The parts of the extracapsular body, which are enclosed in the gelatinous calymma possess a peculiar importance in the Phæodaria; these are firstly the sarcomatrix and the pseudopodia arising from it, and secondly the phæodium. The sarcomatrix, or the layer of extracapsular sarcode (ectoplasm), which immediately surrounds the central capsule is very thick and more strongly developed in all Phæodaria than in all other Radiolaria. Its extraordinary size has been already mentioned by Hertwig (1879, loc. cit., p. 99). It is in direct connection with the intracapsular sarcode (or the endoplasm) only by the openings of the central capsule, and mainly by the astropyle. Very numerous radial pseudopodia arise everywhere from the sarcomatrix and run to the surface of the calymma, usually forming a rich network in it by means of numerous branches and anastomoses (compare Pl. 101-104). On the surface of the calymma the meshes of this network are very numerous, and there arise from its nodal points the terminal pseudopodia, which float freely in the surrounding water. The metamorphoses of this network of sarcode, the perpetual changes in the number and size of its meshes, and the movements of the sarcode streams as well as of the small granules running in it, are always very manifest in the Phæodaria and in the big forms of this legion (mainly in the Aulacanthida) they are better observed than in the majority of other Radiolaria. In many Phæodaria (and perhaps in all) a part of the pseudopodia seems to have undergone a local differentiation, for special physiological purposes; and Hertwig has described a peculiar conical contractile body, which arises in Cœlodendrum between the two parapylæ (loc. cit., p. 100, fig. 3). Further examination of these interesting organs in living Phæodaria is required. Usually the sarcode, issuing from the openings of the capsule, forms a stronger cylinder, with peculiar movements.

The phæodium, or the peculiar dark extracapsular pigment-body of the Phæodaria, is one of the most important and most characteristic parts of their organisation, and has induced me to derive their name from it; it is not less typical for the whole legion, than the astropyle with its radiate operculum and the proboscis; and both these important parts are always in direct topographical and physiological connection. Whilst I have missed the phæodium in no Phæodarium in which the soft body was well preserved, I have not found it in any other Radiolaria; for the similar extracapsular pigment bodies, which are found in some species of Thalassicolla and some other Sphærellaria, have a composition and signification different from that of the phæodium—an exclusive peculiarity of the Phæodaria.

The typical importance of the phæodium for all members of this legion is proved by the following five facts:—(1) its constant presence in all Phæodaria; (2) its constant excentric position in the oral half of the calymma; (3) its constant relation to the astropyle, the operculum and the proboscis of which is always surrounded by it; (4) its constant considerable size, its volume being usually greater than that of the central capsule, the aboral half of which is covered by it; (5) its constant colour and morphological as well as chemical composition. These five facts together demonstrate by their absolute constancy the high morphological and physiological importance of the phæodium for this peculiar subclass of Radiolaria, although its true nature is difficult to make out, and its main function is not yet sufficiently known.

The first remarks that have been made on the phæodium were published in 1862 in my Monograph, where I noticed the peculiar dark brown extracapsular pigment body and its excentric position covering only one-half of the central capsule, in Aulacantha (p. 263, Taf. ii. figs. 1, 2), in Thalassoplancta (p. 262, Taf. iii. fig. 10), and in Cœlodendrum (p. 361, Taf. xxxii. fig. 1). Its general presence and peculiar composition were first recognised by Dr. John Murray, who had, during the Challenger Expedition, the first opportunity of examining many big living Phæodaria brought up from great depths. He gives in his first Report (1876, loc. cit., p. 536), the following important notice:—"The sarcode of all these deep-sea Rhizopods has many large black-brown pigment-cells. Small bioplasts are scattered through the sarcode. These collect into capsular-like clumps when the animal is at rest, and are quickly coloured by carmine." In 1877 Dr. John Murray sent me the wonderful collection of Phæodaria brought home by the Challenger, and I had now the best opportunity for examining the phæodium in hundreds of well-preserved specimens. Supported by these extensive observations, I gave, in 1879, in a preliminary paper,^[273] a fuller description of the phæodium, and of the phæodella (or the peculiar dark pigment-granules composing it). At the same time R. Hertwig published his observations on the big living Tripylea examined by him at Messina, and pointed out particularly that the constant presence, composition, and arrangement of this excentric extracapsular pigment-body was most characteristic, and sufficient in itself to distinguish this group from all the other Radiolaria (1879, loc. cit., p. 99).

The most striking peculiarity of the phæodium, and the most important on account of its absolute constancy, is its excentric position, covering only the oral hemisphere of the central capsule, and wholly or partly wanting on the aboral hemisphere. This constant topographical relation to the capsule never fails, and may be always observed at the first glance, when the body is seen from the side (the main axis of the capsule being perpendicular to the axis of the eye of the observer), (compare Pl. 101, figs. 6, 10; Pl. 102, fig. 1; Pl. 103, fig. 1, &c.). The phæodium envelops, therefore, the oral half of the capsule completely, and especially the astropyle in its centre; hence, the radiate operculum and the proboscis arising from the latter cannot be seen usually before the former is removed. The general form of the entire phæodium, in consequence of this excentric position, is concavo-convex (or crescentic in longitudinal section), its concave face embracing the oral or anterior face of the capsule, and its convex face being turned to the surface of the calymma.

The topographical relation of the phæodium to the surrounding shell is also very characteristic in the suborder Phæogromia, or in those Phæodaria which possess a peculiar shell-mouth placed on the oral pole of its main axis. Here the capsule is always placed in the aboral half of the shell-cavity, the phæodium in its oral half, separating the astropyle from the mouth of the shell, both of which lie in the main axis; as in the Challengerida (Pl. 99), Tuscarorida (Pl. 100), Castanellida (Pl. 113), Circoporida (Pl. 115), and Medusettida (Pl. 118-120). In this suborder (the Phæogromia), the volume of the phæodium may be, on an average, about as great as that of the central capsule, whilst in the majority of other Phæodaria it is much greater than the latter.

A similar topographical relation between the phæodium and the enclosing shell, as in the Phæogromia, also exists in the suborder Phæoconchia, or in those Phæodaria, which are distinguished from all the others by the possession of a bivalved shell (Pl. 121-128). The two valves of this curious shell, which resembles that of the Brachiopoda, are dorsal and ventral, and the tripylean central capsule is always so placed between them that its two lateral parapylæ (right and left) lie in the frontal plane of the shell, where a large frontal fissure opens between the opposed margins of the two hemispherical or cap-shaped valves. The phæodium is also placed here on the oral half of the capsule and surrounds its astropyle; but it exhibits some differences in the three families of Phæoconchia.

The Concharida, the bivalved shell of which is simple, and without tubular apophyses (Pl. 123-125), possess a relatively small central capsule, which usually fills up only the third or fourth part of the shell-cavity. This is the aboral or posterior part, on the apex of which both valves are united by a ligament in some Concharida (Pl. 123, figs. 8, 9). The oral or anterior part of the shell-cavity (usually two-thirds or more) is filled up by the phæodium, and this is usually bifid, being divided by a frontal constriction into two wings or lobes; the dorsal wing is hidden in the upper valve of the shell, the ventral wing in the lower valve; both wings are usually united only by a small central bridge, and this bridge of the phæodium is pierced in its centre by the proboscis of the astropyle (Pl. 124, figs. 6, 10; Pl. 123, figs. 8, 9).

The Cœlodendrida have a different shape (Pl. 121). Their bivalved shell is relatively small and tiny, and bears on the two poles of the sagittal axis two conical apophyses or galeæ, from each of which three or four very large, dichotomously branched tubes arise. The central capsule fills up the cavity of the bivalved shell almost entirely, and the voluminous dark phæodium envelops both to such an extent that the shell and the enclosed capsule are often hidden in it completely. Therefore I arrived in my first description of Cœlodendrum (1862, loc. cit.) at the erroneous conclusion that the capsule lies outside, not inside the shell. The first accurate figure and description of its structure was given in 1879, by Hertwig (loc. cit., p. 99, Taf. x. fig. 3). The central capsule (v) is here separated from the bivalved shell (m) only by a very small distance, and the oral part of both is hidden in the phæodium. I find, however, in the majority of the numerous preparations of the Challenger collection, the volume of the phæodium much greater, and it often envelops the entire shell.

The Cœlographida, finally, have a phæodium of the most remarkable shape, since in their bivalved shell a peculiar reserve store or magazine of phæodella, which we call the "phæocapsa" is developed for it (Pl. 126-128, g.t.m.). The bivalved shell has in these most perfect Phæodaria a structure similar to that in the Cœlodendrida; but they differ from the latter in the stronger development, and greater differentiation of the two apical galeæ, and the large hollow tubes arising from them. These two helmet-shaped cupolæ, the galeæ (g), which arise from the two valves on the poles of the sagittal axis, are in the Cœlographida usually larger than the valves themselves, and are not closed, as in the Cœlodendrida, but open by a tubular apophysis at their base, the nasal tube or rhinocanna (t). The apex of the galea is connected with the open mouth of the rhinocanna by a single or double frenulum (b). The two nasal tubes or rhinocannæ (a dorsal and a ventral) lie in the sagittal plane of the body and run from the base of each galea along the anterior convexity of the valve to its oral margin. Here is placed the proboscis of the astropyle, between the two opposed mouths of the rhinocannæ (Pl. 128, fig. 2). The phæodium is usually hidden entirely in the two phæocapsæ, which are composed of the two galeæ (g), and the two rhinocannæ arising from them (Pl. 127, figs. 4-9). A part of the phæodella is usually thrown out by the mouth of the latter (m).

The characteristic colour of the phæodium exhibits numerous different tints between green, brown, and black. It seems to be in the majority blackish-brown or greenish-brown, very often olive, more rarely almost quite green or red-brown. Usually the colour is so dark, intense and opaque, that the parts enclosed by the phæodium, mainly the oral hemisphere of the central capsule and the astropyle, are completely hidden in it. The chemical composition of the phæodium demands further accurate researches; unfortunately I have not been able to make out its true nature, since numerous different experiments furnished no certain general results.

The phæodella, or the pigment-corpuscles, which compose the phæodium, aggregated in hundreds, and in the bigger species in thousands, are usually spherical, sometimes somewhat ellipsoidal, at other times spheroidal or lenticular; but usually numerous smaller, irregular, roundish particles are intermingled between the larger and more regular corpuscles, and often the main mass forms a very fine black powder. The size of the phæodella is very variable, not only in the different species, but also in one and the same individual. The larger phæodella have a diameter of 0.01 to 0.02, the smaller of 0.004 to 0.008 mm.; but there also occur very big forms of 0.04 to 0.05 mm., or even more, and very small ones of 0.001 mm. or less. Usually the phæodium appears as an aggregate of numerous larger and smaller phæodella, which are very different in size as well as in the intensity and tint of their colour, and are irregularly crowded in a black, powder-like substance.

The morphological nature of the phæodella is also difficult to make out. I have already pointed out in my first description of Aulacantha, Thalassoplancta and Cœlodendrum (1862, loc. cit.), that a great part of these pigment-corpuscles are true cells, composed of a nucleus and protoplasm, which contains granules of pigment, and is enveloped by a membrane. Dr. John Murray, who had during the Challenger voyage the opportunity of examining numerous different living Phæodaria, and staining them by carmine, also asserts that a great part of those dark corpuscles are "large black-brown pigment-cells" (1876, loc. cit., p. 536). Numerous preparations of the Challenger collection, well preserved in glycerine, and stained by carmine, contain Phæodaria belonging to different families, the phæodium of which contains numerous such "pigment-cells," with a dark red nucleus, and so similar are these cells, that every histologist should recognise them. But in strange contrast to this is the fact, that in numerous other mountings, prepared in the same manner, not a single cell of this kind is found in the phæodium, and that the latter is composed only of irregular pigment-granules. In many Phæodaria belonging to different families I, like Hertwig, could not find a single true nucleated cell in the phæodium.

A great part of the Phæodaria, and usually the bigger forms of Aulacanthida, Cannorrhaphida, Cœlodendrida, Cœlographida, &c., exhibit a peculiar structure of the larger phæodella, viz., a fine parallel striation (Pl. 101, figs. 3, 6; Pl. 102, fig. 1; Pl. 103, fig. 1; Pl. 104, figs. 1-3, &c.). In each phæodellum may be counted about ten to twenty such fine parallel stripes (more in the greater, less in the smaller forms); and in the ellipsoidal phæodella the stripes are either transverse rings, perpendicular to their main axis, or ascending obliquely; they often resemble the convoluted spiral filament of a thread-cell or nettle-cell of an Acaleph. Sometimes these parallel transverse stripes are very striking. Another structure is seen in larger phæodella, namely an aggregate or cluster of smaller globules, often of equal size, resembling a small morula. All these minute structures of the phæodella as well as their changes in the living Phæodaria, require a far more extensive examination (by means of strong lenses and different chemical reagents), than I could, unfortunately, devote to them.

The physiological signification of the phæodella, therefore, is at present not yet known; but the general facts quoted above, their constant presence, position, volume, and composition, make it probable that their physiological value in the Phæodarian organism is very great. The following hypotheses may be taken provisionally into consideration:—A. The phæodella are peculiar symbiontes, or unicellular algæ, comparable to the xanthellæ or zooxanthellæ of the other Radiolaria. This hypothesis is probably correct for those phæodella which are true nucleated cells; and the more so, as the majority of Phæodaria do not exhibit those common yellow xanthellæ, which are usually found in the Spumellaria and Nassellaria. It is even possible that the latter are absent in all Phæodaria. B. The phæodella are dark pigment-bodies, which absorb light and heat in a manner similar to the simple "pigment-eyes" of many lower animals, and may therefore be optical sense-organs of the Phæodaria. This hypothesis may be supported by a comparison with the large-eyed unicellular Protist, Erythropsis agilis, described by R. Hertwig. C. The phæodella are organs of nutrition of the Phæodaria and active in their metastasis ("Stoffwechsel"). Regarding them from this point of view, we may suppose that the phæodella are secreted products which serve for digestion, acting like the bile or the saliva of higher animals. Perhaps they too act like the venomous matter produced in the thread-capsules of the Acalephæ. The suggestion that they are mere excretions, or half-digested matters, as Hertwig supposes ("halb assimilirte Nahrungs-bestandtheile," 1879, loc. cit., p. 99) seems less probable. The most important fact illustrating their high signification for the processes of nutrition, digestion and for effecting changes on matter, seems to be the close relation of the phæodium to the astropyle; the radiate operculum of the latter, and the proboscis arising from it, being constantly covered and completely hidden by the central main mass of the phæodium.

The skeleton of the Phæodaria is always extracapsular, and exhibits in the majority of this legion such a characteristic shape, form, and structure, that these organisms may be easily recognised by it, even apart from the central capsule and the phæodium. In a few cases, however, the skeleton is so similar to that of some Nassellaria and Spumellaria, that it may be accidentally confounded with it. In general the skeleton of the Phæodaria is much larger, and much more highly developed, than that of most other Radiolaria, and exhibits the most wonderful appearances, and the most marvellous complications, which are found in the whole world of Protists, or of unicellular organisms. The varied composition and differentiation of the skeleton alone distinguishes the numerous families, genera, and species of Phæodaria described in the sequel; all the fifteen families, however, agree so completely in the structure of the central capsule and the phæodium described, that we may derive them all phylogenetically from a small skeletonless family, the Phæodinida.

The chemical composition of the skeleton seems to be, in the majority of Phæodaria, somewhat different from that of the other Radiolaria. In a few groups only, especially in the Cannobelida (Dictyocha, Mesocena, &c.), and in a part of the Castanellida and Concharida, the substance of the skeleton seems to be of pure silica, as in the Nassellaria and Spumellaria; these flinty skeletons, therefore, may be also found fossil. In the majority of Phæodaria, however, the skeleton does not consist of pure silica, but of an organic silicate; it becomes more or less intensely stained by carmine, and browned or blacked by fire; in many cases it even becomes completely burned and destroyed by the prolonged action of heat. This circumstance explains why Phæodaria in general are rare in deep-sea deposits, as in the common Radiolarian ooze of the Pacific, and why they are generally absent in fossil deposits. Even the pure Radiolarian rocks of the Barbados, &c., contain only a few Phæodaria, mainly Dictyochida.

According to the different forms of the skeleton, we may divide the legion or subclass of Phæodaria into two sublegions, four orders, and fifteen families. Firstly, we may distinguish as two groups the Phæocystina, without a lattice-shell, and the Phæocoscina, with a lattice-shell (compare above, p. 5). The Phæocystina comprise three different families, viz., (1) Phæodinida, without any skeleton (Pl. 101, figs. 1, 2); (2) Cannorrhaphida, with an incomplete skeleton, composed of numerous separate, not radial pieces, which are scattered around the capsule in the calymma (Pl. 101, figs. 3-14; Pl. 114, figs. 7-13), and (3) Aulacanthida (Pl. 102-105), with an incomplete skeleton, composed of numerous hollow radial tubes, which pierce the calymma and come in contact by their proximal ends with the surface of the central capsule.

The Phæocoscina, or the Phæodaria with a lattice-shell (embracing the great majority of the whole legion) exhibit three principal differences in the shape of their shell, and from these we distinguish the three following orders; (A) Phæosphæria, with a spherical, not bivalved shell (rarely of an ellipsoidal or lenticular, or another modified form), without a shell-mouth or a peculiar constant large opening on the lattice-shell; (B) Phæogromia, with an ovate or polyhedral, not bivalved shell (often also of a subspherical, ellipsoidal, or another modified form), constantly provided with a shell-mouth or a peculiar large opening on one pole of the main axis of the lattice-shell; (C) Phæoconchia, with a bivalved shell, composed of two completely separated, hemispherical, cap-shaped or boat-shaped valves (a dorsal and a ventral), comparable to that of the Brachiopoda.

The Phæosphæria, or those Phæodaria the big shell of which is usually spherical, never bivalved and never provided with a peculiar shell-mouth, comprise a great number of common and large-sized Phæodaria, which may be arranged into four different families, according to the different structure of the shell—(1) Orosphærida (Pl. 106, 107), spherical shell extremely big and robust, composed of single piece of coarse lattice-work, the thick bars of which are stratified and contain partly a fine axial-canal, meshes of the network usually irregularly polygonal, no astral septa in the nodal points; (2) Sagosphærida (Pl. 108), spherical shell large-sized, but extremely delicate and fragile, composed of a single piece of arachnoidal lattice-work, the thin bars of which are simple solid threads, without axial-canal, meshes of the network always large and triangular, no astral septa in the nodal points; (3) Aulosphærida (Pl. 109-111), spherical shell large-sized, but very fragile, composed of numerous hollow cylindrical tubes, which are connected (and at the same time separated) by peculiar astral septa in the nodal points, meshes either triangular or polygonal; (4) Cannosphærida (Pl. 112), spherical shell double, composed of two concentric shells which are connected by thin hollow radial tubes, the inner shell simple, solid or fenestrated, with a shell-mouth on the basal pole, the outer shell composed of hollow cylindrical tubes which are connected by astral septa in the nodal points. The structure of this outer shell is the same as in the Aulosphærida, while the basal mouth of the inner shell brings this family in closer relationship to the Phæogromia.

The Phæogromia, or those Phæodaria the shell of which is not bivalved, but provided with a peculiar constant mouth on the oral pole of the main axis, are in general similar to the Nassellaria (Monocyrtida), and may be divided into five different families, viz., (1) Challengerida (Pl. 99), shell ovate or subspherical, also often triangular or lenticular, distinguished by a peculiar diatomaceous structure, an exceedingly fine tracery of regular hexagonal, very delicate network; (2) Medusettida (Pl. 118-120), shell ovate, campanulate or cap-shaped, distinguished by a peculiar alveolar structure, with a corona of peculiar hollow, large, articulated feet around the mouth; (3) Castanellida (Pl. 113), shell spherical or subspherical, of ordinary simple lattice-work, usually with a corona of simple solid teeth around the mouth; (4) Circoporida (Pl. 114-117), shell spherical or polyhedral, with panelled or dimpled surface, distinguished by a peculiar porcellanous structure (numerous thin needles being embedded in a punctulate cement-substance), with hollow radial spines and with a corona of simple solid teeth around the mouth; (5) Tuscarorida (Pl. 100), shell ovate or subspherical, with smooth surface, of the same peculiar porcellanous structure as the Circoporida, but with hollow, very long tubular teeth around the mouth.

The Phæoconchia are the peculiar and most interesting "Phæodaria bivalva," differing from all other Phæodaria, and from all known Radiolaria in general, in the possession of a bivalved lattice-shell, composed of a dorsal and a ventral valve. They may be divided into three families: (1) Concharida (Pl. 123-125), shell with two thick and firm, regularly latticed valves, which bear no hollow tubes and no cupola or galea on their apex or sagittal pole; (2) Cœlodendrida (Pl. 121), shell with two thin and fragile, scarcely latticed valves, which bear a conical cupola or a helmet-shaped galea on their apex, and hollow branched tubes arising from it (without rhinocanna and frenula); (3) Cœlographida (Pl. 122, 126-128), shell with two thin and fragile, scarcely latticed valves, similar to those of the Cœlodendrida, but differing from them in the development of a peculiar rhinocanna or nasal tube upon each valve; this tube is connected by an odd or paired frenulum with the apex of the galea, and both together contain the phæodium.

The phylogenetic affinity of the fifteen families enumerated, and the morphological relationship based upon it, form a very difficult problem. The whole legion of Phæodaria is probably monophyletic, in as much as all the families may be derived from a single ancestral group, the skeletonless Phæodinida (Phæodina and Phæocolla); but at the same time polyphyletic, in as much as probably many families have been derived, independently one from another, from different branches of Phæodinida; or in other words, the characteristic malacoma of the Phæodaria (the cannopylean central capsule and the calymma with the phæodium) may be a monophyletic product, inherited from a single ancestral form; the manifold skeleton, however, is certainly a polyphyletic product, originating from different skeletonless Phæodinida.

Among the independent families of Phæodaria, derived directly from skeletonless Phæodinida by production of a peculiar skeleton, may be the following: Cannorrhaphida (Pl. 101, probably polyphyletic), Aulacanthida (Pl. 102-105), Castanellida (Pl. 113), Challengerida (Pl. 99), Concharida (Pl. 123-125), Circoporida (Pl. 114-117) and Tuscarorida (Pl. 100). The four families of Phæosphæria (the Orosphærida, Sagosphærida, Aulosphærida and Cannosphærida (Pl. 106-112), may be derived perhaps from the Castanellida; and the Medusettida (Pl. 118-120), have been perhaps derived from the Challengerida. The complicated affinities of these groups are however difficult to explain. The Cœlodendrida (Pl. 121) are probably derived from the Concharida, and the Cœlographida (Pl. 126-128) from the Cœlodendrida.

The geometrical fundamental form of the shell is in the majority of Phæodaria monaxonial, corresponding to the main axis of the enclosed central capsule; the astropyle of the latter, placed on the oral pole of the main axis, corresponds to the mouth of the shell in all Phæogromia. In the Phæosphæria, where no peculiar shell mouth is developed, the general fundamental form of the shell is usually homaxonial or spherical, often an endospherical polyhedron, rarely ellipsoidal or spindle-shaped (with prolonged main axis), or lenticular (with shortened main axis). The bivalved Phæoconchia have usually either an amphithect shell (with the same fundamental form as the Ctenophora), or a dipleuric, bilaterally symmetrical shell (with a dorsal and a ventral valve, a right and a left parapyle). A small number of Phæodaria (mainly Circoporida) are remarkable on account of the regular polyhedral form of their shell, the geometrical axes of which resemble crystalline axes and are defined by regular radial tubes; as the octahedral Circoporus (Pl. 117, fig. 6), the dodecahedral Circorrhegma (fig. 2), and the icosahedral Circogonia (fig. 1).

The siliceous or silicate bars, which compose the skeleton of the Phæodaria, are in the majority hollow tubes, filled up by jelly; in some other families, however, they are solid rods, as in the Nassellaria and Spumellaria. Such usual lattice-work, composed of solid rods, occurs only in the families Sagosphærida (Pl. 108), Castanellida (Pl. 113), and Concharida (Pls. 123-125). A quite peculiar structure, a diatomaceous tracery of extremely fine and regular hexagonal frames, distinguishes the Challengerida (Pl. 99). The hollow cylindrical tubes, which are found in the other families, appear in three different forms, simple, articulate, and provided with an axial thread. Simple hollow tubes, which are neither articulate nor provided with an axial thread, occur in the Cannorrhaphida (Pl. 101), Aulacanthida (Pl. 102-105), Cœlodendrida (Pl. 121, 122), and Cœlographida (Pl. 126-128). In all these families the hollow cylindrical tubes have a very thin wall and contain a wide cavity, filled only by jelly. The Orosphærida (Pl. 106, 107), differ in the reduction of the cavity, which becomes very narrow (often rudimentary or lost), whilst the walls of the tubes become extremely thickened and stratified, numerous concentric layers of silica being disposed one over the other. The hollow cylindrical tubes contain an axial filament, or a thin thread of silica, placed in its axis, in the families Aulosphærida (Pl. 109-111), Cannosphærida (Pl. 112), Circoporida (Pl. 114-117), and Tuscarorida (Pl. 100). Usually the axial filament is connected with the thin wall of the tube by numerous horizontal branches. A quite peculiar structure distinguishes the Medusettida (Pl. 118-120); their hollow tubes, extremely prolonged, are articulate owing to the presence of numerous, regular, equidistant transverse septa; these are pierced by a short tubule, similar to the siphon of the shells of Nautilus; this remarkable alveolar structure also occurs in the peripheral part of their shell-wall (and sometimes in the whole shell), numerous small polyhedral chambers or alveoles which communicate by small openings, being developed; they become easily filled with air in the dry shell (Pl. 120, figs. 11-16).

The substance of the siliceous or silicated shell-wall is, in the majority of Phæodaria, homogeneous and structureless, as in the Spumellaria and Nassellaria; but sometimes it acquires a peculiar structure. The thickened wall of the hollow tubes in the Orosphærida and in several Aulacanthida (Pl. 105, figs. 6-10) becomes distinctly stratified, concentric strata being disposed one over the other. A very remarkable structure, differing from that in all other Radiolaria, is found in the porcellanous shell of the Circoporida (Pl. 114-117), and Tuscarorida (Pl. 100). The thickened wall of the opaque shell is here composed of a peculiar silicated cement, which encloses numerous very thin and irregularly scattered needles (Pl. 115, figs. 6-9; Pl. 116, fig. 3). Dry fragments of these shells, observed by a strong lens, appear finely punctulate, and probably air, entering into these fine porules of the cement, causes the white colour and the calcareous or porcellanous appearance of the opaque dry shell. Its surface is smooth in the Tuscarorida (Pl. 100), panelled in the Circoporida (Pl. 114-117).

The hollow or solid spines, which arise from the shell of the Phæodaria, exhibit an extraordinary variety and elegance in the production of different branches, bristles, hairs, secondary spine, and thorns, hooks, anchor-threads, pencils, spathillæ, &c. These appendages are developed similarly to those of many Spumellaria, but exhibit a far greater variety and richness in form. They are organs partly for protection, partly for retention of food. They are much more interesting than in other Radiolaria.

Synopsis of the Orders and Families of Phæodaria.

Order I. PHÆOCYSTINA, Haeckel (1879).

Definition.—Phæodaria without lattice-shell, either without any skeleton, or with an incomplete skeleton, composed of numerous single pieces, which are scattered in the calymma without connection. Central capsule placed in the centre of the spherical calymma.