, may show a regularity of development similar to that offered by the homologies of the parts in the skeleton of the Echinodermata, or of the bones in the skull of the Vertebrata.

The two convex lateral plates of the cephalis, or the right and left sides, both symmetrically equal, do not exhibit the same regularity in the shape, number, and disposition of the pores that the four other sides do. Usually their pores are much smaller and more numerous. In very few forms only a distinct frontal ring is visible (corresponding to that of the Coronida), and in this case the lateral pores are sometimes disposed symmetrically on both sides of this ring. These Spyroidea may have arisen directly from corresponding forms of Coronida.

The sagittal ring (or the primary vertical ring) inherited from the Stephanida and Semantida, shows in the Spyroidea great variety in its form and in its relation to the cephalis. These variations are far greater than Bütschli (1882, loc. cit.) supposed. We may distinguish the following six principal cases:—A. The ring lies completely in the wall of the cephalis, and causes a deep sagittal constriction in it. In this case the lower part or basal segment of the ring separates the basal pores into pairs; its anterior part or ventral rod the facial pores; its upper part or coryphal rod the apical pores; and its posterior part or dorsal rod the occipital pores. No part of the ring is free in the shell-cavity. B. The greater part of the ring lies enclosed in the shell-wall; only its dorsal rod arises free in the shell-cavity and ascends vertically or obliquely to the apex, where it is usually prolonged into the apical horn. This seems to be the most common case. C. The coryphal and the basal part of the ring lie enclosed in the shell-wall, with its dorsal and ventral rod free in the shell-cavity; the dorsal rod ascends vertically to the apex, the ventral rod obliquely upwards to the upper part of the facial plate (the nasal or orbital region). This case, regarded by Bütschli as the usual one, is far less common than he supposed. D. The greater part of the ring lies free in the shell-cavity, its basal rod only is enclosed in the shell-wall, and separates the right from the left group of the cortinar pores. This case seems to be rarer than the preceding. E. The whole ring lies free in the shell-cavity, and is connected with the sagittal constriction of the shell-wall by numerous short radial beams. The distance of the shell-wall from the enclosed ring is usually least on the basal rod, and greatest on the dorsal rod. F. The whole ring lies free in the shell-cavity (as in the preceding case), and is connected by numerous short radial beams with a secondary larger, concentric sagittal ring, which is developed in the longitudinal constriction of the shell-wall.

Comparing these six principal cases, in respect of the relation of the primary sagittal ring to the cephalis of the Spyroidea, we may suppose that they represent together a continuous phylogenetical series of which the first (A) is the original, and the last (F) the latest case; and that the true cause of the various changes is a successive separation of the ring, which becomes more and more free and independent from the lattice-plate of the cephalis. Since the intimate study of the structure of the cephalis in many Spyroidea is very difficult, it requires further accurate observations (from all six sides of the shell). There are also some other more difficult complications of its structure, which cannot be solved without exact study and extended comparison.

Of peculiar importance for the differentiation of the numerous genera of the Spyroidea are the basal feet or the descending and diverging apophyses, which are developed from the base of their cephalis. In general they exhibit the same typical shape which we find in the Plectellaria as well as in the Cyrtoidea, and which we regard therefore as an important common character of the majority of Nassellaria. We may therefore distinguish here also the three primary, and the other ones as secondary feet. The three primary feet, or the "cortinar feet," are the same that we found in the Plagonida and Plectanida, in Cortina and Cortiniscus among the Stephoidea, and in the majority of Cyrtoidea. They also reappear in the same characteristic disposition and connection with the basal plate of the cephalis, in the majority of Spyroidea. The odd caudal foot (c) is the lower prolongation of the dorsal rod of the sagittal ring, which is prolonged upwards into the apical horn. The two paired pectoral feet, however (p′, p′′), are the descending prolongations of the coracal rods (e) which separate the jugular pores (i) from the cardinal pores (k) of the cortinar plate (compare Pl. 95, figs. 2-6).

The three cortinar feet are nearly equally developed in the majority of Spyroidea (compare Pls. 84, 89). But often the odd caudal foot exhibits a different shape from that of the two paired pectoral feet, and in many genera of the group it becomes more or less rudimentary. It has quite disappeared in the subfamily Dipospyrida (Pl. 85); here the two pectoral feet only are developed (often excessively), and are usually opposite in the frontal plane. Such dipodal forms, very common in the Spyroidea, are never found in the Cyrtoidea.

The secondary feet, which we contrast with the three primary feet as later productions, exhibit great variations in number and development. Very frequently three interradial secondary feet become developed, alternating with the three primary cortinar or perradial feet, and opposed to them in pairs (Pl. 95, figs. 7, 8). In these hexapodal Spyroidea (or Hexaspyrida) an odd anterior or sternal foot (z) is opposite to the odd caudal foot, and two paired posterior or tergal feet (t′, t′′) to the two pectoral feet. When the sternal foot becomes rudimentary or lost, pentapodal forms arise (Pentaspyrida, Pl. 95, figs. 9-11), and when the two odd feet (caudal and sternal) disappear, tetrapodal forms arise with two pairs of feet (Therospyrida, Pl. 89, figs. 5, 6). The latter differ from those quadrupedal forms in which two opposite sagittal feet (the caudal and sternal) alternate in the form of a cross with two paired lateral or pectoral feet (Tetraspyrida, Pl. 53, figs. 19, 20). In many Spyroidea the number of basal feet is greatly increased, and they form a regular radial corona around the basal plate, like a circle of tentacles (Polyspyrida, Pl. 87). In this latter case the feet are usually flat and lamellar, whilst in the other groups they exhibit a very variable form, as is seen in Pls. 83-89.

The Central Capsule of the Spyroidea has been accurately observed in only a few genera, and requires a further exact examination regarding the different modifications which occur in their different families. It is very probable that these will agree with the well known modifications in the corresponding groups of Cyrtoidea. The characteristic Monopylean structure of the capsule (with porochora and podoconus) was first described by Richard Hertwig in his Ceratospyris acuminata (loc. cit., p. 72, taf. vii. fig. 2). I found the same afterwards in many other forms. In the Zygospyrida and Tholospyrida the central capsule seems to be usually bilobate (bisected by the sagittal ring), and enclosed in the two chambers of the cephalis. The two lobes of the capsule (right and left) are equal and connected by a smaller middle part, which contains the transverse elliptical nucleus and is enclosed by the sagittal ring; often each lobe contains a large oil-globule. In some forms, however, three or four lobes (sometimes perhaps more) are developed, which pierce the cortinar pores of the basal plate and depend freely between the basal feet (Pl. 53, fig. 19). In the Phormospyrida and Androspyrida the formation of such basal lobes seems to be more frequent; they are here enclosed by the thorax. Usually each lobe contains a large oil-globule. In Nephrospyris (Pl. 90) and in some similar genera the central capsule is violin-shaped, deeply bisected by the sagittal ring; in each of the two lobes a series of oil-vesicles is developed at both poles of the transverse axis (figs. 7, 10); the nucleus, a transverse cylindrical body, is placed in the latter. In the periphery of the voluminous calymma of this singular genus a large number of globular cells are developed (figs. 7, 10), probably symbiotic xanthellæ. In some forms of Nephrospyris these xanthellæ are enclosed by a peculiar inflated girdle, developed in the periphery of the kidney-shaped skeleton (Pl. 90, figs. 1, 4-6).

Synopsis of the Families of Spyroidea.

Family LII. Zygospyrida, n. fam. (Pls. 84-87).

Definition.—Spyroidea without galea and thorax; the shell consisting of the bilocular cephalis only and its apophyses.

The family Zygospyrida is by far the richest among the four families of Spyroidea, the number of genera in the whole suborder amounting to forty-five, in the former to twenty-eight; and the number of species in the latter to two hundred and thirty-seven, in the former to one hundred and seventy-two. The number of individuals also found in many species of Zygospyrida is far greater than in any species of the three other families. The shell of the Zygospyrida is represented by the bilocular cephalis only, and never develops a galea (as in the Tholospyrida) nor a thorax (as in the Phormospyrida and Androspyrida). The three latter families have therefore been derived from the former as their common ancestral group.

The Zygospyrida are very similar and nearly related to the Monocyrtida, and in both groups the cephalis alone represents the whole shell. Therefore in 1882, Bütschli, in the paper mentioned above (p. 1016), maintained the opinion, that these two groups were identical. But there is this important difference between them, that in the Zygospyrida (as in all Spyroidea) the cephalis is bilocular, with a sagittal constriction, separating the right and left chambers. In the Monocyrtida, however (as in all Cyrtoidea), the cephalis is unilocular, forming a quite simple chamber without sagittal constriction. Correspondingly the primary sagittal ring in all Zygospyrida is well preserved and usually complete, while in the Monocyrtida it is never complete, and often quite absent. Another difference is indicated by the form of the central capsule, which in the Zygospyrida is usually bilobed, and more developed in the frontal axis, whilst in the Monocyrtida it is commonly ovate, and more developed in the principal axis.

The important questions of the origin and phylogenetical relation of these two similar groups of Nassellaria form a very complicated and difficult problem, and we do not at present possess the means of solving it. It may be that a part of the Monocyrtida has been derived from the Zygospyrida (as Bütschli erroneously supposes for all), but the contrary is also possible. A third possibility is the independent origin of both groups from the Semantida. But we shall see afterwards, in the description of the Monocyrtida, that a great part of this group may with greater probability be derived from the Plectoidea than from the Stephoidea, and that another part of them has probably been derived quite independently from the Nassellida. Regarding the complicated relations of these similar groups, mentioned above (pp. 892-894), it seems hopeless here to discuss further their difficult affinities; but in any case it seems useful or even necessary to separate the Monocyrtida from the Zygospyrida.

The cephalis of the Zygospyrida possesses all the characteristic features which we have described above as belonging to the Spyroidea in general (p. 1017), and a survey of the Pls. 84-87 exhibits the most important modifications of its shape (compare also Pl. 95, figs. 1-13). Therefore we may avoid useless repetitions by referring to the description given above. We will only point out that the bilocular character of the cephalis (in opposition to the simple cephalis of the Monocyrtida) in the majority of Zygospyrida is apparent on the first view, and that it is always demonstrated by an accurate examination of the sagittal ring and its relation to the longitudinal constriction of the cephalis (compare p. 1019). The basal plate or cortinar plate exhibits usually four typical pores (two smaller jugular, i, and two larger cardinal, k), but this is by no means a general rule, as Bütschli thought; moreover, instead of four cortinar pores, there are frequently found three or six or some other numbers (compare above, p. 1018).

The top of the cephalis in the majority of Zygospyrida bears an apical horn, as an upper prolongation of the dorsal rod of the sagittal ring; it is often of unusual size and shape (Pl. 85, figs. 5-11; Pl. 95, fig. 8, &c.). In several genera two lateral or frontal horns are developed on both sides of the former (Pl. 84, figs. 9-12; Pl. 86, figs. 5-13). Sometimes these two alone are present, while the original apical horn is lost by reduction (Pl. 95, fig. 12). In other genera the apical face bears no horns; probably they are lost by reduction.

The greatest variety of forms is produced in the Zygospyrida by the different number, disposition, and shape of the feet, or the basal apophyses arising from the base of the cephalis. Regarding these remarkable differences, we may distinguish eight subfamilies. The common ancestral group of all are probably the Tripospyrida, which possess the three typical basal feet of Cortina and Cortiniscus, an odd caudal and two paired pectoral feet (Pl. 84). From these may be derived the Dipospyrida (Pl. 85) by loss of the caudal foot, the Tetraspyrida by development of an odd sternal foot (Pl. 53, figs. 7, 8, 19), and the Hexaspyrida by development of three secondary interradial between the three primary perradial feet. The Pentaspyrida may have been derived from the latter by loss of the sternal foot (Pl. 95, figs. 9-11), and the Therospyrida by loss of the two sagittal feet (Pl. 89, figs. 5, 6). The Polyspyrida bear numerous (seven to twelve or more, often twenty to thirty) basal feet, forming a corona around the cortinar plate (Pl. 87); they may have been derived from the Tripospyrida or Hexaspyrida by further multiplication of the feet. Finally, the Circospyrida exhibit no feet at all (Pl. 89, figs. 7-12); they may be derived either from one of the preceding groups by complete reduction and loss of the basal apophyses, or directly from those Semantida which have no feet (Semantis, Semantrum, Clathrocircus, &c.).