On Growth and Form

If I have dealt comparatively briefly with the imorganic skeleton of sponges, in spite of the obvious importance of this -part of our subject from the physical or mechanical point of view, it has been owing to several reasons. In the first place, though the general trend of the phenomena is clear, it must be at once admitted that many points are obscure, and could only be discussed at the cost of a long argument. In the second place, the physical theory is (as I have shewn) in manifest conflict with the accounts given by various embryologists of the development of the spicules, and of the current biological theories which their descriptions embody; it is beyond our scope to deal with such descriptions

in detail. Lastly, we find ourselves able to illustrate the same physical principles with greater clearness and greater certitude in another group of animals, namely the Radiolaria. In our description of the skeletons occurring within this group we shall by no means abandon the preliminary classification of microscopic skeletons which we have laid down; but we shall have occasion to blend with it the consideration of certain other more or less correlated phenomena.

The group of microscopic organisms known as the Radiolaria is extraordinarily rich in diverse forms, or “species.” I do not know how many of such species have been described and defined by naturalists, but some thirty years ago the number was said to be over four thousand, arranged in more than seven hundred genera*. Of late years there has been a tendency to reduce the number, it being found that some of the earlier species and even genera are but growth-stages of one and the same form, sometimes mere fragments or “ fission-products”? common to several species, or sometimes forms so similar and so interconnected by intermediate forms that the naturalist denominates them not “species” but “varieties.” It has to be admitted, in short, that the conception of species among the Radiolaria has not hitherto been, and is not yet, on the same footing as that among most other groups of animals. But apart from the extraordinary multiplicity of forms among the Radiolaria, there are certain other features in this multiplicity which arrest our attention. For instance, the distribution of species in space is curious and vague; many species are found all over the world, or at least every here and there, with no evidence of specific limitations of geographical habitat; others occur in the neighbourhood of the two poles; some are confined to warm and others to cold currents of the ocean. In time also their distribution is not less vague: so much so that it has been asserted of them that “from the Cambrian age downwards, the families and even genera appear identical with those now living.” Lastly, except perhaps in the case of a few large “colonial forms,” we seldom if ever find, as is usual

* Haeckel, in his Challenger Monograph, p. clxxxviii (1887) estimated the number of known forms at 4314 species, included in 739 genera. Of these, 3508 species were described for the first time in that work. in most animals, a local predominance of one particular species. On the contrary, in a little pinch of deep-sea mud or of some fossil “Radiolarian earth,’ we shall probably find scores, and it may be even hundreds, of different forms. Moreover, the radiolarian skeletons are of quite extraordinary delicacy and complexity, in spite of their minuteness and the comparative simplicity of the “unicellular” organisms within which they grow; and these complex conformations have a wonderful and unusual.appearance of geometric regularity. All these general considerations seem such as to prepare us for the special need of some physical hypothesis of causation. The little skeletal fabrics remind us of such objects as snow-crystals (themselves almost endless in their diversity), rather than of a collection of distinct animals, constructed in apparent accordance with functional needs, and distributed in accordance with their fitness for particular situations. Nevertheless great efforts have been made of recent years to attach “a biological meaning” to these elaborate structures ; and “to justify the hope that in time the utilitarian character (of the skeleton} will be more completely recognised*.”

In the majority of cases, the skeleton of the Radiolaria is composed, like that of so many sponges, of silica; in one large family, the Acantharia (and perhaps in some others), it is composed, in great part at least, of a very unusual constituent, namely strontium sulphatey. There is no fundamental or important — morphological character in which the shells formed of these two constituents differ from one another; and in no case can the chemical properties of these inorganic materials be said to influence the form of the complex skeleton or shell, save only in this general way that, by their rigidity and toughness, they may give rise to a fabric far more delicate and slender than we find developed among calcareous organisms.

A slight exception to this rule is found in the presence of true crystals, which occur within the central capsules of certain Radio- + Biitschli, Ueber die chemische Natur der Skeletsubstanz der Acantharia, Zool. Anz. ¥Xx, p. 784, 1906. laria, for instance the genus Collosphaera*. Johannes Miiller (whose knowledge and insight never fail to astonish us) remarked that these were identical in form with crystals of celestine, a sulphate of strontium and barium; and Biitschli’s discovery of sulphates of strontium and of barium in kindred forms render it all but certain that they are actually true crystals of celestine f. In its typical form, the Radiolarian body consists of a spherical mass of protoplasm, around which, and separated from it by some sort of porous “capsule,” les a frothy mass, composed of protoplasm honeycombed into a multitude of alveoli or vacuoles, filled with a fluid which can scarcely differ much from sea-water tf. According to their surface-tension conditions, these vacuoles may appear more or less isolated and spherical, or joining together in a “froth” of polygonal cells; and in the latter, which is the commoner condition, the cells tend to be of equal size, and the resulting polygonal meshwork beautifully regular. In many cases, a large number of such simple individual organisms are associated together, forming a floating colony, and it is highly probable that manv other forms, with whose scattered skeletons we are alone acquainted, had in life formed part likewise of a colonial organism. In contradistinction to the sponges, in which the skeleton always begins as a loose mass of isolated spicules, which only in a few exceptional cases (such as Kuplectella and Farrea) fuse into a continuous network, the characteristic feature of the Radiolarians hes in the possession of a continuous skeleton, in the form of a netted mesh or perforated lacework, sometimes however replaced . by and often associated with minute independent spicules.

Before we proceed to treat of the more complex skeletons, we may begin, then, by dealing with these comparatively simple cases where either the entire skeleton or a considerable part of it is represented, not by a continuous fabric, but by a quantity of loose, separate spicules, or aciculae, which seem, like the spicules of Aleyonium, * For figures of these crystals see Brandt, F. u. Fl. d. Golfes von Neapel, X11. Radiolaria, 1885, pl. v. Cf. J. Miller, Ueber die Thalassicollen, etc. Abh. K. Akad. Wiss. Berlin, 1858.

+ Celestine, or celestite, is StSO, with some BaO replacing SrO. { With the colloid chemists, we may adopt (as Rhumbler has done) the terms spumoid or emulsoid to denote an agglomeration of fluid-filled vesicles, restricting the name froth to such vesicles when filled with air or some other gas. to be developed as free and isolated formations or deposits, precipitated in the colloid matrix, with no relation of form to the cellular or vesicular boundaries. These simple acicular spicules occupy a definite position in the organism. Sometimes, as for instance among the fresh-water Heliozoa (e.g. Raphidiophrys), they lie on the outer surface of the organism, and not infrequently (when the spicules are few in number) they tend to collect round the bases of the pseudopodia, or around the large radiating spicules, or axial rays, in the cases where these latter are present. When the spicules are thus localised around some prominent centre, they tend to take up a position of symmetry in regard toit; instead of forming a tangled or felted layer, they come to lie side by side, in a radiating cluster round the focus. In other cases (as for instance in the well-known Radiolarian Aulacantha scolymantha) the felted layer of aciculae lies at some depth below the surface, forming a sphere concentric with the entire spherical organism. In either case, whether the layer of spicules be deep or be superficial, it tends to mark a “surface of discontinuity,” a meeting place between two distinct layers of protoplasm or between the protoplasm and the water around; and it is obvious that, in either case, there are manifestations of surface-energy at the boundary, which cause the spicules to be retained there, and to take up their position in its plane. The case is somewhat, though not directly, analogous to that of a cirrus cloud, which marks the place of a surface ! of discontinuity in a stratified at- ; mosphere.

We have, then, to enquire what ! are the conditions which shall, apart ae from gravity, confine an extraneous \e body to a surface-film; and we may ASG do this very simply, by considering ; { system. In Fig. 218 we have two \ \ fluids in contact with one another iis, 318 (let us call them water and protoplasm), and a body (b) which may be immersed in either, or may be restricted to the boundary between. We have here three possible “interfacial contacts,” * each with its own specific surface-energy, per unit of surface area: namely, that between our particle and the water (let us call it a), that between the particle and the protoplasm (8), and that between water and protoplasm (y). When the body lies in the boundary of the two fluids, let us say half in one and half in the other, the surface-energies concerned are equivalent to (S/2)a + (S/2)8; but we must also remember that, by the presence of the particle, a small portion (equal to its sectional area s) of the original contact-surface between water and protoplasm has been obliterated, and with it a proportionate quantity of energy, equivalent to sy, has been set free. When, on the other hand, the body lies entirely within one or other fluid, the surface-energies of the system (so far as we are concerned) are equivalent to Sa+sy, or SB+ sy, as the case may be. According as a be less or greater than f, the particle will have a tendency to remain immersed in the water or in the protoplasm ; but if (S/2) (a + B) — sy be less than either Sa or Sf, then the condition of minimal potential will be found when the particle lies, as we have said, in the boundary zone, half in one fluid and half in the other; and, if we were to attempt a more general solution of the problem, we should evidently have to deal with possible conditions of equilibrium under which the necessary balance of energies would be attained by the particle rising or sinking in the boundary zone, so as to adjust the relative magnitudes of the surface-areas concerned.

It is obvious that this principle may, in certain cases, help us to explain the position even of a radial spicule, which is just a case where the surface of the solid spicule is distributed between the fluids with a minimal disturbance, or minimal replacement, of the original surface of contact between the one fluid and the other. In like manner we may provide for the case (a common and an important one) where the protoplasm “creeps up” the spicule, covering it with a delicate film. In Acanthocystis we have yet another special case, where the radial spicules plunge only a certain distance into the protoplasm of the cell, being arrested at a boundary-surface between an inner and an outer layer of cytoplasm; here we have only to assume that there is a tension

at this surface, between the two layers of protoplasm, sufficient to balance the tensions which act directly on the spicule*. In various Acanthometridae, besides such typical characters as the radial symmetry, the concentric layers of protoplasm, and the capillary surfaces in which the outer, vacuolated protoplasm is festooned upon the projecting radii, we have another curious feature. On the surface of the protoplasm where it creeps up the sides of the long radial spicules, we find a number of elongated bodies, forming in each case one or several little groups, and lying neatly arranged in parallel bundles. A Russian naturalist, Schewiakoff, whose views have been accepted in the text-books, tells us that these are muscular structures, serving to raise or lower the conical masses of protoplasm about the radial spicules, which latter serve as so many “tent-poles” or masts, on which the protoplasmic membranes are hoisted up; and the little elongated bodies are dignified with various names, such as “myonemes”’ or “myophrises,” in allusion to their supposed muscular naturet. This explanation is by no means convincing. To begin with, we have precisely similar festoons of protoplasm in a multitude of other cases where the “myonemes” are lacking; from their minute size (-006—-012 mm.) and the amount of contraction they are said to be capable of, the myonemes can hardly be very efficient instruments of traction; and further, for them to act (as is alleged) for a specific purpose, namely the “hydrostatic regulation” of the organism giving it power to sink or to swim, would seem to imply a mechanism of action and of coordination which is difficult to conceive in these minute and simple organisms. The fact is (as it seems to me), that the whole method of explanation is unnecessary. Just as the supposed “hauling up” of the protoplasmic festoons is at once explained by capillary phenomena, so also, in all probability, is the position and arrangement of the little elongated bodies.

Whatever the actual nature of these bodies may be, whether they are truly portions of differentiated protoplasm, or whether they are foreign bodies or spicular structures (as bodies occupying a similar position in other cases undoubtedly are), we can explain their situation on the surface of the protoplasm, and their arrangement around the radial spicules, all on the principles of surface-tension *. This last case is not of the simplest; and I do not forget that my explanation of it, which is wholly theoretical, implies a doubt of Schewiakofi’s statements, which are founded on direct personal observation. This I am none too willing to do; but whether it be justly done in this case or not, I hold that it is in principle justifiable to look with great suspicion upon a number of kindred statements where it is obvious that the observer has left out of account the purely physical aspect of the phenomenon, and all the opportunities of simple explanation which the consideration of that aspect might afford.

Whether it be wholly applicable to this particular and complex case or no, our general theorem of the localisation and arrestment of solid particles in a surface-film is of very great biological importance; for on it depends the power displayed by many little naked protoplasmic organisms of covering themselves with an “agglutinated” shell. Sometimes, as in Difflugia, Astrorhiza (Fig. 219) and others, this covering consists of sand-grains picked up from the surrounding medium, and sometimes, on the other hand, as in Quadrula, it consists of solid particles which are said to arise, as inorganic deposits or concretions, within the protoplasm itself, and which find their way outwards to a position of equilibrium in the surface-layer; and in both cases, the mutual capillary attractions between the particles, confined to the boundary-layer but enjoying a certain measure of freedom therein, tends to the _ orderly arrangement of the particles one with another, and even to the appearance of a regular “pattern” as the result of this arrangement.

The “picking up” by the protoplasmic organism of a solid particle with which “to build its house” (for it is hard to avoid this customary use of anthropomorphic figures of speech, misleading though they be), isa physical phenomenon kindred to that by which an Amoeba “swallows” a particle of food. This latter process has been reproduced or imitated in various pretty experimental * The manner in which the minute spicules of Raphidophrys arrange themselves round the bases of the pseudopodial rays is a similar phenomenon.

Fig, 219. Arenaceous Foraminifera; Astrorhiza limicola and arenaria. (From Brady’s Challenger Monograph.) ways. For instance, Rhumbler has shewn that if a thread of glass be covered with shellac and brought near a drop of chloroform suspended in water, the drop takes in the spicule, robs it of its shellac covering, and then passes it out again*. It is all a question of relative surface-energies, leading to different degrees of “adhesion”? between the chloroform and the glass or its covering. Thus it is that the Amoeba takes in the diatom, dissolves off its proteid covering, and casts out the shell. Furthermore, as the whole phenomenon depends on a distribution of surface-energy, the amount of which is specific to certain particular substances in contact with one another, we have no difficulty in understanding the selective action, which is very often a conspicuous feature in the phenomenon. Just as some caddisworms make their houses of twigs, and others of shells and again others of stones, so some Rhizopods construct their agglutinated “test” out of stray sponge-spicules, or frustules of diatoms, or again of tiny mud particles or of larger grains of sand. In all these cases, we have apparently to deal with differences in specific

* Rhumbler, Physikalische Analyse von Lebenserscheinungen der Zelle, Arch, f. Entw. Mech. vu, p. 103, 1898. + The whole phenomenon is described by biologists as a “surprising exhibition of constructive and selective activity,” and is ascribed, in varying phraseology, to intelligence, skill, purpose, psychical activity, or “microscopic mentality”: that is to say, to Galen’s rexvixh pious, or “ artistic creativeness ” (cf. Brock’s Galen, 1916, p- xxix). Cf. Carpenter, Mental Physiology, 1874, p. 41; Norman, Architectural achievements of Little Masons, etc., Ann. Mag. Nat. Hist. (5), 1, p. 284, 1878; Heron- Allen, Contributions...to the Study of the Foraminifera, Phil. Trans. (B), Ccvt, pp. 227-279, 1915; Theory and Phenomena of Purpose and Intelligence exhibited by the Protozoa, as illustrated by selection and behaviour in the Foraminifera, Journ. h. Microscop. Soc. pp. 547-557, 1915; ibid., pp. 137-140, 1916. Prof. J. A. Thomson (New Statesman, Oct. 23, 1915) describes a certain little foraminifer, whose protoplasmic body is overlaid by a crust of sponge-spicules, as ‘‘a psycho-physical individuality whose experiments in self-expression include a masterly treatment of sponge-spicules, and illustrate that organic skill which came before the dawn of Art.” Sir Ray Lankester finds it “not difficult to conceive of the existence of a mechanism in the protoplasm of the Protozoa which selects and rejects building-material, and determines the shapes of the structures built, comparable to that mechanism which is assumed to exist in the nervous system of insects and other animals which ‘automatically’ go through wonderfully elaborate series of complicated actions.” And he agrees with “Darwin and others [who] have attributed the building up of these inherited mechanisms to the age-long action of Natural Selection, and the survival of those individuals possessing qualities or ‘tricks’ of life-saving value,” J. R. Microsc. Soc. April, 1916, p. 136.

surface-energies, and also doubtless with differences in the total available amount of surface-energy in relation to gravity or other extraneous forces. In my early student days, Wyville Thomson used to tell us that certain deep-sea “ Difflugias,”’ after constructing a shell out of particles of the black volcanic sand common in parts of the North Atlantic, finished it off with “a clean white collar” of little grains of quartz. Hven this phenomenon may be accounted for on surface-tension principles, if we assume that the surfaceenergy ratios have tended to change, either with the growth of the protoplasm or by reason of external variation of temperature or the like; and we are by no means obliged to attribute the phenomenon to a manifestation of volition, or taste, or aesthetic skill, on the part of the microscopic organism. Nor, when certain Radiolaria tend more than others to attract into their own substance diatoms and such-like foreign bodies, is it scientifically correct to speak, as some text-books do, of species ‘‘in which diatom selection has become a regular habit.” To do so is an exaggerated misuse of anthropomorphic phraseology.

The formation of an “agglutinated” shell is thus seen to be a purely physical phenomenon, and indeed a special case of a more general physical phenomenon which has many other important consequences in biology. For the shell to assume the solid and permanent character which it acquires, for instance, in Difflugia, we have only to make the additional assumption that some small quantities of a cementing substance are secreted by the animal, and that this substance flows or creeps by capillary attraction between all the interstices of the little quartz grains, and ends by binding them all firmly together. Rhumbler* has shewn us how these agglutinated tests, of spicules or of sandgrains, can be precisely imitated, and how they are formed with greater or less ease, and greater or less rapidity, according to the nature of the materials employed, that is to say, according to the specific surface-tensions which are involved. For instance if we mix up a little powdered glass with chloroform, and set a drop of the mixture in water, the glass particles gather neatly round the surface of the drop so quickly that the eye cannot follow the

operation. Ifwe perform the same experiment with oil and fine sand, dropped into 70 per cent. alcohol, a still more beautiful artificial Rhizopod shell is formed, but it takes some three hours to do. It is curious that, just at the very time when Rhumbler was thus demonstrating the purely physical nature of the Difflugian shell, Verworn was studying the same and kindred organisms from the older standpoint of an incipient psychology*. But, as Rhumbler himself admits, Verworn was very careful not to overestimate the apparent signs of volition, or selective choice, in the little organism’s use of the material of its dwelling.

This long parenthesis has led us away, for the time being, from the subject of the Radiolarian skeleton, and to that subject we must now return. Leaving aside, then, the loose and scattered spicules, which we have sufficiently discussed, the more perfect Radiolarian skeletons consist of a continuous and regular structure ; and the siliceous (or other inorganic) material of which this framework is composed tends to be deposited in one or other of two ways or in both combined: (1) in the form of long spicular axes, usually conjoined at, or emanating from, the centre of the proto- ‘plasmic body, and forming a symmetric radial system; (2) in the form of a crust, developed in various ways, either on the outer surface of the organism or in relation to the various internal surfaces which separate its concentric layers or its component vesicles. Not unfrequently, this superficial skeleton comes to constitute a spherical shell, or a system of concentric or otherwise associated spheres.

We have already learned that a great part of the body of the Radiolarian, and especially that outer portion to which Haeckel has given the name of the “calymma,” is built up of a great mass of “vesicles,” forming a sort of stiff froth, and equivalent in the physical sense (though not necessarily in the biological sense) to “cells,” inasmuch as the little vesicles have their own well-defined boundaries, and their own surface phenomena. In short, all that we have said of cell-surfaces, and cell conformations, in our discussion of cells and of tissues, will apply in like manner, and under appropriate conditions, to these. In certain cases, even in

so common and simple a one as the vacuolated substance of an Actinosphaerium, we may see a very close resemblance, or formal analogy, to an ordinary cellular or “ parenchymatous”’ tissue, in the close-packed arrangement and consequent configuration of these vesicles, and even at times in a slight membranous hardening of their walls. Leidy has figured * some curious little bodies, like small masses of consolidated froth, which seem to be nothing else than the dead and empty husks, or filmy skeletons, of Actinosphaerium. And Carnoy7{ has demonstrated in certain cell-nuclei an all but precisely similar framework, of extreme delicacy and minuteness, as the result of partial solidification of interstitial matter in a closepacked system of alveoli (Fig. Fig. 220. ‘Reticulum plasmatique.” Let us now suppose that, (After Carnoy.) : : : : in our Radiolarian, the outer surface of the animal is covered by a layer of froth-like vesicles, uniform or nearly so in size. We know that their tensions will tend to conform them into a “honeycomb,” or regular meshwork of hexagons, and that the free end of each hexagonal prism will be a little spherical cap. Suppose now that it be at the outer surface of the protoplasm (that namely which is in contact with the surrounding sea-water), that the siliceous particles have a tendency to be secreted or adsorbed; it will at once follow that they will show a tendency to aggregate in the grooves which separate the vesicles, and the result will be the development of a most delicate sphere composed of tiny rods arranged in a regular hexagonal network (e.g. Aulonia). Such a conformation is

extremely common, and among its many variants may be found cases in which (e.g. Actonomma), the vesicles have been less regular in size, and some in which the hexagonal meshwork has been developed not only on one outer surface, but at successive surfaces, producing a system of concentric spheres. If the siliceous material be not limited to the linear junctions of the cells, but spread over a portion of the outer spherical surfaces or caps, then we shall have the condition represented in Fig. 223 (Hthmosphaera), where the shell appears perforated by circular instead of hexagonal apertures, and the circular pores are set on slight spheroidal eminences; and, interconnected with such types as this, we have others in which the accumulating pellicles of skeletal matter have extended from the edges into the substance of the boundary walls

Fig. 223. Hthmosphaera conosiphonia, Fig. 224. Portions of shells Akl. of two “species” of Cenosphaera : upper and have so produced a system of films, normal to the surface of the sphere, constituting a very perfect honeycomb, as in Cenosphaera favosa and vesparia*. : In one or two very simple forms, such as the fresh-water Clathrulina, just such a spherical perforated shell is produced out of some organic, acanthin-like substance; and in some examples of Clathrulina the chitinous lattice-work of the shell is just as

* Tn all these latter cases we recognise a relation to, or extension of, the principle of Plateau’s bourrelet, or van der Mensbrugghe’s masse annulaire, of which we have already spoken (p. 297). regular and delicate, with the meshes just as beautifully hexagonal, as in the siliceous shells of the oceanic Radiolaria. This is only another proof (if proof be needed) that the peculiar conformation of these little skeletons is not due to the material of which they are composed, but to the moulding of that material upon an underlying vesicular structure.

Let us next suppose that, upon some such lattice-work as has just been described, another and external layer of cells or vesicles is developed, and that instead of (or perhaps only in addition to) a second hexagonal lattice-work, which might develop concentrically to the first in the boundary-furrows of this new layer of cells, the siliceous matter now tends to be deposited radially, or normally to the surface of the sphere, just in the lines where the external layer of vesicles meet one another, three by three. The result will be that, when the vesicles themselves are removed, a series of radiating spicules will be revealed, directed outwards from each of the angles of the original hexagon; as is seen in Fig. 225. And it may further happen that these radiating skeletal rods are continued at their distal ends into divergent rays, forming a triple fork, and corresponding (after a fashion

which we have already described as occurring in certain spongespicules) to the three superficial furrows between the adjacent cells. This last is, as it were, an intermediate stage between the simple rods and the complete formation of another concentric sphere of latticed hexagons. Another possible case is when the large and uniform vesicles of the outer protoplasm are mixed with, or replaced by, much smaller vesicles, piled on one another in more or less concentric layers; in this case the radiating rods

will no longer be straight, but will be bent into a zig-zag pattern, with angles in three vertical planes, corresponding to the successive contacts of the groups of cells around the axis (Fig. 226). Among a certain group called the Nassellaria, we find geometrical forms of peculiar simplicity and beauty,—such for instance as that which I have represented in Fig. 227. It is obvious at a glance that this is such a skeleton as may have been formed

(1 think we may go so far as to say must have been formed) at the interfaces of a little tetrahedral group of cells, the four equal cells o the tetrahedron being in this particular case supplemented by a little one in the centre of the system. We see, precisely as in the internal boundary-system of an artificial group of four soap-bubbles, the plane surfaces of contact,- six in number; the relation to one another of each triple set of interfacial planes, meeting one another at equal angles of 120°; and finally the relation of the four lines or edges of triple contact, which tend (but for the little central vesicle) to meet at co-equal solid angles in the centre of the system, all as we have described on p. 318. In short, each triple-walled re-entrant angle of the little shell has essentially the configuration (or a part thereof) of what we have called a “Maraldi pyramid” in our account of the architecture of the honeycomb, on p. 329*.

There are still two or three remarkable or peculiar features in this all but mathematically perfect shell, and they are in part easy and in part they seem more difficult of interpretation. We notice that the amount of solid matter deposited in the plane interfacial boundaries is greatly increased at the outer margin of each boundary wall, where it merges or coincides with the superficial furrow which separates the free, spherical surfaces of the bubbles from one another; and we may sometimes find that, along these edges, the skeleton remains complete and strong, while it shows signs of imperfect development or of breaking away over great part of the rest of the interfacial surfaces. In this there is nothing anomalous, for we have already recognised that it is at the edges or margins of the interfacial partition-walls that the manifestation of surface-energy will tend to reach its maximum. And just as we have seen that, in certain of our “multicellular” spherical Radiolarians, it is at the superficial