planes; communicate with the interseptal canals; and open on the surface of the cord respectively.

In some, if not in all, specimens, the spicules not only exist throughout the cord (for they can be seen on its inner aspect where the cord is in contact with the outer margin of the chambers), but are continued down over the interseptal spaces even to the innermost

turn.

Now, if the substance of the cord were homogeneous, the structures mentioned in it could not be defined. If it were simply the "peculiar manner" in which its homogeneous substance were "traversed by the set of canals" which it contains, as stated by Dr. Carpenter, then portions of the cord on transverse fracture could not be made to present the ends of spicules at the fractured points, nor could portions of the cord fall out on fracture longitudinally, in the form of spicules. Nothing but certain portions of the cord being harder than the rest, and of these portions being of a spicular form, could give rise to either of these appearances. While, if it be the "peculiar manner in which the homogeneous substance of which it is composed is traversed by the set of canals," which gives the cord an appearance of spicular structure, how is it that this spicular appearance exists over almost all the interseptal spaces of some specimens, where there is no plexus and no canals but those which pass through it almost perpendicularly? It is, however, useless to have recourse to argument for conviction when the fact can be demonstrated, so we will turn our attention to another point in the economy of this shell, viz. the "canal-system."

As regards the use of the canal-system, nothing yet has been definitely assigned. I thought formerly that it subserved the purpose of a water-circulation as in Sponges, viz. the water going in by the ends of the small canals which open on the horizontal surface of the test, and coming out through the orifices of the larger ones on the surface of the spicular cord, and I now think that this may be a part of their function, at the same time that they may draw in nutritious particles by the small pores also like the Sponges. The anastomosing canals, like also in appearance and function to the mycelium of Fungi, serve to convey portions of the sarcode (upon which the canals themselves are first moulded) to the points from which new portions of the organism are to be developed; while they undoubtedly too, in part, perform the office of excretory channels, for in the recent and living specimens of Operculina Arabica (which I obtained on the coast of Arabia), the sarcode of the interseptal canals remains, after the calcareous matter

of the test has been removed by acid, in connection with the membranous chambers, by short branches, through which globular bodies (to which I shall more particularly allude presently), which are more or less present in the chambers, readily pass, on pressure, into the larger interseptal canals; and again through these, probably by the openings on the spicular cord, during the living state, they would have obtained an exit. That the sarcode of the canal-system also carries on the development of the organism independently of the chambers, is proved by the development of the test continuing after the chambers have ceased to be formed,—as will be hereafter mentioned. Lastly, the substance covering the horizontal surface of the test, which I have likened to the cuticle of shells, in accounting for the formation of the horizontally laminated structure of the test, and have inferred to be connected with the sarcode of the chambers through the vertical tubuli of the test, MM. d'Archiac and Haime have more properly likened to the "épiderme des échinides ou de l'épithèque des polypes" (p. 69). But what I meant is seen by the context in my "Paper" viz. that, in its dry state, it was merely like the cuticle of shells in appearance and not identical with it.

Of this substance Dr. Carpenter states nothing in his "Paper" on Operculina; but in his description of Orbitolites he observes, in a "foot-note" (p. 207), “I have little doubt that the greenish' cuticle described by Mr. Carter as covering his Operculina Arabica, and supposed by MM. d'Archiac and Haime to be specially connected with the formation of the shell, is of the same nature," that is, of the nature of "a covering of vegetation, chiefly composed of Diatomaceæ, Dismidiex, and other minute Algæ." To which I must simply reply, that, "such a mistake is impossible with a practical microscopist." Moreover, lately I have had to examine some "deep-sea soundings " from the Arabian Sea, in which there were many minute foraminifera, and as it is at such depths that the foraminifera are most likely to be taken up alive or with the living sarcode of the animal in their tests, so most of these were covered with the so-called "cuticle " first seen on the substance of Operculina Arabica.

But are not the horizontal, or "spiral" laminae (as they have been termed by D'Archiac and Haime in Nummulites,) of Operculina composed almost entirely of vertical tubuli which establish a direct connection between the cavity of the chambers and the surface, and between the chambers of the overlying layers in Nummulites? Is not the sarcode which fills the cavities of the test of Operculina, &c., like

that of the Rhizopoda generally, viz. whose portions flow together when they come in contact; and would they thus not form a layer over the surface of the test? And, could the horizontal layers of the test be formed in any other way, or are they likely to be so, under such circumstances? Lastly, is not all this in favour of what I have stated, viz. that there is a substance, in appearance, like the "cuticle" of shells, over the dried specimens of foraminifera which contain the living organism when they are taken out of the water? But as I have already observed respecting the spicular structure of the cord, the fact does not rest upon argument but can be demonstrated, and upon demonstration I made the statement ten years ago!

The "vertical tubuli," as just stated, connect the chambers with the surface, not only in Operculina, but in the tests of Nummulites, Orbitoides dispansa, and Orbitolites Mantelli, and it is through their agency chiefly that the layers of shell and the chambers are vertically formed.

The openings on the horizontal surface over and about the septal spaces are those of canals connected with the great interseptal system. They are the same as MM. d'Archiac and Haime's "canaux d'une troisième grandeur," or middle-size-canals.

But besides these openings, there are spaces and lines in Operculina which are composed of shell-substance alone, that is, without the presence of the vertical tubuli, or the middle-size-canals, and these, in the test of recent Operculina, have the appearance of homogeneity and transparency, but are opaque and white in the fossilized one, where they evidently are identical with the opaque white portions of Nummulites which have afforded MM. d'Archiac and Haime some of their chief distinguishing characters; and thus, the latter are proved not to be what they supposed them, viz. remains of "larges canaux," but originally, transparent portions of the shell, unaccompanied by any canal, except accidentally, as will be more particularly shown hereafter,

Animal of Operculina.-Hardly anything more of the animal of Operculina is known now than when my description of the test was published; and I now as then, cannot help thinking that the existence of the animal matter of the Robulina which I examined at sea, and thought to be in the form of a worm in "loops" in the chambers, united by contrictions where the chambers joined close to the spicular cord, was a fallacy; for the observations were made at sea, in a little vessel, on the deck, in the open air, with simple, though powerful, lenses, and, since then all that I have been able to obtain from the specimens

of Operculina which had living animals in them when they were taken, is a number of membranous sacs, corresponding in form with that of the chambers, and united by a like membranous structure at the base, where the latter are in contact with the spire or spicular cord; together with the membranous portions of the interseptal canals; those of the marginal plexus; those which unite the chambers together through the interseptal spaces; and lastly, others which unite the chambers to the interseptal canals.

Besides this, the chambers have generally been more or less filled with minute, spherical, nucleated bodies, whose size I have mislaid, similar to those which I have described, and figured in the fresh-water Rhizopod, called "Euglypha," and which I have considered to be embryonal cells (Ann. vol. xviii. pl. v. fig. 26, and vol. xx. pl. 1, fig. 196, &c.), which again are like those figured by Dr. Carpenter in the cells of Orbitolites (Phil. Trans. pl. iv. fig. 3), and which he views as the sarcode broken up into propagative "gemmules." My own view of them in Operculina also, is, that they are propagative bodies of the species; but the most interesting point which their presence in the chambers of Operculina has elicited with me is that, by slight pressure they can be easily made to pass through the short channels of communication which exist between the chambers and the interseptal canals, into the latter, showing thus, as before stated, that one of the uses of the interseptal canals is to give exit to these bodies through the branches of the marginal plexus which open on the surface of the last or outermost turn of the spicular cord.

The bond of union between the chambers at the spicular cord is not tubular, or at least, I never saw the propagative spherules pass from one chamber to the other through it; while I am inclined to think that this bond is composed of part of the sarcodal plexus of the spicular cord, and that from this the chambers are developed, as will be better understood presently; also that the part of the plexus which is more directly concerned in uniting the chambers occupies the free surface of the spicular cord, and gives rise to that arched opening which exists between the septa and the spicular cord. The free margin of the septum here also frequently presents a scolloped form, as if it had rested on a plexus of sarcodal filaments, while in some cases I have not been able to distinguish any aperture between the septum and the cord at all, indicating that the two are sometimes in contact. All this seems to show that the chambers are developed from the marginal plexus and not from each other.

As regards the function of the chambers, the presence of the nucleated spherules above mentioned in them seems to indicate that they are the reproductive organs; and the fact that in Alveolina elliptica they are frequently almost wanting altogether, while in other instances they are interrupted two or three times by several turns of the cortical part alone (which part is analogous to the spicular cord of Operculina); and that, in the globose nummulites the chambers are frequently not distinguishable in the outer turns, shows that the development of the test can go on without the presence of the chambers, and therefore that they are probably supplementary and propagative. Indeed, the approximation of the turns of the spire or those of the spicular cord, which, of course, must entail a corresponding diminution in the size of the chambers, will be found by-and-by to indicate the full size of the species, in which the subsidence of the generative force appears to be thus indicated.

As yet, however, we know very little about the animal of the foraminifera, chiefly because we are so ignorant of the allied forms to it. The same is the case with the animal of Spongilla, which I have described (Ann. vol. xx. 1857). It fails to elicit much attention, because at present it has no known alliances, but by -and-by, when these are found out, and more forms of the same kind are discovered for comparison, then the nature and position in organic development of these beings will be realised, and their component parts understood. Till then they must remain in abeyance.

Mode of growth.-The mode of growth in Operculina and Nummulites is the same. That is, the horizontal portions, or spiral lamina as they have been termed by MM. d'Archiac and Haime, are developed from the sarcode of the chambers passing through the vertical tubuli; while the spicular cord and the chambers in the first instance spring from the marginal plexus of sarcodal filaments. The latter is shown in Alveolina as above stated, where the chambers cease to be developed and then appear again after several turns of the spire have been completed by the cortical layer alone, which part, as before stated, is analogous to the spicular cord. Both the segments of Operculina and those of Nummulites begin to be formed from the spicular cord, and three or more of the last are generally in successive stages of development, the last of all being the least formed. This must not, however, be confounded with the last chambers of the fully developed Operculina, which, like those of Nummulites, are also successively less in size.

More recent observations.-Since the above was written I have again