132. Dictyosis of the Acantharia.—The lattice-structures of the Acantharia differ essentially from those of other Radiolaria in several particulars. Firstly, they consist not of silica but of acanthin (§ 102); secondly, they are always secondary formations, usually developed from transverse processes of the primary centrogenous radial spines; thirdly, their formation is not simultaneous (at the same time over the same shell), but successive (proceeding from the individual radial spines tangentially towards the middle of the intervals); fourthly, the configuration of the network is due to the relative position of the spines and the mode of union of their transverse apophyses. Since they are at right angles to the spines, and since the branches of the apophyses are at right angles to them, the original ground-form of their dictyosis is a lattice-work with quadrangular meshes; these are often quite regular and square (Pl. 130, figs. 5, 6; Pl. 136, figs. 2, 9, &c.); more commonly they are rectangular or irregularly quadrangular (Pl. 131, fig. 10; Pl. 133, figs. 2, 3, &c.). In the majority of the Acantharia the quadrangular form of the meshes passes over into an irregularly polygonal or roundish one (Pls. 137, 138). Very often the primary meshes of the lattice-shells, which immediately surround the radial spines, are larger and more regular ("aspinal pores"), whilst the numerous secondary meshes between them are smaller and irregular ("coronal pores"; Pl. 135, figs. 1-4, &c.).
133. Dictyosis of the Nassellaria.—The siliceous lattice-structures of the Nassellaria are formed on the whole like those of the Spumellaria, with which they were formerly united under the name "Polycystina." In this group also there may be distinguished as two main forms the regular and irregular. In the Nassellaria the regular lattice-structures generally exhibit hexagonal or circular meshes, whilst the irregular are either polygonal or roundish; the irregular forms are, however, much more abundant than the regular, and a further distinction from the Spumellaria consists in the fact that the primary skeletal elements, from which the lattice is secondarily developed, exercise a predominant influence upon their form. These primary elements in the majority of the Nassellaria are to be seen in two morphologically most important structures:—first, the primary sagittal ring, which embraces the central capsule in the median plane (§ 124); and secondly, the basal tripod (§ 125), whose three diverging rays proceed from the base of the central capsule, whilst commonly a fourth vertical ray supports the dorsal side of latter (compare Pls. 81-91, p. 892). In the majority of the Nassellaria these two primary elements appear in combination, whilst in others only one of them is recognisable. In addition there occur numerous monaxon lattice-shells in which neither of these elements can be recognised, but a simple ovoid lattice-shell (cephalis) alone forms the whole skeleton or its primary part (Pl. 51, fig. 13; Pl. 98, fig. 13). The great difficulty in the morphological interpretation and phylogenetic derivation of the Nassellarian skeleton lies in the fact that each of these three elements—the primary sagittal ring, the basal tripod, and the latticed cephalis—may form the whole skeleton by itself or be combined with one or both of the others (p. 893). Even nearly related or at all events very similar forms may differ very greatly in this respect. With regard to the manifold forms of their dictyosis it follows that it is partly dependent upon one of the two first elements, partly independent. In the Plectellaria (or those Nassellaria which do not possess a complete lattice-shell) the lattice-work is usually irregular and arises by union of the ramifications, which proceed either from the primary sagittal ring (Pls. 81, 82, 92-94) or from the basal tripod (Pl. 91). In the Cyrtellaria (or Nassellaria with a complete lattice-shell, Pls. 51-80), on the other hand, the lattice-work is sometimes regular, sometimes irregular, being often very different in the different joints of a segmented shell (Pl. 72); a great part of it arises independently of the two chief morphological elements, and develops according to laws similar to those which regulate the dictyosis of the Spumellaria.
134. Dictyosis of the Phæodaria.—The lattice-structures of the Phæodaria, which consist of a silicate of carbon (§ 102), are on the whole not developed in such variety as those of the other Radiolaria, but exhibit several essentially different types of structure, not reducible to a common primitive type of lattice-work. In one portion of this legion there occurs an ordinary simple lattice-work (as in Spumellaria and Nassellaria), with solid trabeculæ; of these the Castanellida (Pl. 113) and Concharida (Pls. 123-125) have usually regular or subregular, circular meshes, sometimes hexagonally framed; the Orosphærida (Pls. 106, 107) large irregular polygonal meshes with thick trabeculæ, the Sagosphærida (Pl. 108) large triangular meshes with thin filiform trabeculæ. The Challengerida (Pl. 99) are characterised by a very delicate regular lattice-work, with minute hexagonal pores, like a Diatomaceous frustule. The Medusettida (Pls. 118-120) show a peculiar alveolar structure, numerous small compartments being enclosed between two parallel plates. In the Circoporida (Pls. 114-117) and Tuscarorida (Pl. 100) the opaque porcellanous shell has a peculiar cement structure (§ 104), and the lattice-structure is confined for the most part to characteristic rings of pores at the base of the hollow tubes, which arise from the shell. The most peculiar lattice-work, however, appears in the segmented shell of the Aulosphærida (Pls. 109-111) and Cannosphærida (Pl. 112). In the former the large meshes of the lattice-work are usually subregular and triangular, in the latter polygonal; the trabeculæ are hollow cylinders, filled with jelly, and containing usually a central axial thread. In each nodal point of the lattice, in which three or more tangential tubes meet, these are separated by stellate or astral septa.
135. Radial Spines of the Skeleton.—The skeleton in the great majority of Radiolaria is armed with radial spines, which are of great importance in the development of their general form and of their vital functions. From a morphological point of view the number, arrangement, and disposition of the spines is usually the determining factor as regards the general form of the skeleton. Physiologically they discharge distinct functions, as organs of protection and support; they act also, like the tentacles of the lower animals, as prehensile organs, since their points, lateral branches, barbed hooks, &c. serve to hold fast nutritive materials. In general main-spines and accessory spines may be distinguished in most Radiolaria; the former are of pre-eminent importance in determining the figure of the skeleton; the latter are merely appendicular organs. The main-spines present such characteristic and important differences in the various legions of Radiolaria that they must be considered separately.
136. Radial Spines of the Spumellaria.—The radial spines, which exhibit most manifold variations in the large order Sphærellaria, present characteristic differences in its four suborders. In the Sphæroidea their number and disposition serve for the separation into families (p. 59); the Cubosphærida (Pls. 21-25) always possess six radial main-spines, which stand opposite to each other in pairs and lie in three diameters of the shell, which are at right angles to each other and correspond to the axes of the regular crystallographic system. The Staurosphærida (Pl. 15) have four spines, which form a regular cross and stand opposite to each other in pairs, in two axes at right angles. The Stylosphærida (Pls. 13-17) show only two main-spines, which are opposed to each other in the vertical main axis of the body. Finally, the Astrosphærida (Pls. 18-20, 26-30) are characterised by a larger and variable number of radial spines (eight, twelve, twenty or more), sometimes regularly, sometimes irregularly arranged. Among the other Sphærellaria the Prunoidea (Pls. 13-17, 39, 40) are most allied to the Stylosphærida with two opposite main-spines; the Discoidea (Pls. 31-47), on the other hand, to the Staurosphærida with four crossed spines; there exist, however, Discoidea with two opposite, three marginal, or numerous radial main-spines; it is characteristic of this suborder that they all usually lie in the horizontal median plane of the lenticular shell, arising from its equatorial margin. The Larcoidea (Pls. 9, 10, 49, 50) show a great variety in the number and arrangement of their radial main-spines, which in the different families of this suborder stand in direct causal relation to the various forms of growth of the shell; usually the primary main-spines lie either in the three different dimensive axes, at right angles to each other, whose differentiation is characteristic of the lentelliptical Larcoid shell (§§ 34, 122) or in definite diagonal axes, which cut the former obliquely. The radial spines of the Spumellaria are never united in the centre of the body, but arise separately from the surface of the primary central lattice-shell (medullary shell), more rarely from one of the secondary (cortical) shells, which enclose it. Their form is originally three-edged (sometimes pyramidal, sometimes prismatic); the cause of this is to be found in their origin from the nodal points of the lattice-shell, whose meshes are primitively hexagonal; hence three trabeculæ unite in each nodal point, and are produced into three edges of the spine. Very commonly, however, the spines are round (conical or cylindrical), more rarely polygonal. The three edges are often delicately toothed, not unfrequently spirally twisted around the axis of the spine (Pl. 21, figs. 1, 12).
137. Radial Spines of the Acantharia.—The radial spines of this legion have a much greater significance than in the other three classes of Radiolaria, since here alone they are the primary determining factors in the skeletal structure, and grow outwards from the middle of the central capsule. This centrogenous origin of the radial spines is as characteristic of the Acantharia as their chemical constitution, which is not siliceous but acanthinic (§ 102). Furthermore, their form is in most cases so peculiar that even an isolated Acantharian spine can be generally distinguished from one belonging to either of the other three legions. In the great majority of the Acantharia (all Acanthonida and Acanthophracta) twenty radial spines are constantly present, which, disposed according to a definite geometrical law, make up the skeleton (compare § 110 above and p. 717). The twenty spines are generally simply apposed to each other in the centre (either by the surfaces or the edges of their pyramidal base); more rarely they are completely united and form a single star-like piece of acanthin (Astrolithium). Very rarely (Acanthochiasma) each two opposite spines are united so that ten diametric bars cross in the middle of the central capsule. Whilst in the great majority of Acantharia these twenty radial spines are present, the small group Actinelida is characterised by the possession of an inconstant, often very large number, sometimes over one hundred. Among these Actinelida are probably to be found the stem-forms of the whole legion. The variously modified spines of the Acantharia may be grouped in three main categories: (1) round (cylindrical or conical); (2) four-edged (prismatic or pyramidal); (3) two-edged (leaf- or sword-shaped). The latter very commonly bear two opposite transverse processes, the former four crossed ones. By ramification and union of these apophyses arise the lattice-shells of the Acanthophracta (excepting the Sphærocapsida).
138. Radial Spines of the Nassellaria.—The radial spines in this legion show as great a variety in their form as in the Spumellaria, and, as in them, are solid, siliceous bars, usually three-edged (prismatic or pyramidal), or round (cylindrical or conical); more seldom they are polygonal in section. The great majority of the Nassellaria are, however, distinguished by a triradial structure, three primary radial bars diverging from the base of the central capsule (usually from the centre of the porochora); there is usually in addition a fourth apical spine, which passes upwards vertically or obliquely on the dorsal aspect of the central capsule. These three or four typical radial spines of the Nassellaria may be derived with great probability from the basal tripod of the Plectoidea (Plagoniscus, Plectaniscus, &c., Pl. 91); and since this tripod is very characteristically combined in Cortina and Cortiniscus with the primary sagittal ring of the Stephoidea, the three typical rays may be generally designated "cortinar feet," in contradistinction to the other radial processes of the Nassellarian skeleton. One of the three descending basal feet ("pes caudalis," Pls. 91-95, C) is always unpaired, and lies in the vertical median plane (or sagittal plane), just as does the vertically directed apical spine, which originally forms the dorsal bar of the sagittal ring, and is produced upwards into the "apical horn," (marked a on the plates). The other two basal feet are paired, and diverge right and left, forwards and downwards ("pedes pectorales," p.p.). Six-rayed Nassellaria, in which three secondary (interradial) feet are intercalated between the three primary (perradial) cortinar feet, are less common than the three-rayed forms. In some groups the number rises still higher, nine, twelve, or even more secondary feet being intercalated between the three primary. Besides, accessory radial spines may be developed on different parts of the shell, which have sometimes a definite relationship to the typical radial spines, sometimes not. Their form and ramification are very various (Pls. 51-98).
139. Radial Spines of the Phæodaria.—The radial spines of the Phæodaria are very clearly distinguished from those of other Radiolaria by the fact that they are usually hollow tubes, rarely solid bars. As a rule, the tubes are cylindrical, often slightly fusiform or conical, their siliceous wall is very thin, and their lumen filled with jelly; a fine thread of silica usually runs in the axis, and in several families is connected by fine transverse threads with the wall of the tube (Pl. 110, figs. 4, 6; Pl. 115, figs. 6, 7). The peculiar family Medusettida is characterised by a very remarkable segmentation of the hollow spines (Pls. 118-120). Each tube is divided by a series of septa into chambers, which communicate by a central or excentric opening in each septum, an arrangement resembling the siphuncle of the chambered Cephalopod shells. The number and arrangement of the radial tubes in most Phæodaria is indefinite and very variable; only in a few families is the number constant in each species and genus, and the disposition regular. The Medusettida (Pls. 118-120) resemble the Nassellaria, inasmuch as equal radial feet diverge from the base of their shell, sometimes three in number (Cortinetta, Pl. 117, fig. 9), sometimes four (Medusetta, Pl. 120, figs. 1-4), sometimes six (Gazelletta); Gorgonetta is specially distinguished by the possession of six ascending and six descending spines regularly alternating (Pl. 119). The Tuscarorida (Pl. 100) usually have three or four equidistant feet. The Circoporida (Pls. 115-117), on the other hand, rather approach the Sphæroidea, their spherical or regular polyhedral shell having a definite number of tubular radial spines, which arise at regular intervals from their angles; Circoporus has six, Circospathis nine, Circogonia twelve, and Circorrhegma twenty radial tubes. Very rarely the tubes of the Phæodaria are angular, usually they are round, more or less cylindrical, though they are often bifurcated or even ramified, and exhibit a great wealth of the most delicate appendages; siliceous hairs, bristles, spines, barbed or anchor-like hooks, spathillæ, brushes, circlets, &c. (compare Pls. 99-128).
140. Main-Spines and Accessory Spines.—As accessory spines (Paracanthæ) we oppose to the main-spines (Protacanthæ), just described, all those processes which have no determining influence upon the formation of the skeleton as a whole, but are to be regarded as secondary constituents of the skeleton, or appendicular organs of inferior significance. They are developed in the utmost variety, sometimes as hairs or bristles, sometimes as thorns or clubs, either straight or curved (often zigzag), smooth or barbed; sometimes standing vertically upon the shell, or directed towards the centre, sometimes obliquely, or rising at a definite angle. In those Spumellaria whose lattice-shell consists of several concentric spheres, the accessory spines generally arise from the outermost, the main-spines, on the contrary, from the innermost. In the Nassellaria, multifarious forms of accessory spines are especially developed in the order Plectellaria. In the Phæodaria they are often furnished with delicate appendages, e.g., anchor-hooks, spathillæ, coronets, &c. Among the Acantharia the accessory spines which arise from the surface of the shell in the Acanthophracta are very characteristic. They are not radially disposed (like the similar superficial spines of the Spumellaria), but parallel to the radial main-spines from whose transverse processes they arise. Since in all these Acanthophracta the twenty radial main-spines are opposite to each other in pairs, all the accessory spines (often several hundred) are parallel to ten different regularly disposed axes of the lattice-shell (Pls. 134-138).
The skeletons of the Radiolaria, in addition to the general relations which have been discussed above, present numerous and important special differences in the various larger and smaller groups. These are indicated in detail in the descriptions of the legions, orders, and families in the systematic portion of this Report.
A SKETCH OF OUR KNOWLEDGE OF THE DEVELOPMENT OF THE RADIOLARIA IN THE YEAR 1884.
141. Individual Developmental Stages.—The germinal history of the Radiolaria presents great obstacles to direct observation, and hence is very incompletely known. The fragmentary observations, however (having been made on Radiolaria of very various groups and supplemented by comparative anatomical considerations), allow us to draw a general picture of the essential developmental processes in this great class. It may probably be assumed that in all Radiolaria, after maturation, the central capsule discharges the function of a sporangium, and its contents are broken up into numerous flagellate swarm-spores (zoospores). After these flagellate swarm-spores (resembling Astasia) have emerged from the ruptured central capsule, they probably pass over into a Heliozoan-stage (Actinophrys) and then after the formation of a jelly-veil into the condition of Sphærastrum. Afterwards, when a membrane is formed between the outer jelly-veil and the inner nucleated cell-body, an Actissa-stage arises, which exhibits in its simplest form the differentiation of the spherical unicellular body into the central capsule and calymma. Actissa thus represents both ontogenetically and phylogenetically the primitive condition of the Radiolarian organism, and may thus be regarded as the point of departure of all other forms.
142. The Astasia-Stage.—The formation of flagellate zoospores in the mature central capsule is probably to be regarded as the common form of individual development in all Radiolaria; since the whole contents are utilised in the formation of these swarm-spores, and since the extracapsulum takes no share in the process and perishes after they are evacuated, the central capsule may be regarded as a sporangium (see note A, below). The zoospores of the Radiolaria generally arise in the following way:—the nucleus of the unicellular organism, sometimes early, sometimes late (and in several different ways, §§ 63-70) breaks up into numerous small nuclei, and each of these surrounds itself with a small portion of the endoplasm. Very often, perhaps generally, this endoplasm contains one or several fat-granules and sometimes also a small oblong crystal; from the protoplasm of the small roundish or ovoid cells protrudes one or more vibratile flagella. The fully developed spores, which commence their vibrations even within the central capsule, emerge when it ruptures, and swim about freely in the surrounding water by means of the flagellum. At this stage of its existence the young Radiolarian represents essentially the simplest form of the Flagellata, such as Astasia or Euglena; the unicellular body is for the most part ovoid or subcylindrical, sometimes fusiform or reniform, usually from 0.004 to 0.008 mm. in diameter (Pl. 1, fig. 1c; Pl. 129, fig. 11). In the anterior part of the flagellate cell, immediately behind the base of the flagellum, lies a homogeneous, spherical nucleus, whilst in the posterior part are usually several small fat-granules and often also a small oblong crystal (hence the name "crystal-spore," "Krystall-Schwärmer"). The number of vibrating flagella, which are extremely long and fine, seems to be variable, usually one, sometimes two, occasionally perhaps three, or even four or more (see note B).
A. The formation of the motile spores in the central capsule was first observed by J. Müller in Acanthometra (1856, L. N. 10, p. 502), then by A. Schneider in Thalassicolla (1858, L. N. 13, p. 41), and finally by myself in Sphærozoum (1859, L. N. 16, p. 141). These older observations were, however, incomplete, for the origin of the motile corpuscles from the contents of the central capsule was not observed. The first complete and detailed observations upon the formation of spores in the Radiolaria were published in 1871 by Cienkowski (L. N. 22, p. 372, Taf. xxix.); they relate to two different Polycyttaria, Collosphæra and Collozoum. These investigations were supplemented by R. Hertwig on Collozoum and Thalassicolla (1876, L. N. 26, pp. 28, 43, &c.); on Collozoum he made the important discovery that the Polycyttaria form two kinds of spores, one with and the other without crystals, and that the latter are divided into macrospores and microspores (compare the chapter on "Reproduction," §§ 212-216). Quite recently Karl Brandt has confirmed these observations, and has extended them to all the genera of Polycyttaria (1881, L. N. 38, p. 393, and 1885, loc. cit.).
B. The number of flagella, projecting from each spore, is very difficult to determine, owing to their extraordinary length and slenderness. It appeared to me that in the majority of those Radiolaria whose spores I investigated only a single flagellum could be demonstrated with certainty, although sometimes two, springing from a common base, seemed to be present. Compare the chapter on "Reproduction," (§ 215) and the recent work of Karl Brandt on Sphærozoea (1885, L. N. 52, pp. 145-174).
143. The Actinophrys-Stage.—The fate of the flagellate zoospores which emerge from the mature central capsule of the Radiolaria has not hitherto been decided by actual observation; all attempts to rear the swarming zoospores have been in vain, for they have soon died. From what we know, however, of the comparative morphology of the Protista, the hypothesis is fully justified, that between the Astasia-stage of the flagellate swarm-spores, and the well-known Actissa-stage of the simplest Radiolaria, there lies an intermediate developmental stage, which may be regarded as being essentially the simplest Heliozoan form, Actinophrys or Heterophrys. The swarm-spore is very probably converted directly in to a simple floating Heliozoon by its elongated or ovoid body becoming spherical and by fine pseudopodia protruding all round instead of a single flagellum; the nucleus at the same time assuming a central position.
144. The Sphærastrum-Stage.—The Actinophrys-stage of the young Radiolaria, which proceeds immediately from the flagellate zoospore, is probably connected with the Actissa-stage by an intermediate form, which may be regarded as a simple skeletonless Heliozoon with a jelly-veil; a well-known example of such a form is Sphærastrum (in the solitary, not the social condition) and Heterophrys. This important intermediate form has arisen from the simple Actinophrys-stage by the excretion of an external structureless jelly-veil, such as is formed in many other Protista (e.g., in the encystation of many Infusoria). The young Radiolarian in this second Heliozoon-stage becomes a simple cell with pseudopodia radiating on all sides; its body consists of three concentric spheres, the central nucleus, the protoplasmic body proper, and the surrounding calymma or jelly-veil. When a firm membrane is developed between the last two spheres this Sphærastrum-stage passes over into the Actissa.
The gap in our empirical knowledge which still exists between the flagellate stage (§ 142) and the simplest Radiolarian stage (Actissa, § 145), can be filled hypothetically only by the assumption of several Heliozoon-stages following one upon another. It is possible also that the capsule-membrane is not formed between the endoplasm and exoplasm (as here supposed), but that the membrane was formed first outside the cell and the extracapsulum subsequently secreted around it.
145. The Actissa-Stage.—The first Spumellarian genus, Actissa, is not only the simplest form actually observed among the Radiolaria, and the true prototype of the whole class, but also the simplest form under which the Radiolarian organisation can be conceived. It is therefore extremely probably that Actissa not only forms the common stem-form of the whole class in a phylogenetic sense, but is also its common ontogenetic or germinal form. Probably in all Radiolaria the Sphærastrum-stage develops immediately into the typical Actissa-stage, by the formation of a firm membrane between the protoplasmic body of the spherical Heliozoan cell and its jelly-veil. Thus arises the characteristic central capsule, which is wanting in the nearly related Heliozoa. It is further probable that all Radiolaria in their early stage will so far conform to the state of things in Actissa as to have the capsule-membrane of the spherical skeletonless cell perforated everywhere by fine pores. This structure is retained in all Spumellaria, whilst in the other three legions those structural relations of the capsule which are characteristic of each develop from the Actissa-stage.
146. The Ontogeny of the Spumellaria.—In the simplest case the individual development in the Spumellaria ceases with the Actissa-stage. In all other genera of this legion diverging forms proceed from this, of which the different growth of the three dimensive axes on the one hand (§§ 44, 45), and the differentiation of the various parts of the unicellular organism with the formation of the skeleton on the other, are of pre-eminent significance. Even in the varying growth of the central capsule in the different dimensions of space in the skeletonless Colloidea, four different modes may be distinguished, which further, in the corresponding development of the skeleton, furnish the basis for the origin of the four orders of Sphærellaria. The most primitive and simplest form of growth, equal extension in all directions, is found in the spherical central capsule and the concentric spherical skeletons (Procyttarium, Sphæroidea). When the growth of the central capsule proceeds more rapidly in the direction of the vertical main axis than in any other direction, the ellipsoidal or cylindrical central capsule (Actiprunum) arises, and the vertically elongated skeleton of the Prunoidea, which is derived from it. When, on the contrary, the growth of the central capsule and lattice-shell is less in the direction of the vertical main axis than in any other direction, the lenticular or discoid central capsule (Actidiscus) arises, and the corresponding lenticular shell of the Discoidea. Finally, even quite early in many Spumellaria, the growth of the central capsule and of the corresponding lattice-shell in the three dimensive axes is different, and hence arise the lentelliptical forms whose geometrical type is the triaxial ellipsoid or the rhombic octahedron (Actilarcus, Larcoidea). Thus the origin of the four orders of Sphærellaria is simply explained by a varying growth in the different dimensive axes. The primary (innermost) lattice-shell is in this legion always simultaneously developed (suddenly excreted at the moment of lorication from the sarcodictyum). The secondary lattice-shells, on the other hand, which surround the former concentrically, and are united with it by radial bars, arise successively from within outwards.
147. The Ontogeny of the Acantharia.—The individual development of the Acantharia in the simplest case (Actinelius) stops at a point which differs from the Actissa-stage only in the change of radial axial threads into acanthin spines. In the small group Actinelida, their number remains variable and usually indeterminate (Adelacantha), whilst in the great majority of the legion (Acanthonida and Acanthophracta) the number is constantly twenty, and those spines are regularly arranged according to the Müllerian law in five parallel circles, each containing four crossed spines (Icosacantha). The simplest form among these latter is Acanthometron, which may be regarded both ontogenetically and phylogenetically as the common starting-point of all the Icosacantha. Within this extensive group variations in the length of the dimensive axes appear, similar to those observed in the Spumellaria. In the Astrolonchida and Sphærophracta the central capsule remains spherical, extending equally in all directions; and correspondingly the lattice-shell, which is excreted on the surface of the spherical calymma, remains spherical. In the Belonaspida (just as in the Prunoidea) this form passes over into an ellipsoid by prolongation of one axis; on the contrary, in the Hexalaspida (as in the Discoidea) the discoidal or lenticular form arises by shortening of an axis. Finally, in the Diploconida, and in some Hexalaspida in which the growth is different in all three dimensive axes (as in the Larcoidea), both the central capsule and the shell assume the lentelliptical form. The lattice-shell of the Acanthophracta is usually successive in its development, since from each of the twenty radial spines two or four tangential apophyses proceed, whose branches subsequently unite and combine to form the lattice-shell. Only in the peculiar Sphærocapsida can the pavement-like shell arise simultaneously or in a moment of lorication.
148. The Ontogeny of the Nassellaria.—The individual development of the Nassellaria in the simplest instance remains stationary at the skeletonless Nasselid stage (Cystidium, Nassella), which can be immediately derived from the foregoing Actissa-stage by the disappearance of the pores in the upper (apical) hemisphere of the central capsule, whilst in the lower (basal) portion they are modified to form a porochora; the podoconus is developed within the endoplasm upon this latter. Usually the spherical form of the central capsule passes over into an ovoid or ellipsoidal one, the vertical axis which passes through the centre of the porochora being elongated. From the skeletonless Nassellida the other Nassellaria may be derived both ontogenetically and phylogenetically by the excretion of an extracapsular siliceous skeleton. Unfortunately, the earliest stages in the formation of this skeleton are unknown, and hence no answer can at present be given to the important question, in what order the three primary skeletal elements of the Nassellaria (the basal tripod, sagittal ring, and latticed cephalis) appear (compare §§ 111 and 182). If, for example, in Cortina and Tripospyris the basal tripod were to appear first in the ontogeny, and the sagittal ring were developed from this, then the Plectoidea would be rightly considered to be the oldest forms in the phylogeny of the skeleton-forming Nassellaria; and in the contrary case the Stephoidea would be so regarded. The relations of growth in the three dimensive axes are very variable in the Nassellaria; the three most important factors in this respect (partly separately and partly in combination) are; (1) the development of the basal tripod to a triradial stauraxon form (the ground-form being a three-sided pyramid); (2) the development of the sagittal ring in the median plane of the body (the vertical axis having the poles different); (3) the development of the latticed cephalis outside the central capsule (the poles of the vertical axis being again different). Since the development both of the skeleton and of the malacoma is characterised in most Nassellaria by the stronger growth of the vertical axis and the differentiation of the two poles, the allopolar monaxon ground-form acquires a predominant significance in this legion (§ 32); the starting point of most of the further modifications is the basal pole of the vertical main axis. Next to this the sagittal axis is usually the most important determining factor (its dorsal and ventral poles being usually different), more rarely the frontal axis (with equal right and left poles). In the zygothalamous Spyroidea (as in the Stephoidea) the formation of the shell proceeds from the sagittal ring, whilst in the polythalamous Cyrtoidea the latticed cephalis is always the starting point, from which a series of joints (thorax, abdomen, and in the Stichocyrtida, the numerous post-abdominal joints) successively arise (unipolar growth).