THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA PRESENTED BY PROF. CHARLES A. KOFOID AND MRS. PRUDENCE W. KOFOID PHILOSOPHICAL TRANSACTIONS. I. Researches on the Foraminifera. Ey WILLIAM B. CARPENTER, M.D., F.E.S., F.G.S., &c. Eeceived June 17, Eead June 17, 1858. PAET III. ON THE GENEEA PENEEOPLIS, OPEEOULINA, AND AMPHISTEGINA. 119. IN my preceding memoirs, I have shown that two very dissimilar types of structure present themselves among Foraminifera ; one characterized by its simplicity, the other by its complexity. In the former, of which OrMtolites, Orbicutina, and Alveolina are typical examples, the calcareous skeleton does not present any definite indications of organization, but seems to have been formed by the simple calcification of a portion of the homogeneous sarcode-body of the animal ; that sarcode-body is but very imperfectly divided into segments, the communications between the cavities occupied by these segments being very free and irregular ; the form of the segments themselves, and the mode of their connexion, are alike inconstant ; and even the plan of growth, on which the character of the organism as a whole depends, though preserving a general uniformity, is by no means invariably maintained. In the latter, to which Cycloclypeus and Hete- rostegina belong, we find the calcareous skeleton presenting a very definite and elaborate organization ; the several segments of the body are so completely separated from each other, that they remain connected only by delicate threads of sarcode ; each segment thus isolated has its own proper calcareous envelope, which seems to be moulded (as it were) upon it, and this envelope or shell is perforated with minute parallel tubuli, closely resembling those of dentine except in the absence of bifurcation or ramifica- tion ; the partition-walls between adjacent segments are consequently double, and are strengthened by an intermediate calcareous deposit, which is traversed by a system of inosculating passages that seems properly to belong to it. The form of the segments, their mode of communication, and consequently the general plan of growth, have a very considerable degree of constancy ; and altogether the tendency is strongly manifested in MDCCCLIX. B 2 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. this type, to the greater individualization of the parts of the composite body, which in the preceding must be looked upon rather as constituting one aggregate whole. 120. I purpose in the present memoir to carry on this contrast, by presenting a detailed comparison of the structure of two generic forms, which, whilst they so far agree in general plan of growth as not only to have been ranked by M. D'ORBIGNY side by side in his order Helicostegues, but to have been placed by other systematists in close apposition, differ in the most marked manner as to all the particulars just enumerated. Both these types are of peculiar interest : the first, Peneroplis, on account of the very wide range of variation it presents, which has led to the establishment of three genera, apparently distinguishable by well-marked differences in conformation, upon what I feel satisfied will prove to be but individual modifications of one and the same specific type ; the second, Operculina, as being the nearest existing representative of Nummulites, and consequently as affording not merely the key to the elucidation of the structure, but also the basis for the determination of the value of the reputed species, of that genus, by the study of the range of variation which it presents ; this range being, though more restricted than in the preceding case, still quite sufficient to justify a large multiplica- tion of species, in the estimation of those who do not practise that extended method of comparative inquiry, on the importance of which I have dwelt in a former Memoir (^f 74). With the latter of these genera, as also with Nummulites, I shall prove that the genus Amphistegina is closely allied ; although M. D'ORBIGNY, misled by the marked want of symmetry and by the alternation in the disposition of the chambers, which are exhibited by certain forms o f that type, has placed it in a different order, Entomostegues. For I shall have to show that a gradational variety in this respect, ending in complete symmetry, may coincide with such a uniformity in general structure, that even a very decided departure from symmetry must be regarded as a character of little value in classification, compared with agreement in the organization of the shell and in those peculiarities in the conformation of the animal which are indicated by it ; and further, that a most marked difference in degree of organization exists between two species of Amphistegina, which so closely resemble each other externally that the young of one may easily be mistaken for the adult of the other. Genus PENEROPLIS. 121. History. The genus Peneroplis was first instituted by MONTFORT* to distinguish a peculiar type of minute polythalamous shells, which had been previously described and figured by FICHTEL and MoLLf, and had been ranged by them with numerous others under the comprehensive designation Nautilus; and MONTFORT correctly indicates its distinctive character, as " bouche de toute la longueur de la base, et percee serialement par une file des pores," though he seems to have very erroneously interpreted the signi- * Burrow de SONNINI, 1802-1805 ; Mollusques, tome iv. p. 1. pi. 42 ; and Concbyliologie Systeinatique, 1808, p. 258. f Testacea Microscopica, Vindob. 1798-1803, p. 91. tab. 16. figs, a, i. GENUS PENEKOPLIS: HISTOET AND DISTRIBUTION. 3 fication of those pores. For he goes on to state, " Cette coquille est encore pellucide, et permet de lire au travers de son tet la serie et la disposition des nombreuses cellules de chaque concameration ; ces cellules deviennent plus grandes a mesure que la coquille prend plus d'accroissement. II est probable que leur nombre repond a celui des animaux qui les habitent, et qui les constituent simultanement pour former un nouveau rang." This sub- division of the principal chambers into cells, each occupied by a separate animal, exists only in the imagination of MONTFORT, who seems to have been misled by the peculiar markings of the surface of the shell, which, as will be presently seen, are merely superficial. LAMARCK, not adopting MONTFORT'S genus, referred the Nautilus planatus of FICHTEL and MOLL to the genus Cristellaria, with which it has no relationship whatever ; and having copied their figures into the ' Encyclopedic Methodique *,' he designated one set of forms as C. planata, and another as C. dilatata, subsequently reuniting them, however, under a new specific name, C. sguammula^. The genus Peneroplis has been recognized by BLAINTILLE J and by EHRENBERG ; and D'ORBIGNY has applied this name, in his various writings on Foraminifera, to the form described by FICHTEL and MOLL, whilst he has created a new generic term, Dendritina, for a' series of closely allied forms, in which there is a single large dendritic aperture instead of a linear series of separate apertures. The distinctive characters of these two genera, as last given by him||, are as follows: " Peneroplis ; coquille nautiloide comprimee, pourvue de nombreuses ouvertures sur une seule ligne a la demiere loge seulement. Cavite simple : Dendritina ; ce sont des Peneroplis dont les ouvertures anastomosees ferment une dendrite." In addition to these, I shall cite his definition of another genus, established by LAMARCK in 1801, which, as I shall presently show, consists, like Dendritina, of mere varieties of Pene- roplis: " Spirolina ; coquille nautiloide dans le jeune age, projetee en crosse dans 1'age adulte." 122. The genus Peneroplis is very widely diffused through warmer latitudes; indeed few collections of Foraminifera from sands or dredgings taken from the Mediterranean, the ^Egean Archipelago, the Red Sea, the East or West Indies, the Philippine seas, or the shores of Australia or the Polynesian Islands, will fail to present numerous examples of it. A few specimens have been found on British coasts ; but it is surmised by Pro- fessor WILLIAMSON that these have been brought by the Gulf-stream from the West Indian Seas. " Amongst the multitude of West Indian seeds and other light objects thus thrown upon our north-western shores, it was to be expected that some of the tropical Foraminifera would be entangled ; and such may have been the case with the species under consideration^." The finest examples I have met with were contained in * Tab. 467, figs. 1, a, b, c, 2 ; a, b, c. f Animaux sans Vertebres, torn. vii. p. 607. J Malacologie, p. 372. " Tabellarische Charakteristik der Bryozoen-Classe und sammtlicher Familien und Gattungen der Polythalamien," in Berlin Transactions, 1838. || Cours lilementaire de Paleontologie, 1849, torn. ii. p. 198. If On the Eecent Foraminifera of Great Britain, published by the Bay Society, 1858, p. 46. u2 4 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEKA. Mr. CUMING'S Philippine Collection ; whilst the most numerous were kindly furnished me by Mr. J. GWYN JEFFERYS from his dredgings in the Gulf of Genoa. We shall hereafter find that, as in the case of Orbitolites, particular varieties of conformation prevail, though by no means exclusively, in particular localities. 123. External Characters. The ordinary form of the shell of Peneroplis (of which an ideal representation is given in Plate I. fig. 2) is an extremely flat spire of about two turns and a half, opening out rapidly in its last half turn (Plate II. fig. 1). In the young shell, each whorl is merely adherent to the preceding, so that the first-formed portion is not concealed by the subsequent growth ; but it very commonly happens that the last whorl spreads itself out to such a degree as partially to invest the preceding (Plate II. figs. 1, 3), the extension towards the centre, however, seldom reaching so far as to conceal the original umbilicus. Although this lateral extension is sometimes con- fined to the inner margin of the spire, as is seen in fig. 1, yet it often occurs along the outer margin also, as is seen in fig. 3 ; being usually limited, however, to the last four or five chambers. But sometimes, instead of opening out and partially investing the previous whorls, the spire is prolonged in a straight line, and several of its successive chambers present little or no progressive increase in size (Plate II. fig. 5) ; though even in this variety the last four or five chambers are often seen to spread themselves out rather suddenly, so as to extend along the inner margin of the straight portion. Between these different extremes of conformation, every intermediate gradation presents itself. 124. The surface of the shell is very strongly marked by depressed bands, which indicate the place of the septa between the chambers ; and between these septal bands the walls of the chambers rise in flattened arches. In a direction transverse to the septal bands, we almost uniformly observe a strongly-marked striation ; the striae running parallel at tolerably regular intervals, which average about TTtfoth f an inch, from one septal band to another. This striation, which imparts a very characteristic physiognomy to these minute shells, seems due to a sort of plication or ridge-and-furrow arrangement of the shelly wall (Plate II. fig. 20), which may not improbably have the effect of imparting to it increased strength. The plication generally disappears at the junction of the walls of the chambers with the septa ; and consequently we do not usually see it at the mouth when the shell is viewed endways. Sometimes, however, it is continued on to the septum itself; and its character is then extremely well displayed, as in Plate II. fig. 15, which, however, represents not the typical form, but one of the varieties to be hereafter noticed. On the prominences of the plicae, there are frequently to be seen rows of extremely minute puncta (fig. 20) ; these, however, are not the apertures of pas- sages through the shell, as might not unnaturally be supposed ; but are, like the punc- tations of Orbiculina (^[ 88), mere depressions of its surface, as I have ascertained by the careful examination of very thin sections. It is remarkable that the plication of the shell is sometimes wanting, though the punctations may still present themselves in rows, as shown in fig. 23 ; whilst in other cases, not only are the plicae deficient, but the punctations are distributed uniformly over the entire surface, as shown in fig. 24. That GENUS PENEKOPLIS : OEGANIZATION. these variations are not indicative of any specific difference, is at once proved by the fact that the shells which exhibit them in one part, present the ordinary character of surface in another. In Plate II. fig. 22 is shown a case in which similar punctations present themselves on the septum closing the mouth of the shell of a Dendritina, which, as will presently appear, is only a modified Peneroplis. 125. The septum which closes the mouth of the shell is perforated by numerous isolated pores, arranged in a single linear series (Plate I. fig. 2, Plate II. fig. 1); the number of these pores depends upon the length of the septal plane (a very convenient term which I adopt from Professor WILLIAMSON), and thus it usually increases with the age of the individual, each chamber opening externally by a larger number of pores than did that which preceded it. The typical form of these pores seems to be circular, though they are apt to present various departures from that shape ; they usually lie in a sort of furrow, formed by the projection of the lateral borders of the mouth somewhat beyond the septum ; and, as in Orbitolites and Orbiculina, each one is surrounded by a prominent annulus of shell. 126. The texture of the shell very closely resembles that of Orbitolites and Orbiculina, but is somewhat more porcellanous. As in those genera, the shell presents an opaque white hue, when it is viewed by light reflected from its surface ; whilst thin sections examined by transmitted light are of a brownish yellow or dark amber colour. Its sub- stance is apparently quite homogeneous, no other trace of structure presenting itself than the plications and punctations already referred to ; and its texture is not nearly so firm as that of those Foraminiferous shells which possess a minutely-tubular organi- zation. 127. On examining a thin section of a typical Peneroplis, taken through the median plane between the lateral surfaces (Fig. I.), the central chamber is seen to have the Fig. I. Section of Peneroplis, parallel to its surface. globose form which has been shown to characterize the primordial segment of the Fora- minifera already described ; from this first chamber a single passage leads to the second, 6 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMIXIFEEA. which communicates in like manner by a single passage with the third, as does the third with the fourth ; the fourth chamber, however, communicates by two passages with the fifth, as does the fifth with the sixth, and the sixth with the seventh. In the septum between the seventh and eighth, with which the second whorl may be considered as commencing, there are three apertures ; and this number continues for the four consecu- tive partitions which divide the chambers forming the next half-convolution. Then, however, commences a very remarkable increase ; for whilst in each of the next two partitions there are four passages, the numbers in the four succeeding partitions which divide the chambers completing the second turn are respectively 6, 9, 11, and 14 ; whilst in the last eight partitions which divide the chambers of the outer half-whorl, the numbers of the apertures are respectively 14, 20, 26, 28, 30, 35, 44, and 48. The average distance of the apertures from each other remains nearly the same throughout ; so that their number pretty closely corresponds with, and may be taken to represent, the length of the septal plane which they traverse in each case; and it is not a little remarkable, that whilst this number should only increase from 1 to 4 in the first con- volution and a half, it should so rapidly augment from 4 to 48 in the last half-con- volution. 128. As I have not been fortunate enough to obtain any other than dried specimens of this organism, I have not had the opportunity of examining the structure and arrange- ment of its soft parts. Such an opportunity, however, has presented itself to Professor EHEENBERG, who collected living specimens in his expedition to the Red Sea; and by treating these with dilute acid, he has freed the animal body from its enclosing shell, precisely as I was able to do in the case of Orbitolites. The figure * which he has given of the animal corresponds in every important particular with what an examination of the shell would lead me to expect ; for it represents a series of segments of a generally- homogeneous substance, corresponding in form, dimensions, and connexions with the successive chambers of Fig. I. The only departure that I can discover from what I should myself have anticipated, lies in this that the successive segments are connected along the inner margin of the convolutions by a band much broader and thicker than the threads which pass between other parts of the segments ; so that this band would seem to establish a principal connexion, to which the other threads might be considered as secondary. Now after a very careful examination both of the septal planes of numerous specimens, and of sections taken in the direction of Fig. I, I feel myself justified in the positive assertion that no such principal aperture exists at the inner margin of each septum, as would be required to give passage to such a band as is figured by Professor EHRENBEKG. Consequently I can only account for this feature in his delineation of the animal (the idea of a difference in the conformation of the shell being negatived by the precise correspondence between his figures of it and my own, as well as by my familiarity with the Red Sea type of Peneroplis), by supposing that, like some of his other figures, it rather represents his idea of the structure of the animal, than what he actually saw in *v Abhandl. der Komgl. Akad. der Wiasenschaft. zu Berlin, 1838. Physik, taf. 2, fig. 1. GENUS PENEEOPLIS: OKGANIZATION AND AFFINITIES. 7 its body, this principal band being apparently regarded by him as an intestinal canal, by which he supposed all the segments to be connected together. Professor EHRENBERG'S figure represents a marking-out of spherules in the midst of the otherwise homogeneous substance of the body, closely corresponding to those which I have described in the animal of Orbitolites (^[ 43); and it also shows the skeletons of various Diatomacese imbedded in the substance of the four outer segments, as I have myself found them in Orbitolites (^[ 34). Consequently I can have no hesitation in the belief that the general structure and mode of life of these two forms are analogous ; and that there does not exist in Peneroplis, any more than in Orbitolites, that hypothetical organization which Professor EHRENBERG has attributed to both. 129. Monstrosity. The only example of monstrosity which I kave met with in this type, is obviously analogous to the " monstrosities by excess " which I have described as not unfrequent in Orbitolites; consisting (at least apparently) in the formation of a secondary body by outgrowth or gemmation from the first. This secondary body does not seem, however, to have its origin in the primordial segment, to which similar out- growths may always be traced in Orbitolites (^[ 62). 130. Affinities. As regards its general organization and plan of growth, it seems clear that Peneroplis bears a very near approximation to that form of Orbiculina in which the later increase takes place like the earlier upon the spiral type (^[ 87), and in which there is but a single plane of chambers with a single row of marginal pores (^[ 90): the only essential difference, in fact, consisting in this, that each successive increment is formed in the case of Peneroplis by a continuous segment of sarcode, occupying an undi- vided chamber of the shell ; whilst in the case of Orbiculina each segment is subdivided into sub-segments strung like the beads of a necklace on a continuous stolon of sarcode, the chamber being correspondingly subdivided by transverse partitions having a con- tinuous passage through all. The very close alliance of the two forms is shown by the fact, that specimens are occasionally to be met with in collections of Orbiculince, which, while presenting the general physiognomy of that genus, exhibit the deficiency of trans- verse partitions which is the distinguishing character of Peneroplis ; so as to render it uncertain whether such specimens should be considered as Orbiculince in which (as certainly happens occasionally in Orbitolites, ^[61) the usual partitions are deficient, or whether they truly belong to Peneroplis, their resemblance to Orbiculina being super- ficial only, and their presence among specimens of that genus merely accidental. In either case it is obvious that the affinity between these two types is very close ; and in fact young Orbiculince frequently present such a strong external resemblance to Pene- roplis, that their true nature cannot be determined without such an examination of their internal structure as will serve to disclose the transverse partitioning of their chambers. It may be fairly inquired whether the peculiar striation of the surface of Peneroplis, which so strongly suggests the idea of internal partitioning as to have led MONTFORT to assert its existence, is not really a rudiment of that structure. 131. Varieties. The relationship just indicated being borne in mind, it will scarcely 8 DE. CAEPENTEE'S EESEAECHES OX THE FOEAMINIFEEA. excite surprise to find that Peneroplis, like Orbiculina, presents very considerable diver- sities of conformation. The first and simplest departure from what has been described as its typical character, consists in a duplication of the series of pores in each septum (Plate II. fig. 7 a), the spire being at the same time less compressed, so that the septal plane is wider in proportion to its length. Now it is not a little remarkable that this is almost uniformly the case with specimens furnished by particular localities, whilst those obtained from others not very remote exhibit almost as uniformly the extremest elonga- tion and narrowing of the septal plane with only a single row of apertures ; and hence it might not unreasonably be maintained that this difference should be accounted of specific value. In reply to this, however, there is not only the analogy of Orbitolites and Orbiculina, in which an indefinite multiplication in the rows of marginal pores may take place during the growth of the individual, but also the fact that in Peneroplis the two forms cannot be distinguished at an early age, either by the shape of the shell or by the disposition of the pores, which are often arranged neither in a single nor in a double row, but on a sort of mixture of both plans, as shown in Plate II. figs. 9 and 10 ; whilst among the more advanced examples of each type, it is not at all uncommon to meet with individuals which present a combination of the characters of both, the septal plane having a single row of pores in one part of its length with a double row in another. Sometimes, moreover, in one of the less compressed forms of the shell, although there is but a single row of pores, it is obvious from the elongated shape of these that they indicate a tendency to duplication (fig. 8). Moreover we find in the variety with a com- plete double row the same disposition as in the ordinary Peneroplis to the substitution of the rectilineal for the spiral mode of growth, as we see in fig. 7. Hence I consider that it may be unhesitatingly asserted that the duplication of the row of pores, and the increased turgidity of the spire which it accompanies, are but features of individual variation, and cannot be admitted to rank as specific differences. And in this view I am glad to find myself borne out by Professor WILLIAMSON, who defines Peneroplis*, not (like M. D'ORBIGNT) as having only a single row of apertures, but as having " septal orifices scattered over the long narrow septal plane ; " whilst his figure of Peneroplis planatus shows these orifices arranged for the most part in a double row, one portion having them even more multiplied. 132. What is the relationship to the typical Peneroplis, however, of that group of forms to which D'ORBIGNY has given the generic designation Dendritina, is a question of more difficulty. These are characterized, as we have seen, by the possession of a single large aperture sending out dendritic ramifications in each septum (Plate I. fig. 1, Plate II. figs. 12, 13) ; but this is by no means the whole of their differentiation. For the spire, instead of being compressed, is very turgid ; and its successive whorls not merely surround but also invest those which have preceded them ; so that what may be appropriately termed alar prolongations (al, figs. 12 a, 12 b, Plate II.) of the chambers of even the last whorl often extend nearly to the umbilicus. The geographical distribution * The Eecent Foraminifera of Great Britain, p. 44. RELATIONSHIP OF DENDEITINA TO PEXEEOPLIS. 9 of Dendritina, moreover, is peculiar ; for, so far as I am aware, this type is restricted to the tropical ocean. I have not met with it in dredgings from any part of the Mediter- ranean or the Red Sea, where Peneroplis abounds ; while the largest specimens I have seen are those contained in Mr. CUMING'S Philippine explorations, some of these measuring 078 inch in diameter, and - 030 in thickness or breadth. To such as content themselves with glancing at strongly-marked examples of this type, the propriety of its generic, or at any rate of its specific separation from Peneroplis, would seem indubitable. Never- theless I think that I shall be able to show adequate grounds for the belief, that the two forms cannot be separated by any definite line of demarcation ; and that they must there- fore be ranked as belonging not merely to the same genus, but even to the same species. 133. In the first place I would refer to the fact that the peculiar plication and puncta- tion of the surface of the shell, which are such marked features in the physiognomy of Peneroplis, are repeated in Dendritina (Plate II. figs. 21, 22) in a manner so precisely similar, as strongly to impress every one who has his attention directed to the aspects of these two forms respectively, with the idea of their very close relationship. 134. Secondly, the differences of general configuration between Peneroplis and Den- dritina are differences of degree, and present themselves in very variable amount in different individuals. For, starting from those forms of Peneroplis in which the spire is least compressed, the transition is easy to those Dendritina whose spire is least turgid, and whose septal plane is not broader in proportion to its length than it often is in Peneroplis. From the most compressed forms of Dendritina to those which have the most turgid spires and the widest septal planes, the gradation is insensible, scarcely any two individuals according in their proportions ; thus we find that whilst the septal plane is sagittate in some (Fig. II. A, c), it tends to become reniform in others (B, D), the Fig. II. Front views of four specimens of Dendritina. margin of the spire, which is almost carinated in the first case, becoming obtuse and even cresentic hi the second. Again, the extent of the investment of the earlier whorls MDCCCXLIX. c 10 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. by the later, varies as much as the degree of turgidity ; for whilst in some of the most compressed forms, as in Peneroplis, each whorl does little more than apply itself to the margin of the preceding (Fig. III. A, c), the more turgid the spire becomes, the more completely (generally speaking) does it embrace the preceding, the alar prolongations of the chambers thus coming to bear a large proportion to their principal cavity (Fig. III. B, D). 135. Thirdly, we observe in Dendritina precisely the same tendency to rectilineal extension in the later period of growth, as in Peneroplis ; for although a distinct generic term Spirolina has been given to the form presenting this modification, I think it must be apparent to every one that the example delineated in Plate II. fig. 11, bears just the same relation to the typical Dendritina, that those shown in figs. 5, 7 bear to the typical Peneroplis. The transition from Dendritina to Spirolina is well seen in figs. 12, 13; in the first of which we see the last whorl apparently about to disengage itself from the earlier ones, whilst in the second that disengagement has been completed. I cannot conceive that any one could refuse to regard either of these specimens as a Dendritina ; and yet it is obvious that their continued increase upon the plan which has already manifested itself in the formation of the later chambers, would convert them into Spiro- lincB. Such continued increase has obviously taken place in the specimen represented in fig. 11, which, up to the time of the substitution of the rectilineal for the spiral mode of growth, has all the appearance of an ordinary Dendritina. 136. Still it may be said that, notwithstanding all these points of resemblance, the difference between Peneroplis and Dendritina is clearly marked out by the difference in the modes of communication between the chambers of the two types ; and that such a difference is sufficiently important to constitute a valid generic character. I freely admit that this would be the case, if the difference were constantly to present itself between all the individuals of the same type, as it does between their characteristic examples (Plate I. figs. 1, 2); but the fact is far otherwise. We have seen that among those which would be unhesitatingly ranked under the designation Peneroplis, there is not only a tendency to multiplication of the rows of separate pores, but also an occasional fusion of two or more pores, so as to form a single large pore of irregular shape. On the other hand, among the unquestioned Dendritince, we observe not merely that the form of the single large dendritic aperture is extremely variable, but that it is frequently so simple as to suggest the idea of having been formed by the coalescence of a linear series of pores. The most characteristic forms of the dendritic aperture that I have met with, are shown in Plate II. figs. 12 a, 13 a; two examples of a remarkable departure from this have just been seen in Fig. II. B, D, where the proportions of the aperture are altogether reversed, its breadth being much greater than its length, and its central part being enlarged at the expense of its ramifications ; while through such an aperture as that represented in Plate II. fig. 22, we are conducted back to the ordinary type, though its form is much simplified. On the other hand, in Fig. II. A, c, we have marked examples of a narrowing and elongation of the aperture, with such a reduction KELATION OF DENDEITINA TO PENEKOPLIS. 11 of its dendritic ramifications, that it comes to present little more than a linear fissure. From a comparison of these cases, it will be seen that the form of the aperture bears a pretty constant relation to that of the septal plane ; the broadest apertures presenting themselves in the individuals which have the most turgid spire, and the narrowest in those whose spire is most compressed ; whilst the proportionate development of the two principal alar prolongations seems related to the degree of that alar extension of the chambers over the whorl they enclose, of which I have already spoken. But the most satisfactory proof of the wide extent of range of variation in the form of the aperture in Dendritina, is afforded by a comparative examination of the apertures connecting different chambers of the same individual. Thus in the interior of the very shell that presented the peculiarly characteristic example, Plate II. fig. 12 a, we find a form, b, closely corresponding to that shown in Fig. II. c ; and in four septa of the inner part of another shell, we have the simple forms of aperture represented in Fig. III. A, B, c, D. Fig. III. Septal planes and apertures from different parts of the same specimen of Dendritina. 137. But further, not only do we thus meet with examples of each type which pre- sent more or less of approximation towards the other, but we also not unfrequently encounter individuals in which the characters of the two types are so blended that it is difficult, if not impossible, to say to which they should be referred. Thus in Plate II. fig. 8, we have a young specimen with a linear series of apertures, the marked elongation of which shows each to be formed by the fusion of two, whilst there is an additional pair precisely hi the situation of the alar prolongations in fig. 12 b ; and it is obvious that a longitudinal coalescence of these pores would produce an aperture exactly resembling that of fig. 12 b. Again, in another young specimen shown in Plate II. fig. 16, a partial fusion of the separate pores into a single dendritic aperture has actually taken place (fig. 16 ), whilst another broad aperture is seen just below this. In fig. 10 we see numerous pores, some small and rounded, others large and irregular, each of the latter being obviously formed by the coalescence of two or three of the separate pores ; and it is evident that a closer approximation of the whole would produce a single large den- dritic aperture. Other varieties of the same kind are presented in figs. 9 and 14. But a yet more remarkable example is shown in fig. 15 ; for here the coalescence has pro- c2 12 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. ceeded so far as to produce a number of separate branching apertures, and nothing is wanting but the removal of the line of shell which passes down the middle of the septum, to unite these into the most characteristic form of the single ramifying orifice. This individual, like the one represented in fig. 6, was already beginning to assume the Spirolina form, the rounded shape of the mouth showing that the spire has detached itself completely from the previously-formed convolutions: in figs. 4 and 11 a are shown the mouths of more advanced examples of the same type, which present such a com- bination of the large dendritic aperture of Dendritina with the isolated pores of Pene- roplis, as to complete (in my opinion) the proof that no valid distinction can be drawn between these two types, either from the number, the isolation, the position, or the shape of the apertures in the septa. This conclusion will probably become less sur- prising than it may at first sight appear to such as are accustomed to the study of the more constant characters presented by the shells of Mollusca, when it is borne in mind that the apertures seem to have no other function than to give passage to threads of sarcode ; and that it is the general character of the Ehizopod type, for such extensions of the substance of the body to be entirely destitute of constancy either as to position or number. 138. If the foregoing conclusion be admitted, it follows that not merely must the genera Dendritina and Spirolina be relinquished, but that both these forms must be regarded as mere varieties of Peneroplis planatus. From the circumstance that the Den- dritina-tyTpe is found only in tropical seas, and that it attains a far larger size than the ordinary Peneroplis-type, whilst the latter seems to be almost starved out (as it were) in the Mediterranean, presenting itself there under its humblest form, and scarcely extending itself at all into northern seas, I am disposed to think that temperature has a considerable influence in producing these varieties. I should be far, however, from attributing the entire result to this agency alone, since we find Peneroplis presenting itself abundantly in tropical seas, along with Dendritina. But as it is in the Mediter- ranean forms that we meet with the extreme of flattening and with the most uniform singleness of the row of pores, whilst it is among the tropical forms that we find Pene- roplis not merely attaining its greatest dimensions, but also presenting the closest approximation to Dendritina in the turgidity of its spire and in the arrangement of its septal apertures, it seems obvious that the development of this type of structure is favoured by a constantly high temperature ; and further, that Dendritina may be regarded as the highest form, towards which Peneroplis tends in proportion as it is subjected to that influence. As I have not had the opportunity of examining the fossil forms of this type, I am as yet unable to say how far the foregoing conclusions will be borne out by the phenomena which they present. Genus OPEECULINA. 139. The collection of Mr. CUMING contains a large number of Nautiloid Forami- nifera, differing considerably from one another not only in size and proportions, but also GENUS OPEECTJLINA : HISTOET. 13 in general aspect and particular features, yet having so many structural characters in common as to be obviously referrible to the same generic type. This type I shall designate, provisionally at least, as Operculina ; since, although it does not correspond with the definition of this genus as given by its founder M. D'ORBIGNY, it is obviously related most closely to the fossil forms which have been described under that designa- tion by M. D'ORBIGNY himself, as well as by MM. D'ARCHIAC and HAIME, and to the recent type described as Operculina by Mr. CARTER. 140. History. The genus Operculina was created by M. D'ORBIGNY in 1825, for the reception of a group of Foraminifera bearing a strong general resemblance to Nummulites, but considered by him as differing from it in the extreme compression of its discoidal shell, in the smaller number of its whorls, in the non-investment of the earlier whorls by the later, so that the former remain apparent on both sides through the whole of life, and also in the form of the aperture, which is a transverse slit in Nummulites, but triangular in Operculina. The following is the definition of this genus given by him in his ' Foraminiferes Fossiles de Vienne ' (p. 117) : " Coquille libre, equilaterale, ovale ou discoidale, tres comprimee, formee d'une spire non embrassante, reguliere, egalement apparent de chaque cote, a tours contigus, et croisant tres rapidement. Loges nombreuses, etroites, la derniere formant saillie de toute la largeur de la spire, a tous les ages percee d'une ouverture visible triangulaire, centre le retour de la spire." In his ' Cours Ele- mentaire de Paleontologie ' (torn. ii. p. 197), he defines the genus by the following characters alone : " Coquille comprimee a tours apparents, pourvue d'une ouverture triangulaire centre le retour de la spire." I think it necessary also to quote from the same work M. D'ORBIGNY'S definitions of three other genera nearly allied to Operculina ; since, as will presently appear, the characters which he assigns to them are based on an entirely wrong idea of the distinctive features presented by their respective types. Nonionina. " Coquille nautiloide, pourvue d'une seule ouverture en fente transversale, non masquee, placee contre le retour de la spire." Nummulites. " Coquille lenticulaire a tours embrassants, pourvue d'une ouverture en fente transversale, souvent masquee, contre le retour de la spire." Assilina. " Ce sont des. Nummulites dont les tours de la spire, ne rejoignant pas le centre, sont tous plus ou moins apparents." In the admi- rable Monograph of MM. D'ARCHIAC and HAIME *, the genus Nummulites is defined as follows : " Coquille libre, symmetrique, orbiculaire, plane, discoide, lenticulaire ou subglobuleuse, composee d'une lame calcaire pliee et enroulee suivant un meme plan, criblee de pores ou canaux de trois grandeurs differentes, a tours plus ou moins serres, plus ou moins embrassants, quelquefois simplement juxtaposes. Dans le jeune age, le dernier est saillant ; plus tard, les tours se rapprochent de telle sorte qu'a 1'etat adulte le dernier parait se souder tout a fait a 1'avant dernier. Canal spiral divise par des cloisons transverses, plus ou moins nombreuses, percees a leur base, contre le retour de la spire, d'une ouverture lineaire transverse. Ces derniers caracteres semblent s'etre attenues vers la fin de la vie de Tamma!" (p. 70). The genus Nonionina is said by these * Description des Animaux Fossiles du Groupe Nummulitique de PInde. Paris, 1853. 14 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. authors to differ from Nummulites in respect of its flatter form, the comparative short- ness of its spire, the smaller number of chambers, and the non-embrace of its convolu- tions ; as also in absence of that narrowing of the later convolutions which leads to the entire closure of the spire in Nummulites. The genus Operculina is said (loc. cit.) to be distinguished from Nummulites only by the depression of its form, the small number of its convolutions, and the rapid increase in breadth of the last whorl, which seems constantly to remain open. The genus Assilina, having no sufficiently distinctive characters, must (according to these able investigators) be merged in Nummulites. It is obvious, from these remarks, that MM. D'ARCHIAC and HAIME had fully recognized the two fundamental errors of M. D'ORBIGNY'S definition of Operculina; viz. his state- ment that the convolutions are non-embracing, and his description of the form of the aperture as triangular. And although Mr. CARTER * does not allude to the discrepancy between the structure of the recent type described by him and the generic definition of M. D'ORBIGNY, yet it is evident that in assigning to it the designation Operculina he was guided rather by its general resemblance to the fossils on which that generic name had been conferred, than by the conformity of its structure to M. D'ORBIGNY'S definition. The errors of that definition are easily accounted for ; since it is only by such an exami- nation of thin transverse sections as M. D'ORBIGNY (I have reason to believe) never made, that the fact of the earlier whorls being really embraced by the later, notwithstanding the apparent freedom of the latter, can be substantiated ; and the determination of the form of the aperture is very likely to be erroneously made, when the examination is limited to fossil specimens, in which it is frequently but very indistinctly traceable. 141. The minute structure of this type has. been already investigated by two excellent observers. Under the designation of " an undescribed species of Nonionina" Professor WILLIAMSON f has given an account of the structure of a shell abounding in the Manilla sand, which, not merely from his figures and descriptions, but from a comparison of the specimens which he has kindly enabled me to make, I know to be one of the smaller forms of the Philippine Operculina. Mr. H. J. CARTER J has still more minutely described the organization of an Operculina which he obtained in great abundance on the south- east coast of Arabia, the shells coming up attached to the grease of the sounding lead from sandy bottoms of between ten and twenty fathoms depth ; and of the identity of this with a larger form of Philippine Operculina I am enabled to speak, from comparison of specimens which Mr. CARTER has obligingly transmitted to me. I find the descrip- tion of each to be in the main correct so far as it goes, but to be defective in some essential particulars. Thus Professor WILLIAMSON does not notice the canal-system except in the marginal cord ; and Mr. CARTER, though he has more fully described the canal-system, has not only missed one important part of it, but has misapprehended the structure of the marginal cord. The investigations of each, moreover, have been limited * Annals and Magazine of Natural History, Sept. 1852. t Transactions of Microscopical Society, First Series, vol. iii. p. 112. J Annals and Magazine of Natural History, Sept. 1852. GENTS OPEECULINA : EXTEENAL CHAEACTERS. 15 to one particular form of this protean type ; and neither seems in the least degree aware of its extraordinary tendency to variation. The rich store of material placed at my disposal by Mr. CUMING has not only enabled me to prosecute my inquiries into the minute structure of this organism upon specimens of unparalleled size and degree of deve- lopment, but has also enabled me to bring together a great body of data for comparison as to the extent of variation which it may undergo, alike in external conformation and in internal organization. And as the information hence obtained has a most important bearing upon the study of the closely-allied genus Nummulites, I venture to think that, notwithstanding all which has been done by Professor WILLIAMSON and Mr. CARTER, it is desirable that I should give a detailed account of my own investigations into the structure of Operculina, as well as a summary of the results of my comparison of its multiform varieties. It will be advantageous in the first place to examine into what appears to be the characteristic structure of the type, and then to inquire into the degree of modification to which this may be subject on each point. 142. External characters. A large proportion of Mr. CUMLNG'S specimens accord in their general aspect with the one represented in Plate III. fig. 7, which is closely con- formable to Mr. CARTER'S type of Operculina Arabica, except in its somewhat larger dimensions. It is a compressed spiral of about '25 inch in diameter, and about '015 inch in thickness, consisting of between three and four convolutions gradually increasing in breadth ; these are in general nearly flat, but are sometimes a little arched between their inner and outer margins, sometimes depressed so as to present a slight concavity, espe- cially near the outer margin of the last whorl. The chambers are about seventy-five in number, commencing from a primordial spheroidal cell and progressively increasing in dimensions with the widening of the spire ; their septa have for the most part a radial direction, but they bend backwards near the outer margin of each whorl ; and they are marked externally by bands which are distinguished by their semitransparent aspect from the dull brownish hue of the general surface. These septal bands are commonly on the same plane with the intervening portions of the shell ; but sometimes the walls of the chambers are a little arched between the septa that bound them, so that the septal bands are slightly furrowed ; whilst the walls of the chambers are sometimes a little depressed, so that the septal bands are prominent ; and such varieties may present themselves in different parts of one and the same shell. Not unfrequently the whole surface is seen to be marked by veiy minute punctations ; but more commonly they are larger and fewer in number, and are often arranged in pretty regular lines parallel to the septal bands, one, two, or three rows of such punctations being seen on the wall of each chamber (Plate V. fig. 7) ; these punctations, when sufficiently magnified, are found to be spots of semitransparent shell-substance resembling that of the septal bands ; and, as in the case of these, their surface is sometimes on the same plane with that of the general surface of the shell, sometimes a little elevated so as to form papillae, and sometimes a little depressed into minute fossae (fig. 8). These punctations often present themselves abundantly on some parts of the surface, whilst they are entirely absent from others. Sometimes, instead of a limited number of comparatively large and 16 DB, CAEPENTEE'S EESEAECHES ON THE FOEAMINTFEEA. regularly arranged punctations, we find a vast number of very minute papillae, scattered without order over the entire surface of the chambers (fig. 6) ; these, again, may be absent from some parts of a shell, over other portions of which they are abundantly distributed. There is commonly a large semitransparent tubercle at the umbilicus, and smaller tubercles are often seen along the septal bands, especially of the earlier whorls : the umbilical and the septal tubercles are well seen in Plate III. fig. 6, and (less charac- teristically) in several other figures of the same Plate; and the septal tubercles are shown on a much larger scale in Plate V. fig. 11. Not unfrequently, however, the umbilicus is depressed, instead of being elevated into a tubercle ; and the moniliform tubercles along the septal bands are wanting. 143. The departures from this typical form, however, are very wide. A glance at any considerable aggregation of specimens reveals to us an extraordinary variety of size and shape ; and our attention is specially attracted by a series of which an example is represented in Plate III. fig. 9, which are not merely distinguished by a size greatly above the average their long diameter reaching nearly -4 inch, but also by the extraordinary flattening of the later convolutions, and the rapidity with which the spire opens out. The approximation between the two lateral walls of the chambers is here so close, that not only are the septal bands rendered very prominent by the depression of the outer surface between them, but even the outer marginal band stands up as a ridge, from which the walls of the chambers slope down. In fact, an exami- nation of this form leaves the observer impressed with surprise that any room can be left for the animal, the segments of which must be extraordinarily attenuated, losing in thickness what they gain in area. Now a careful comparison of this form with the ordinary type, not only makes it obvious that the former differs from the latter in no other particular than this attenuation, which (as already pointed out in the cases of Heterostegina and Peneroplis) is a common feature of the later growths in Foraminifera, but also that the attenuation takes place in such different degrees in different individuals, that any attempt to use it as a differential character is completely baffled by the con- tinuous gradation of forms that is presented, between the one which has been assumed as the typical, and such as most widely depart from it in this particular. 144. The collection of Mr. CUMING also includes, however, a considerable proportion of comparatively small specimens, ranging from -08 to -20 inch in diameter, which pre- sent a very different configuration. In the smallest of these, represented in Plate III. fig. 1, the spire, instead of being flat or even somewhat hollowed, is arched from its inner to its outer margin, so that the breadth of the septal plane is equal to its length, or nearly so ; and the like is seen in the somewhat larger specimens of which fig. 2 is an example. This variety of conformation, however, being limited to shells whose earlier period of growth is evinced by their smaller number of convolutions and of chambers, and being seen, moreover, in the earlier whorls of those which afterwards present the greatest flattening, may safely, I think, be regarded as a character of age. It will be recollected that the young of the flattened Peneroplis more resembles Dendritina in the comparative turgidity of its spire ; but I do not find that Operculina ever continues long GEXUS OPERCULINA: EXTERNAL CHAEACTERS. 17 to increase upon such a plan ; for the compression of the spire always shows itself in the third whorl, if it has not previously done so, acd is accompanied with a corresponding augmentation of its breadth, so that the septal plane becomes narrowed and elongated, as is seen in comparing fig. 4 with fig. 3. That no basis for specific distinction is afforded by the most marked differences in the form of the spire, as shown in the pro- portions of the septal plane (A, B, c, D), will be made evident by the examination of such a transverse section as is shown in Fig. V. (p. 21) ; in which it will be observed that differences equal to those presented by the most compressed and the most turgid forms of the spire, are exhibited by the successive convolutions of one and the same individual. 145. Another marked feature of difference among the specimens of this collection, is the depression or elevation of the central region relatively to the peripheral. In what I have assumed as the typical form, the central region presents the same general level with the rest, though the umbilicus itself is often marked by a prominent tubercle. In such forms as are represented in Plate III. figs. 1 and 2, the umbilical region is rather depressed than elevated ; and this depression is often observed in older specimens whose early growth has taken place on this type. But there is a group of specimens, of which three sticcessive ages are represented in Plate III. figs. 3, 4, 5, 8, that have the whole central region so exceedingly prominent as to form a cone, whose apex is marked either by one large tubercle, or by a cluster of smaller ones ; and this conformation gives so peculiar a physiognomy to the shells which present it, that few systematists would hesi- tate in placing them apart as specifically different from the rest. On a careful compari- son of a large number of individuals, however, it becomes apparent that this difference, like the preceding, is gradational ; every degree of prominence being traceable from the individuals which have the umbilicus marked only by a tubercle, as in fig. 7, through those in which the region generally is slightly elevated, as in fig. 8, to those in which it presents the most marked projection, as in figs. 3-5. We shall presently find (*f[ 153) that this difference depends mainly on the degree in which the investing layer, prolonged from the later convolutions over the surface of the earlier, is separated from that surface by the extension of the alar prolongations of the chambers of the investing whorls ; as to which point there is a most remarkable diversity, not only among different individuals, but between the several convolutions of the same individual. 146. A fourth very obvious character of differentiation among the individuals of this collection, consists in the presence or absence of tubercles on the septal bands. In what I have described as the typical form (Plate III. fig. 7), there are no considerable pro- minences over the greater part of the surface ; the septal bands are generally smooth and continuous ; and it is only in the central region that we observe any departure from this uniformity, the umbilicus being occupied by a small tubercle, and the smooth septal bands being replaced in the first whorl by moniliform rows of little tubercles. In other instances, however, we find not only the central tubercle, but the rows of tubercles marking the septa, much larger and more prominent ; and this marking-out of the septal bands by elevated tubercles is not limited to the first whorl, but extends to the second, and even to the third, as is well shown in fig. 6. The specimens whose central region is JfDCCCLIX. D 18 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. very prominent (Plate III. figs. 3, 4, 5), usually show this feature in the most decided manner ; but it is occasionally presented also in no less a degree by those whose umbi- licus is flat or even depressed. The most remarkable departure from the ordinary type is seen in a small group of specimens in which the tubercles are not only extremely large and prominent (Plate IV. fig. 6), but are distinguished by their opaque whiteness (Plate V. fig. 12), which contrasts strongly with the semi-transparency ordinarily charac- terizing these prominences. But even this character has no diagnostic value ; for a speci- men which has the tubercles opaque in one part may have them semi-transparent in another ; and as to the size of the tubercles, their degree of prominence, and the pro- portion of the entire spire over the septa of which they occur, there is every degree of variety. In one of my largest and flattest specimens, the central region is strongly tuberculated, but the tubercles almost suddenly cease when the spire begins to open out, and the septal bands are thenceforth as smooth as in the ordinary type. 147. A less obvious but still very decided feature of individual difference, consists in the presence or absence of papillary elevations between the septal bands. I have already spoken (^[ 142) of the frequent existence of symmetrically arranged spots, sometimes slightly depressed, but more commonly elevated, that are distinguished from the rest by the semitransparency of their shell; these spots are sometimes considerably enlarged, and their elevation increased, so that they become prominent papillae, closely resembling the tubercles upon the septal bands. Their size and disposition vary considerably. Sometimes they are small, numerous, and scattered without any definite arrangement over the entire surface of the wall of the chamber, whilst in other cases they are con- siderably larger, and form single, double, or even triple rows between the septal bands (Plate V. fig. 10). Another remarkable variety of external aspect is produced by the elevation of the general surface into rounded eminences closely abutting on one another (like the pustules of the skin in a case of confluent small-pox), and distinguished from the preceding by the absence of any peculiarity in the texture of the shell. That these and other analogous variations of surface-marking have no value as differential characters, is at once demonstrated, not merely by their gradational approximation in different indi- viduals, but by the fact that they are presented in very different degrees on different parts of the surface of the very same shell ; the chambers of one part of the spire being strongly marked by certain of these peculiarities, whilst those of another may only present indications of them, and those of a third may be perfectly smooth. 148. The collection of Mr. CUMING, however, contains a group of forms which are at once distinguished from the rest by their general physiognomy (Plate III. figs. 11, 12), and which, when their characters are examined in detail, appear to be separated from them by well-marked differences. Their aspect is much more lustrous, and their hue much whiter, varied, however, by a tinge of green diffused in irregular patches ; the spire does not in the largest specimens make above three turns, and it begins to open out sooner than in the type already described; the septa are usually considerably more convex anteriorly, and are also rather more distant from each other, so that the interval between them is greater in proportion to the breadth of the spire, the shape of the GENUS OPERCULINA: EXTERNAL CHARACTERS AND INTERNAL STRUCTURE. 19 chambers being thus modified, and the number of chambers in each whorl being dimi- nished ; the septal bands, not merely of the earlier whorls, but even of the last-formed portion of the shell, are raised into prominent tubercles ; and multiple rows of large tubercles are seen between the septal bands. It is to this multiplication of smooth tubercles composed of a variety of shell-substance which reflects light much more strongly than the rest, that the more glistening aspect of this type is chiefly due ; and I have not met in it with any instances of that general brownish coloration of the surface which is the ordinary characteristic of the other. The most constant and remarkable distinctive feature of this type, however, is the presence^of a large hemispherical cluster of semitransparent tubercles in the centre of the spire (Plate III. fig. 10). 149. Having had no difficulty in setting apart a large number of specimens agreeing very closely with each other, and differing from the rest, in all the foregoing characters, I should have arrived at the unhesitating conclusion that this type deserves to take rank as a species of Operculina distinct from the preceding, were it not for the circumstance that every here and there I met with an example in which the differential characters were less strongly marked than usual. Thus in some individuals which preserve the general proportions of the spire, the green coloration is wanting, the large central cluster of tubercles is replaced by a single tubercle of extraordinary size, the septal bands of the neighbourhood are not more tuberculated than in many examples of the ordinary type, and the rows of tubercles over the chambers are either wanting alto- gether, or are not more prominent than in many individuals of the preceding type. This evidence of the negative value to be attached to the number or prominence of the tubercles as a specific character, is confirmed by a curious fact of an opposite nature ; namely, that in certain individuals we find them developed to an extraordinary and obviously abnormal degree (Plate V. fig. 9). From the difficulty of deciding to which type particular specimens are to be referred, I had been almost led to adopt the con- clusion of the specific identity of this with the ordinary form ; when Dr. GOULD, of Boston (N. E.), kindly placed in my hands some Operculina, which had been collected on the coast of Japan by the recent American expedition to that country. These speci- mens combined in so remarkable a manner the most distinctive features of the two types, namely, the general form and proportions of the one, with the umbilical hemispheric cluster of tubercles and the general abundance of tubercular elevations characteristic of the other, as to remove all doubt from my mind with regard to their specific identity. 150. Internal Structure. The study of the internal organization of Operculina may be prosecuted in two modes ; by the examination, under sufficient magnifying powers, of thin transparent sections taken in different directions ; and by viewing under a low power, as opaque objects, fragments obtained by breaking the shell, especially (as Mr. CARTER was the first to suggest) after these have been allowed to absorb carmine or indigo by being placed upon water in which either of these colours has been rubbed up. By the former method alone can certain minutiae of structure be detected ; but the latter is extremely serviceable (as I found in my study of Nummulites, op. cit. ) in enabling the observer to trace out the relations of various parts, which sections exhibit to him D2 20 DE. CABPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. disconnected from one another. In Plate I. fig. 3, an attempt is made to bring together in an ideal representation the most important features of the internal structure of Oper- culina, as disclosed by these two methods of examination ; the warranty for all its details being supplied by the delineations of the separate parts that are given in other figures, which are all taken from preparations in my possession. 151. Owing to the circumstance that large specimens of Operculina are rarely if ever quite flat, it is next to impossible to make a section through the median plane that shall traverse all the whorls to the central cell ; a sufficiently near approach to this, however, has been made in the section represented in Plate VI. fig. 3, and Plate IV. fig. 10, to bring into view the general disposition of the chambers. This is much less regular than would be supposed from a superficial examination of the exterior. For while there is a certain general average in the proportion which their long diameter (that is, the breadth of the whorl) bears to the short (or distance between the septa), which may be stated as about 4^ to 1, this is by no means constantly maintained, the long diameter being some- times as much as 7 times, and sometimes no more than 2^ times, the short. The ordi- nary course of the septa, too, is often strangely departed from, as is seen in Plate VI. fig. 3. There are few individuals which do not present besides abnormal sinuosities, greater or less in degree very marked irregularities in the conformation of the cham- bers, analogous to those which I have described inNummulites*. Frequently a septum, instead of passing continuously from the inner to the outer margin of the whorl, stops short without reaching the latter, and bends backwards to join the last-formed septum; and sometimes a second septum unites itself to the first in the same manner, as shown at a, Plate VI. fig. 3. The abortion is often still more marked ; thus at 6 we see three septa thus interrupted, of which two do not traverse half the distance, and the third not a quarter, thus dividing the space between two complete septa into four small cham- bers along the inner margin, and one large irregular chamber extending to the outer. Another case of the same kind is seen in an earlier part of the same whorl; and in Fig. IV. an additional illustration on a larger scale is given of the like irregularity, for Fig. IV. Irregular disposition of septa in Operculina : a, a, a, normal apertures at inner margin of spire ; b, V, b", apertures of communication between abnormally divided chambers. . the purpose of showing the arrangement of the communicating passages. Generally speaking, the more nearly we approach the centre of the spire, the more regularly do we find the septa disposed, until we come into close proximity with the central cell. * Quarterly Journal of the Geological Society, vol. vi. 1850, p. 23. GENUS OPEECULINA: INTEENAL STEUCTUEE. 21 In Plate IV. fig. 5 we have an illustration of what may be considered the normal mode of commencement of the spire ; from which it will be seen that it originates, as in Fora- minifera generally, in a spheroidal cell, from which others are successively developed around it, the earlier chambers having no very definite shape, but those which succeed them gradually coming to assume the characteristic form and proportions. The like is shown in vertical section in Fig. VII. B. 152. Each chamber communicates with the neighbouring chamber on either side, by a long narrow crescentic fissure left by the non-adhesion of the septum to the outer margin of the preceding whorl (Plate I. fig. 3). This fissure (Figs. IV., VI. a, a, a) is frequently not to be seen in such a section as is represented in Plate VI. fig. 3, owing to the circumstance that this does not happen to pass through the plane in which it lies ; and it is best brought into view either by making thin transverse sections (as Fig. V.), or Pig. V. Vertical or transverse section of three outer convolutions of Operculina. by breaking a specimen transversely and examining its fractured edges, by which such views will be obtained as are presented in Fig. VIII. (p. 26) A, B, c, D. Besides this prin- cipal aperture, we observe in the septa, especially of the larger chambers, a variable number of secondary pores (Plate I. fig. 3 e), generally circular, and of comparatively small size; these are dispersed without any regularity, as is shown in Figs. V., VI., VIII. They may or may not be brought into view in a horizontal section, according as its plane does or does not happen to pass through them ; but when they are thus traversed (Fig. VI. b, b, 5), it is seen that these secondary pores, like the principal aperture, establish a direct communication between adjacent chambers*. In the irregular formation repre- Fig. VI. Section of two septa of Operculina through the median plane, showing the secondary pores, I, b, I, in addition to the principal orifices, a, a, at the inner margin of the spire. * These secondary pores were first observed by me in Nummulites l., Zool. torn. iii. p. 312. GENERAL INTRODUCTION. 185 merits of locomotion, are effected by means of pseudopodial prolongations of the sarcode, put forth through apertures in the shell, and capable, when retracted again, of coalescing with the general mass. In the case of the composite forms, he con- sidered the entire animal to be made up of a series of segments which are essentially repetitions one of another, each possessing an independent vitality of its own*. These statements have been subsequently confirmed and rendered more precise by several other observers ; their truth has been admitted by M. D'ORBIGNY, who, in all his recent works, has described the animals of the Foraminifera in accordance witli them (though without any allusion to the fact, that he had himself previously laboured under an entire misconception of their character, and without any mention of the discoverer of their real nature) ; and they have been recently placed beyond all doubt, by the admirable researches of Professor SCHULTZE (Op. cit.). It cannot but seem surprising, that notwithstanding the light thrown upon this inquiry by M.DUJARDIN in 1835, Professor EHRENBERG should in 1838 have announced to the Academy at Berlin, his conclusion, professedly based on observations of certain forms of these animals in their living state, that their true place in the animal kingdom is among the Bryozoa-\-. He described them as possessing a distinct alimentary canal, which extends from segment to segment ; this, however, instead of being single, as in Nonionina, may (he tells us) be multiple, as in Geoponus; so that we must regard each segment of the latter, however apparently resembling the simple segment of the former, as in reality composed of several adhering bodies. In one instance (he affirms) he found the mouth surrounded by a plumose sensory and pre- hensile apparatus, like that of the Fiustrce and Halcyonellue (see ultra, ^| 4.), but generally speaking he admits that this is altogether wanting, the mouth being a simple aperture. He saw minute extensile tentacula proceeding from all parts of the sieve- like shell, as described by DUJARDIN, and admitted their resemblance to the pseudo- podia of Difflugia, &c., but he remarks, " the rest of their organization, which DUJARDIN has overlooked, removes them from the Infusoria, quite as far as from a chaotic primitive substance." Besides the alimentary canal, Professor EHRENBERG describes a yellowish-brown granular mass as accompanying and sometimes surround- ing it up to the last of the spirals ; this he considers as an ovary. As I have reason to believe that Professor EHRENBERG stands quite alone in this opinion (if, indeed, he still maintains it), and that the real nature of the segments of sarcode and of their connecting threads, is no longer a matter of question among those Naturalists who have given their unprejudiced attention to the subject, I do not think it requisite to occupy either time or space with any further discussion of the question, and therefore dismiss it with this brief mention. * See his " Histoire Naturelle des Infusoires," Paris, 1841 ; and Art. Rhizopodes in Diet. Univ. d'Hist. Nat., torn. xi. p. 115, Paris, 1848. t See his Memoirs in the Transactions of the Royal Academy of Berlin, for 1839 and 1840; also TAYLOR'S Scientific Memoirs, vol. iii. p. 319. 186 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. It may, however, be well here to remark in limine, that it obviously makes a most essential difference in our appreciation of the value of the characters afforded by the form, position, and multiplication of the apertures of communication between the chambers of the shell, whether we regard these as giving passage to an organ of such fundamental importance as an alimentary canal, or whether they merely serve for the connexion of the different segments by stolons of sarcode. For variations, which in the former case must be regarded as indicative of such essential differences, both in structure and function, as would rightly characterize distinct genera or even distinct families, may easily be admitted, on the latter view, to be of such comparatively trivial moment, as to rank no higher than specific characters, or perhaps even to be matters of individual difference. That the latter is the true view of the case, I have become completely assured in the course of my researches ; and I shall hereafter be able to adduce some curious illustrations of it. Turning now to the more recent History of research, I shall briefly notice those investigations which have done most towards the advance of our knowledge of the organization and physiology of the Foraminifera ; the mere collection, description, and systematic arrangement of new forms, without any such advance, being no more a feature of progress, than is the building-up of an edifice, which must necessarily fall, through the insecurity of its foundation, before it shall have been completed. The first series of these, made by Professor W. C. WILLIAMSON of Manchester, upon Polystomella crispa*, not only established several important facts in regard to its minute structure, but may be regarded as having furnished the starting-point for all future investigations of a like kind. Among these facts were several that became of essential value to myself, in the inquiry on which I was engaged at the same time, in regard to the structure of Nummulites ; and served to confirm the inferences which I had deduced from the other features of that important type, as to its participation in the characters of the Foraminifera generally. In the course of that inquiry I made the discovery-^-, not only of a most elaborate and previously-unsuspected structure in the shell itself, but also of a system of interseptal canals, which established a com- munication between the inner segments and the external surface, much more direct than that which they possess through the series of segments which form the outer turns of the spire. The existence of this system of canals has been verified, not merely in Nummulites by MM. D'ARCHIAC and HAIME (Op. cit.), but also in several recent types ; thus Professor WILLIAMSON has detected it \nAmphlstegina and JVonionina^, and more recently in Faujasina^ (which furnishes one of the most, remarkable examples of it) ; whilst Mr. CARTER of Bombay has discovered it in Operculina\\. My own inquiries, which have been carried-on with scarcely an intermission, from the time of my first * Transactions of the Microscopical Society, 1st ser. vol. ii. p. 159. t Quarterly Journal of the Geological Society of London, vol. vi. February 1850, p. 22. { Transactions of the Microscopical Society, 1st series, vol. iii. p. 105. Ibid. 2nd series, vol. i. p. 87. || Annals of Natural History, 2nd series, vol. x. p. 161. GENUS ORBITOLITES: HISTORY. 187 discovery of this remarkable point of structure, have been specially directed to the determination of the extent to which it presents itself in the different sections of the group, and of its value as a distinctive character; and I think that I shall be able to show that it is a feature of the utmost importance, the presence of which marks an elevation of type, and its absence a corresponding degradation. It is much to be regretted, that the recent investigations of Professor SCHULTZE should have been so entirely restricted to the structure of the animal, which can only be ascertained in a comparatively small number of cases ; and that he should have failed so completely in the determination of the internal organization of the shell*, which in a large proportion of instances is the only guide we possess to the nature of the being which formed it. The new classification which he proposes, whilst in many respects an improvement upon that of M. D'ORBIGNY, is essentially vitiated by this defect; and being in itself, therefore, just as provisional as that for which it is proposed as a substitute, can scarcely be expected to supersede it. My own researches I offer simply as materials to serve as a basis for classification ; feeling assured that the time is not yet come, in which the superstructure can be erected with any prospect of permanent stability. I shall commence with a minute analysis of one of the lowest types, Orbitolites ; and propose to show hereafter, that Orbiculina and Alveolina, though ranked in a different order by M. D'ORBIGNY, are in reality closely allied to it ; whilst a new genus (Cycloclypeus), which, in M. D'ORBIGNY'S arrangement, would rank close to Orbitolites, is physiologically separated from it by the widest possible interval. Genus ORBITOLITES. I. History, 1. The Orbitolite has been chiefly known, until very recently, rather by its fossil, than by its existing forms. The abundant occurrence of its disks in the Calcaire grassier of the Paris basin, early attracted attention ; but Orbitolites were not clearly distinguished by the older observers from Nurnmulites, and their true nature was entirely misunderstood. Thus we find them designated, often in association with Nummulites, under the title of Umbilicus marinus by PLANCUS (BIANCHI), who * Professor SCHTJLTZK states (op. cit. p. 15) his inability to discover the canal-system above described ; and as there is no question of his competency and accuracy as an observer, I can only impute his failure to his ignorance of the proper mode of preparing thin sections of these minute shells ; which consists in cementing them to a slip of glass by hardened Canada-balsam ; grinding them down as far as may be desirable on one side ; then loosening them by heat and turning them over, so that the flattened surface shall now be attached ; and finally grinding down the other side, until the requisite degree of thinness shall have been attained; after which a drop of liquid Canada-balsam is laid upon the specimen, gentle warmth applied, and a cover of thin glass put on. Having myself thus prepared sections of Faujasina, which answer in every respect to the figures accompanying Professor WILLIAMSON'S memoir above cited, I can bear the most explicit testimony to their exactness. 188 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. imagined them to be opercula of Ammonites*; of Porpitce nummulares by STOB^US-|~ and BROMELL|, who seem to have regarded them as representing the disks of the existing Porpitae; of Helicites and Operculites by GUETTARD , who considered them as opercula of Gasteropods ; of Discolithes by FORTIS ||, who supposed them to be skeletons of mollusks ; of Madreporites by DELUC, and of Milleporites by FAUJAS DE ST. FOND, whose idea of their nature is sufficiently indicated by the names they assigned to them. 2. The genus Orbitolites seems first to have been erected, and distinctly separated from Nummulite, by LAMARCK, in the first edition of his 'Animaux sans Vertebres,' its type being the O. complanata of the Paris basin. The following is his definition of the genus, which he ranks between Lunulites and Millepora, among his "Polypiers ForamineV': "Polypiarium lapideum, liberum, orbiculare, planum seu concavum, utrinque vel margine porosum, nummulitem referens. Pori minimi, adamussim dis- positi, conferti, interdum vix conspicui." These bodies, he says, are distinguished from Nummulites by the opening of their marginal pores, and by the absence of spiral arrangement in their minute chambers or cells. In his second edition (1816), he altered the name from Orbitolites to Orbulites ; but the latter designation having been previously employed in Malacology, the first appellation has been restored by M. MILNE -EDWARDS in his posthumous edition of LAMARCK'S work. Under one of the designations, Orbitolites or Orbulites, the genus has been recognized by SCHWEIG- GER^[, BRONGNIART and CUVIER**, LAMOUROux-f-f-, DESLONGCHAMPS^|, DEFRANCE^, BLAINVILLE ||||, BRONN^^, GOLUFUSS ***, MICHELIN "f~f~f", PICTET^^, and DUJAR- DIN; none of whom, however, have either given any account of its internal struc- ture, or made any essential modification in the definition of the genus, which they all left in the place which LAMARCK had assigned to it. 3. The existence of more than one recent species of the same type was indicated or expressly mentioned by several of the foregoing writers. Thus FORTIS tells us * De Conchis minus notis, 1739 (fide D'ARCHIAC et HAIME), and App. Phytol. F. Coll. 1764 (fide RUPERT JONES). t Dissertatio epist. ad W. GROTHAUS de nummulo Brattenburgensi, 1732; Opera petrefactorum, 1752; Opusculis, p. 6 (fide D'ARCHIAC et HAIME). J De Nummulo Brattenburgico, in Act. Litt. Suec., vol. ii. p. 50 (fide D'ARCHIAC et HAIME). Memoires sur differentes parties des Sciences et des Arts, 1770. || Memoires pour servir a 1'Hist. Nat. de 1'Italie, 1816, vol. ii. ; and Journal de Physique, 1801, vol. lii. p. 106. f Beobacht. auf Naturg. (1819), pi. 6. ** Ossemens Fossiles (1822), vol. ii. pt. 2. p. 270. ft Expos. Method, des gen. des Polypiers (1821), p. 44. pi. 73. U Encyclop. Method., Zooph. (1824), p. 584. Dictionn. des Sci. Nat. (1825), vol. xxxvi. pp. 294, 295. HII Manuel d'Actinologie (1830), p. 411. pi. 72. f f Lethsea Geognostica (1836-37), pi. 35. *** Petrefacten (1826-33), vol. i. p. 41. pi. 12. ftt Icon. Zoophyt. (1845), p. 167. pi. 46. JU Trait^ Element, de Paleontologie (1844-45), torn. iv. Diet. Univ. d'Hist. Nat., tome ix. (1847), p. 162. GENUS ORB1TOLITES : HISTORY. 189 (op. cit.), on the authority of a voyager in the Indian seas, that living Discolites have been found there ; and as there are no existing Nummulites, this statement probably refers to an Orbitolite. LAMARCK, in his second edition, describes, under the title of O. marginalis, a small form of Orbitolite, only 2 millims. ('03 inch) in dia- meter, found upon fuci, corallines, &c. in the European seas ; this he speaks of as the only living example of the genus then known, and he defines it as O. utrinque plana, margins poroso. Besides this species, however, DEFRANCE (loc. cit.) mentions another, more closely resembling the O. complanata of the Paris basin, as existing on the shores of New Holland ; and this seems the first clear indication of the body (afterwards found by MM. QUOY and GAIMARD in that locality, and erected by them into the distinct genus Marginopora}, the structure of which constitutes the chief subject of my present communication. Both these existing forms are described by BLAINVILLE (op. cit.), the first from actual observation, the second on the information of MM. QUOY and GAIMARD, to the manuscript of whose 'Voyage de FAstrolabe' (then unpublished) he refers as his authority. It is singular, however, that after an attentive search through the published "Zoology" of that work, I have not been able to find, either in the text or in the plates, any mention of Marginopora or of Orblto- lites. Of the Orbitolites marginalis, M. DE BLAINVILLE says (op. cit. p. 412), " Nous 1'avons etudiee avec soin ; et nous sommes presque convaincu que ces petits corps cre'tace's ne sont pas de veVitables polypiers ; mais bien quelque piece inte"rieure, qui s'accroit par la circonference. II est en effet evident, qu'il n'y a pas de cellules pro- prement dites, a moins qu'on ne veuille regarder comme telles les deux plans de locules qui occupent le bord, et qui n'offrent rien determine*. Tout le reste est cou- vert d'une legere croute cre'tace'e, qui ferme les anciens pores." I think it obvious, from this description, that it was founded on specimens resembling that in Plate VII. fig. 8, in which the marginal row of cells has been laid open above and below by accidental abrasion ; and that the true marginal pores, opening between the protu- berances formed by the cells (Plate V. fig. 1), were overlooked. The genus Margi- nopora, placed by M. DE BLAINVILLE in immediate sequence to Orbitolites, is thus characterized (loc. cit.): " Animaux inconnus, contenus dans les cellules poriformes, excessivement petites, rondes, senses, ^parses dans les sinuosity's, tres fines et tor- tueuses, qui guillochent la circonference d'un polypier calcaire, libre, un peu irre- gulier, disco'ide, concave ou concentriquement strie en dessus comme en dessous, et plus e"pais sur les bords." The two surfaces, M. DE BLAINVILLE further tells us, only exhibit striae of increase, without any trace of pores ; but the turned-up edge is entirely riddled with very fine rounded pores, which are situated in the sinuosities of a very close but shallow engine-turning (guillochis). And when one of its surfaces is rubbed away, the disk is found to be formed of concentric canals, separated by partitions, and themselves divided into cells, thus recalling in some degree the struc- ture of Orbitolites. Having myself had the opportunity of inspecting, by the kind- ness of M. VALENCIENNES, the specimens of Marginopora on which the foregoing MDCCCLVI. 2 c 190 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. description was founded (these being now contained in the ' Muse'e d'Histoire Natu- relle'), I have been able to compare them with my own; and finding that they correspond with the peculiar type of the latter, which is represented in Plate V. figs. 2 & 3, I have no hesitation in saying that in this description also the true marginal pores, represented in Plate V. fig. 6, have been overlooked; and that what are described by M. DE BLAINVILLE as pores, are nothing else than incomplete cells left open in the frilled edges which bound the marginal furrow above and below (see ^[ 25). A similar description has since been given by M. DUJARDIN *, who does not hesitate to regard the disk as a polypary, and to speak of the animals whereby it is formed, as polypes. 4. In the Memoir of Professor EHRENBERG already referred to, we find the genus Orbitolitea for the first time associated with true Foraminifera, as a member of his class BRYOZOA, order Polythalamia, suborder Polysomatia, family jlsterodiscina, wherein it is placed next to Lunulites, which undoubtedly belongs to the group of Bryozoa (Polyzoa) as now restricted. This family he characterizes as follows : " Gemmis in eodem piano prodeuntibus, polypiaria plana, discoidea, formantibus, osculis distinctis post mortem apertis ;" and it is by the last of these characters that he distinguishes it from the family Soritidce, consisting of the two genera Sorites and jjmphisonts, of which he says, "Osculis contracto corpore, tanquam operculo duro clausis." If any faith whatever is to be placed in Professor EHRENBERG'S figures and descriptions, his Sorties is nothing else than LAMARCK'S Orbulites marginalis ; whilst his Amphisorus, which differs from Sorites merely in having two layers of cells instead of one, is (as I shall hereafter show) the same type in a higher grade of development. I cannot conceal my astonishment, however, that so practised a microscopic observer should have entirely overlooked the real marginal openings between the cells; still more, that he should have described the entirely-closed cells of the surface as covered in by a moveable operculum, which merely shuts their orifices when the animal is contracted ; and further, that, mistaking an accidental for a normal opening of some of the cells, he should have ventured to figure an eight-armed Bryozoon as issuing forth from one of them, a phenomenon which, I do not hesitate to say, is entirely irreconcileable with our existing knowledge of the organization of the animal of which these disks are the skeletons. 5. The earlier publications of M. D'ORBIGNY on the subject of the Foraminifera do not include any notice of this genus ; and neither in the systematic arrangement which he put forth in his article in the 'Diet. Univ. d'Hist. Nat.' tome v. (1844), nor in that contained in his 'Forarn. Foss. de Vienne' (1846), is the order Cyclostegues recognized, which makes its appearance for the first time in his 'Cours Elementaire de Paleontologie,' tome ii. (1852), between the Monostegues and the Helicostegues, with the following definition (p. 192) :" Animal compose" de segments nombreux, places en lignes circulates. Coquille disco'idale, composee de loges, concentriques, * Diet. Univ. d'Hist. Nat., tome vii. p. 777. GENUS ORBITOLITES: HISTORY. 191 simples on multiples; point de spirale." Under this order are ranged the genera CycloUna (D'ORB. 1839), Orbitolites (LAMARCK, 1801, and Marginopora, QUOY et GAIMARD, 1836), Orbitollna (D'ORB. 1847), and Orbltoides (D'ORB. 1847). As I shall have occasion to show that the first three of these genera cannot be separated from each other by any valid distinctions, and that the greater number of the species ranked under them by M. D'ORBIGNY (see his 'Prodrome de Pale"ontologie Strati- graphique') really belong to the same specific type, I must here cite the generic cha- racters which he assigns to them : " CycloUna, Coquille disco'idale, chaque loge perc6e de nombreux pores, faisant un circle entier autour des autres. L'espece con- nue est de 1'etage cenomanien. Orbitolites, Coquille disco'idale, plane, e*gale, et encroutee des deux cote's, pourvue de lignes concentriques. Loges nombreuses, par lignes irre'gulieres, transverses, visibles seulement au pourtour. Nous connais- sons deux especes ; les premieres, de 1'^tage suessonien ; le maximum, dans les mers actuelles. Orbitolina, Ce sont des Orbitolites a cote's ine"gaux ; 1'un, convexe, eucroute, a lignes concentriques ; 1'autre, concave, non encroute', rnontrant des loges nombreuses, par lignes obliques sur le cote", au pourtour. Nous connaissons de ce genre perdu six especes ; les premieres, de l'e"tage albien ; les dernieres, de l'6tage senonien." 6. The first approach towards a more accurate knowledge of the real nature of the Orbitolite, through an examination of its internal structure, was made (I believe) by myself, in my Memoir " On the Microscopic Structure of Nummulina, Orbitolites, and Orbitoides," read before the Geological Society of London in November 1849, and published in its Quarterly Journal for February 1850. The place assigned to this genus in the system of M. D'ORBIGNY not having been at that time made public, and all other zoologists and palaeontologists having ranked it in close approximation to Lunulites and other Polyparies of the Bryozoic (Polyzoic) kind, I entered upon the examination without the least suspicion that this organism was to be regarded in any other light ; and that Iwas not undeceived in the course of it, may be attributed to the small number of specimens then placed at my disposal for the inquiry by my late friend Professor E. FORBES, and to the circumstance that these specimens were of the type that presents most resemblance to that of a Bryozoic polypary, and were all deficient in the central nucleus, which is the portion most indicative of their Foraminiferous nature. Nevertheless, the marked dissimilarity in structure which I found to exist between the disk of Orbitolite and the polypary of Lunulite or of any other undoubted members of the Bryozoic group, made me even then express myself doubtfully as to its title to be associated with them. In this Memoir, published two years previously to M. D'ORBIGNY'S first announcement of the fact, I showed that the genus Marginopora must be abolished, since its sole representative is so closely allied in structure to the Orbitolite of the Paris basin, that no doubt of their generic iden- tity can be entertained ; the existing M. vertebralis, in fact, being only specifically distinguishable from the fossil O. complanata, by a difference in the form of its super- 2 c 2 192 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. ficial cells, a character to which the more extended knowledge of this type has made it clear to me that not the slightest value is to be attached (see Sect. IV.). I showed that in one as in the other, the cells are normally closed-in over the whole surface; that the cells of the two surfaces are separated from each other by an intervening stratum, traversed by a set of round cells or passages of its own, with which each superficial cell communicates by two small apertures ; and that the only real exter- nal orifices are the minute pores at the margin of the disk, which communicate, not directly with the cells of the superficial layers, but with the passages of the inter- mediate stratum. 7. A very important addition to this limited measure of information was made not long afterwards by Professor WILLIAMSON of Manchester, who, in a paper read before the Microscopical Society of London, June 12, 1850, and published in its Transac- tions (First Series, vol. iii. March 1851), first gave a minute description of the small recent Orbitolites marginalis of LAMARCK (which, however, he designated as O, com- planata); contrasted it with that of a large recent Orbitolite from Tonga, shown by his description to be identical with the Marginopora of QUOY and GAIMARD, to my pre- vious account of which he added some important particulars ; and compared both with the well-known Orbiculina adunca of the Bahamas. His investigations led him to the conclusion, that these three forms should rank under the same genus Orbi- culina; their most important structural features being common to all, whilst their differences are only of specific value. His Orbiculina complanata (really Orbitolife-s marginalis) he characterizes by its spiral commencement, and by its possessing only one layer of cells ; his O. Tonga (Marginopora) he characterizes by its cyclical com- mencement, and by its possession of two superficial layers of cells, with an interme- diate stratum, as I had previously pointed out; whilst the O. adunca he shows to commence on the spiral type, and to carry it on much further than his O. compla- nata, but finally to assume the cyclical, and then to correspond very closely with his O. Tonga. The title of the genus Orbitolites to a place in the group of Foraminifera, in near proximity to, if not in union with, Orbiculina (which had been ranked as such by M. D'ORBIGNY from an early period of his investigations), and the entire absence of any ground for ranking it among the Bryozoa (Polyzoa), were clearly established by Professor WILLIAMSON in this valuable memoir; and though I shall hereafter have occasion to show that some of his conclusions were erroneous, yet I regard them as fully justified at the time, by the information which the materials at his command afforded ; my own means of correcting them being only supplied by the comparison I have been enabled to institute through a much wider range of specimens. And it is with great satisfaction that I am enabled to add, that, after a careful inspection of my preparations and drawings, Professor WILLIAMSON authorises me to express his full accordance in the results which I shall now proceed to detail. 8. As the value of such details cannot be truly estimated without some knowledge of the range of the observations from which they are derived, I think it right to state GENUS ORBITOLITES: GENERAL PLAN OF ORGANIZATION. 193 in limine what are the opportunities I have enjoyed. The first and most important of these was afforded me by the great abundance of Orbitolites of various sizes, from o to ( ), besides a vast multitude of fragments, in Mr. JUKES'S Australian dredgings ; besides which, Professor E. FORBES kindly put into my hands several spe- cimens which had been taken in their living state by Mr. JUKES, from the marine plants to which they were attached. All these I have carefully examined under the microscope, so as to be able to make an exact determination of their external cha- racters ; and of a large proportion I have made microscopic sections in various direc- tions, that I might assure myself of every particular respecting their internal struc- ture. When I had nearly exhausted these sources of information, I found a new and most interesting series of specimens in Mr. CUMING'S Philippine collection; and by the study of these I was enabled to test the validity of the conclusions, to which I had been led by the examination of the Australian forms. The kindness of various friends has further enabled me to examine specimens obtained from other widely-distant localities, such as different parts of the Indian, Southern, and Pacific Oceans, the Red Sea and the /Egean. And finally, I have had placed at my disposal, through the instrumentality of Professor QUEKETT, several Orbitolites of various sizes and ages, obtained on the shores of the Feejee Islands by the late Sir EVERARD HOME, in which the animal substance occupying the interior had been preserved by the immer- sion of the fresh specimens in spirits. 9. Besides the existing forms, I have examined a large number of specimens of Fossil Orbitolites, both from the Paris basin and from other localities ; and I have instituted the same kind of minute comparison of these specimens, both with each other and with the recent types, that I had previously made among the diversified forms of the latter. 10. Postponing, until adequate means shall have been supplied by the details of their organization, to be presently given, the inquiry into the relationship of these different forms (Sect. VI.), we shall proceed in the first place to consider the general plan of organization of the Orbitolite, and then to study the variations to which this is subject. II. General Plan of Organization. 11. In studying the organization of the Orbitolite, we shall have recourse, on the one hand, to the structural characters presented by the Animal, as displayed by the fleshy residuum left after the decalcification of specimens in which it has been pre- served ; and on the other, to the structural characters presented by the Shell, of which some are visible on its surface, and without any preparation, whilst others may be seen in thin specimens by transmitted light, but of which the greater part can only be brought into view by thin sections taken in various directions, especially horizontal or parallel to the surface, and vertical or perpendicular to the surface. 194 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. 12. The state of preservation of tlie animal body of the Orbitolite, in the spirit- specimens which I have examined, is so complete, as to leave me no room for hesi- tation in affirming that it corresponds in every particular with the 'sarcode' which we have seen to have been first described by M. DUJARDIN, as constituting the bodies ' of many of the lowest organisms, and especially as the component of those of the Rhizopoda. A small portion of this substance, sufficiently magnified to exhibit its nearly homogeneous jelly-like aspect, with minute granules and somewhat larger par- ticles scattered through it, is shown in Plate IV. fig. 2. Although it is so far decolo- rized in spirit-specimens as to present only a brownish hue, yet as specimens that have been gathered fresh and have been then dried, possess a reddish aspect, and as this is not due to the shelly substance, it may be presumed that the sarcode of the living Orbitolite has the same bright red colour as that of Rotalia and many other Forarninifera. The entire animal body (Plate IV. fig. 1) is composed of a numerous assemblage of minute segments, arranged at tolerably regular intervals in concentric zones around a sort of central 'nucleus;' the segments composing each zone being united with each other by a continuous annular 'stolon' or band of sarcode, and being connected with those of the adjoining zones by peduncles of the same material. I have not met with the least indication that the sarcode is contained within any proper membrane ; and the absence of any such indication, notwithstanding the various manipulations to which I have subjected its segments, may be taken, 1 think, as strong negative evidence that it has no more existence in this animal, than it has in the species of Foraminifera which have been so well studied by M. DUJARDIN and Professor SCHULTZE. Nor is there the slightest trace of distinct organs, either in the mass of sarcode which forms the central nucleus, or in that which constitutes each one of the surrounding segments ; and he would, I think, be a mere speculator, who should maintain the presence of a digestive cavity in any of these parts, or the exist- ence of an intestinal canal in the peduncular threads which connect them together. The homogeneity of the component substance of the central nucleus, and of the entire assemblage of multiple segments, seems, indeed, to be conclusively established by the following facts: In all the spirit-specimens which I have examined, the cavi- ties of the outer zones are completely void, whilst those of the nucleus and of the inner zones are quite filled with their animal contents. This drawing-together of the soft body towards the centre, is evidenced also in many of the larger specimens which have been dried when collected in the living state, by the limitation of the red colour that indicates the presence of the sarcode, to the inner portion of the disk. In both cases it may be presumed that the animal matter has shrunk together, in the former through the corrugating action of the spirit, in the latter through desiccation. Now if the polypidom of a zoophyte be similarly treated, there is no such drawing together of the entire body, but each cell is found to contain the shrunk contents of its own polype-segment ; and this difference seems to me to indicate a complete dis- similarity in the characters of the two organisms. For it is obvious that the sub- GENUS ORB1TOLITES: GENERAL PLAN OF ORGANIZATION. 195 stance of the peripheral segments of the Orbitolite-body can only be brought together towards the centre, through being completely unattached to the walls of the cavities which it occupies, and through having a form so alterable, as to be capable of being- drawn in threads through the narrow connecting passages, and of then coalescing together again so perfectly, that the masses they form do not present the least trace of having been thus spun out. There is no known kind of animal texture, except sarcode, that is susceptible of this kind of alteration ; and the evidence of it, which I have adduced seems to me extremely valuable, not only as establishing the general nature of the animal body of the Orbitolite, but also as fully justifying the assump- tion, that, in the living state, the sarcode is projected in pseudopodia through the marginal apertures, and that alimentary particles are introduced by their instrumen- tality, as in other Foraminifera. 13. Turning from the animal body to the calcareous disks which enclose it, we find that, whether large or small, these are almost invariably circular, or nearly so; that they are usually nearly flat, any difference in thickness being generally in favour of the marginal portion; and that if, as sometimes happens, there is a slight central projection, this is formed by the nucleus alone. By these characters we may distin- guish Orbitolites from Orbiculina ; for although the discoidal forms of the latter so strongly resemble Orbitolites, that by the structure and arrangement of their mar- ginal portion they could not be distinguished, yet they may always be discriminated by the knobby protuberance of their centre, which is occasioned by the mutual investment of the earlier whorls of the spiral in which they commence. The same entire absence, or very small size, of the central elevation, together with the uni- formity or even slight increase of thickness towards the circumference, also helps us to separate Orbitolites from Orbitoides ; the centre of the latter being always considerably elevated, and the thickness of its disk ordinarily diminishing gra- dually towards its margin*. Around the ' nucleus' which occupies the centre of the disk (Plate V. figs. 1, 6), are seen an indeterminate number of concentric zones of cells (c, c, c), the shape of which differs in different individuals (see Sect. IV.) ; these, although completely closed (unless laid open by abrasion), have their form * I wish this statement to be understood with reference to the genus Orbitoides, as characterized by the structure which I have shown it to possess (Quart. Journ. of Geol. Soc., Feb. 1850), and not to the genus as denned by M. D'ORBIGNY (Cours Elementaire de Paleontologie, tome ii. p. 194), who, notwithstanding that he has shown himself to be acquainted with my Memoir (by copying from it a figure of Nummulite), has not profited in any degree by my investigations, but has left the generic characters of Orbitolites, Orbitolina and Orbitoides in the state in which they might have been, and probably were, before that Memoir was published. The true distinction, however, has been fully recognized by M. D'AHCHIAC, who, in his ' Description des Ani- maux Fossiles du Groupe Nummulitique de 1'Inde,' p. 349, has designated as Orbitoides dispansa and Orbitoides Fortisi, the bodies which, in the account of them he had previously given in the ' Me'm. Soc. Geol. de France,' 2nd ser., vol. iii., he had designated as Orbitolites ; thus correcting the error into which Mr. CARTER has fallen in his description of the same fossils, through reliance on M. D'ORBIGNT'S insufficient and indeed erro- neous characters of these genera. 19(5 DR. CARPENTER'S RESEARCHES ON THE FORAMIN1FERA. distinctly indicated by the surface- mark ings. The only external orifices which communicate with these cells, are the minute pores (d, d, d) forming one or more ranges at the margin of the disk, each pore lying in a vertical furrow, between the projecting walls of two contiguous cells. In the smallest and thinnest disks (Plate V. fig. 1) we find but a single row, or sometimes two rows, of such pores; in disks somewhat thicker, there are three or four rows; and in the largest and thickest Orbitolites (Plate V. fig. 6), no fewer than ten or twelve such rows. This multiplication in the number of ranges of marginal orifices, indicates a like mul- tiplication in the number of floors (so to speak), of which the disk is composed ; and just as the total number of chambers in a building may be increased, either by extending its base over a larger area, or by additions to its number of storeys, so may an increase in the number of segments of which this animal is composed be provided for, either by the marginal addition of a new zone resembling the last, so that the diameter of the disk is alone augmented, or by an increase in the thick- ness of the newly-forming zone, so that it contains a larger number of superposed layers. The new zones, however, never invest or cover those which they surround, each being simply a continuation of the margin of the preceding; and in this respect the mode of growth of Orbitolites at every stage is pointedly distinguished from the early mode of growth of Orbiculina, just now specified as the cause of the protube- rance of its centre. 14. The shelly substance of the calcareous disk, although firm, is by no means so dense and bony as that of the shells of many other Foraminifera of higher organiza- tion. It is apparently quite homogeneous, rarely presenting the least appearance of 'structure,' and this being probably fallacious; I refer to the punctated marking sometimes seen on the outer surface of the nucleus, which shows itself under the aspect represented in Plate VI. fig. 5, when the thin layer of shell which presents it is viewed by transmitted light. Although this appearance might be considered to indicate the existence of a cellular structure in the shell, yet I believe that such an inference would be fallacious ; since I have not been able to detect the least trace of such a structure in the decalcified residuum, which, on the other hand, seems to me to be a substance as structureless as sarcode itself. Coupling these appearances with those which I have found to exist more distinctly in Orbiculina, I am disposed to interpret them as proceeding from minute depressions on the surface; and these are perhaps to be regarded as .the rudiments of those minute closely-set apertures, which, in many Foraminifera, give passage to pseudopodial extensions of the sarcode from every part of their bodies. 15. In all the forms of Orbitolite that I have examined, the central Nucleus pre- sents the same essential characters. When the interior of any disk, whether large or small, is laid open by a horizontal seotion passing through the central plane, the nucleus is seen to be occupied by a large cavity (Plate V. fig. 1) somewhat irregu- larly divided by a sinuous partition, which always, however, marks out a central GENUS ORBITOLITES : GENERAL PLAN; SIMPLE TYPE. 197 cell (a) of a somewhat pyriform shape, as distinct from the space (b, b) which sur- rounds it. The meaning of this feature is at once made apparent, by reference to the disposition of the sarcode which occupies the cavity of the nucleus ; for we then see (Plate IV. fig. 5) that the large central pear-shaped cell is occupied by a mass (a) of corresponding shape, from the small extremity of which a peduncular process extends, that dilates again into a still larger mass (b, b) completely surrounding that from which it springs; the former may be conveniently designated as the 'central,' the latter as the ' circumambient ' segment. In a vertical section of the disk, passing through the centre, such as that seen in Plate V. fig. 4, the nucleus seems to present three chambers ; but this is simply due to the fact that such a section will traverse the circumambient cell twice, that is, will cut it through on both sides of the central cell. In the section represented in Plate V. fig. 9, there are but two central cham- bers ; in this case the plane of division seems to have traversed the nucleus just where the neck of the central cell touches its margin, so that the circumambient cell is only on one side of it. If, on the other hand, the plane of division should happen not to pass through the central cell at all, so as to traverse the circumambient cell alone, a single broad cavity will present itself in the vertical section, as shown in Plate V. fig. 7- Frequently, however, it happens that the circumambient segment is partially subdivided on one side by an interposed partition (Plate VII. fig. 4) ; and then a ver- tical section will show four chambers, as is seen in Plate V. fig. 10, the central seg- ment having a single portion of the circumambient segment on one side of it, and a double portion on the other. Some remarkable varieties in the size of the nucleus, and in the mode of its connexion with the surrounding parts, will be noticed here- after (f ^[ 44-46, and 54, 55). 16. In describing the structure of the Concentric Zones which successively sur- round the nucleus, it will be requisite to make a distinction between the simple and complex types according to which the Orbitolite-disks may be generated; the former being characterized by the existence of only one layer or ' floor ' of segments, the latter by the presence of two or more such layers. For although, as will subsequently appear, I can show, by a series of transitional gradations between these two types of structure, and by their occasional coexistence in the same individual, that they are not to be held to characterize distinct genera (as Professor EHRENBERG supposed), or even distinct species (as Professor WILLIAMSON has urged with more apparent reason), yet, when most characteristically displayed, they differ so much from one another, and each is so remarkably distinguished by features of its own, that it seems more advantageous to describe them separately in the first instance, and then to discuss their relationship to each other. 17. The Simple type is found to prevail in those minute Orbitolite-disks, which occur in greater or less abundance in sands and dredgings from almost every part of the globe, but which are particularly numerous in those of the Philippine shores. Their ordinary diameter is about '05 of an inch, and they usually contain from ten MDCCCLVI. 2 D 198 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. to fifteen concentric zones. The arrangement and connexions of these zones may be made out in the thinnest and most translucent specimens, by examining them by transmitted light, after mounting them in Canada balsam ; this, however, gives such transparence to the thin shelly layer which is continuous over both surfaces, that it may escape notice (if not carefully looked for), so as to lead to the conclu- sion that the cells are open. Most specimens require to be somewhat reduced in thickness, by slightly grinding down one surface, to enable the arrangement of their interior to be distinctly made out ; and this may be examined either by transmitted or by reflected light. Each zone thus seen in horizontal section (Plate V. fig. 1, c, c), consists of a circular set of small ovate cells, excavated, as it were, in the shelly sub- stance of the disk, and communicating with each other laterally by passages which unite them together into a continuous annulus. The zone which immediately sur- rounds the nucleus is connected with it by passages which extend from the outer margin of the large circumambient segment to the several cells of which it is itself composed ; and each zone communicates with the one on its exterior by similar passages, which usually extend, however, not from the cells of the inner zone to those of the outer, but from the connecting passages of the inner zone to the cells of the outer (Plate IV. figs. 8, 9) ; and thus it comes to pass, that the cells of each zone usually alternate with those of the zones that are internal and external to it. A ver- tical section of the disk, such as is shown in Plate V. figs. 4, 5, exhibits the same arrangement under a different aspect. The cells of the concentric zones are seen to be much higher than they are broad, so that they present a somewhat columnar form ; the proportion of their height to their breadth, however, may vary greatly in different parts of the same disk, the former often increasing from the centre towards the periphery (fig. 4), whilst the latter remains constant, or nearly so ; and the columns, instead of being straight, are generally more or less curved, and are some- times bent in the middle at an obtuse angle (Plate V. fig. 7, a, b). The gradation which presents itself from one of these forms to the other, and their coexistence even in the same specimens, clearly proves that no value can be attached to the form and proportions of the cells, thus seen in a vertical section, as furnishing specific charac- ters. In every perfect specimen, the columnar cells are seen to be closed at their two extremities by a thin shelly wall ; and this is sometimes flat, sometimes more or less convex *. The meaning of these arrangements is clearly seen, when we turn our attention to the structure of the animal (Plate IV. fig. 1). For the outer margin of * In a large proportion of the specimens obtained from sands or dredging, the cells have been laid open by attrition ; either throughout the surface of the disk, if it should be flat, or at its margin only, if it should be at all saucer-shaped. The constancy of this last character in a certain set of forms, resembling that repre- sented in Plate VII. figs. 8, 10, might at first sight lead to the idea that they constitute a distinct specific type ; but, as will hereafter appear, these plate-shaped disks cannot be separated by any definite line of demarcation from such as are quite plane ; and in specimens of them which have not suffered attrition, the marginal cells are closed, like all the rest. GENUS ORBlTOLITES: GENERAL PLAN; SIMPLE TYPE. 199 the circumambient segment of the nucleus is seen to give off a number of slender prolongations or 'stolons' of sarcode, which radiate from it to a short distance and then enlarge into columnar segments having a circular or ovoidal base, which are united with each other laterally by an annular 'stolon'; and from the portions of this ' stolon' which intervene between the segments of each annulus of sarcode, are given off the radiating 'stolons' that go to originate the next zone, the arrangement of whose parts is precisely similar to that just described. 18. In this manner, any number of concentric zones may be formed, which are exact repetitions of each other, except that the number of segments in the outer zones is greater than that of which the inner zones are composed. It does not increase, however, in the regular ratio of the respective diameters of the zones ; for the cells of the outer zones, being usually both larger and more widely separated from each other than are those of the inner, are less numerous in proportion ; thus in a specimen before me, there are twenty-eight cells in the innermost row and only forty-nine in the outermost, though the latter is more than twice the diameter of the former. The increase in the number of the segments is accomplished by the occa- sional interpolation of an additional segment, communicating directly with the one immediately interior to it, between the two segments which spring from the annular stolon on either side of the latter, as is shown in Plate IV. figs. 8, 9, a. Hence it is obvious that prolongations of sarcode giving origin to new segments, although ordi- narily put forth rather by the connecting 'stolons' than by the segments themselves, may originate from any part of the annulus. This is shown still more forcibly by the occasional occurrence of irregularities, such as that represented in Plate VI. fig. 6. 19. The cells of the last-formed zone communicate with the exterior by the very same kind of radial passages, as in other instances communicate with the next zone ; and the external orifices of these form the pores which present themselves at the margin of the disk (Plate V. fig. 1). Thus it is seen, on the one hand, how it hap- pens that these pores are intermediate between the cells, instead of opening directly into them ; and on the other, how each pore leads, by the divarication of its passage, into two cells, one on either side of it. When a new zone is formed, each pore opens into one of its cells ; and this zone in its turn communicates with the exterior, through a new set of pores at its own margin. Each pore is often surrounded by a rather prominent annulus of shell (Plate V. fig. \,d); and it is obvious that when the section passes through this, it will be indicated by a little ' beak' on either side of the entrance to the passage ; such ' beaks' (which are of course repeated through the entire disk) are shown in their ordinary aspect in Plate IV. fig. 9, but they are fre- quently more prominent, as is shown in Plate VI. fig. 1. 20. In all cases in which the growth of the disk takes place with normal regula- rity, a complete circular zone is added at once. Exceptions to this regularity are rare, and they can be generally traced with probability to some accidental interrup- tion. It can scarcely be doubted, I think, that when a new ring of cells is about to 2 D 2 200 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. be formed, the prolongations of sarcode issuing from the several pores of the pre- ceding zone coalesce, so as to form a complete zone of segments and connecting stolons around the margin of the previously-formed disk ; and that the deposit of calcareous matter forming the shelly walls of the cells and passages, takes place upon, or rather in, the superficial portion of this zone of sarcode. But I cannot find any evidence in the ordinary growth of the disk, that the sarcode extends itself over the surface of the portion previously formed ; although occasional appearances will be hereafter described (^[ 53), that seem to indicate that it may do so. 21. It is a fact of much importance in the due appreciation of the relations of Orbitolite and its allied forms to other tribes of Foraminifera, that the calcareous partition which separates each cell of any one zone from its neighbours on either side, is not double, but single. And this is in great part the case, even with regard to the partitions that separate the cells of successive zones ; the inner or central boundary of one being chiefly formed by the peripheral wall of the other. It is not easy even in thin sections to distinguish the boundary between the walls of one zone and those of another, so absolutely continuous do they appear to be. But it not unfrequently happens, that in fracturing these disks, their component zones come apart from each other, along their natural lines of junction, so as to disclose the real inner (or central) margin of the outer segment, which then presents a set of wide apertures, through which we look at once into its cells ; thus proving their incom- plete enclosure by proper walls on that side (Plate V. fig. I,//). Thus in the forma- tion of each new zone, the calcareous envelope seems to be only generated where the sarcode is not already in contact with a solid wall. 22. There cannot be any reasonable doubt, that the number of concentric zones which any disk may present, is entirely determined by its stage of growth, and that it affords no basis whatever for specific distinction. Just as in the case of the con- centric layers of wood in the stern of a tree, a minute nucleus, surrounded by only a single annulus of cells, may come in time to be the centre of a large disk consisting of many scores of concentric zones. Although, as already stated (^[ 17), most of the Orbitolites formed upon this simple type are of comparatively small size, yet there does not seem to be any definite limit to the multiplication of zones ; for I possess specimens attaining -15 of an inch in diameter, and consisting of about forty zones (much larger, therefore, than the younger zones of the complex type), in which there is no appearance of any departure from the original mode of growth. That com- paratively few specimens, however, attain so large a size upon this simple type of structure, is due, I believe, to the circumstance that they early tend to develope themselves upon the more complex plan which I shall presently describe. 23. Although I have spoken of these disks as typically plane or nearly so (there being usually no great difference between the thickness of their central and that of their peripheral parts), yet it not unfrequently happens that the successive zones gra- dually increase in thickness from within outwards (as is shown in Plate V. fig. 5), GENUS ORB1TOLITES : GENERAL PLAN; COMPLEX TYPE. 201 so that the height of the columnar segments progressively increases, and the entire disk becomes somewhat biconcave. Sometimes, again, without any alteration in the thickness of the several parts, the disk comes to assume, by the depression of its central portion, the shape of a plate, or that of a watch-glass, or (by the more com- plete upturning of its edges) that of a saucer. In any case in which either surface of the marginal zone is more exposed by its projection than those of the zones which it encloses, there will be a special liability to a laying-open of its cells (as shown in Plate VII. figs. 8, 10) if the disk should be subjected to attrition; and I believe that not only the recent species O. marginalis, but the fossil O. macropora, are nothing else than examples of this type, the figure of the latter given by GOLDFUSS* corre- sponding exactly with a form of it which I have frequently encountered. I have not met with any examples in this simple type, of that marginal thinning away as age increases, which is observable in many other Foraminifera. 24. From the simplest, it will be convenient to pass at once to the most complex type of structure presented by the Orbitolite, the existence of which is marked (as already noticed, ^[ 13) by a multiplication of the horizontal ranges of marginal pores. I have met with this form in specimens obtained by dredging, from the coast of Australia, from various parts of the Polynesian Archipelago, from the neighbourhood of the Philippine Islands, from the Red Sea, and from the ^Egean ; and as the sands of all these localities present the simpler type in great abundance, I am disposed to believe that the former is really not the less widely diffused than the latter, and would be discovered wherever it abounds, if properly searched for. The largest specimen in my possession, measuring seven-tenths of an inch in diameter, is from the coast of Australia, where these Orbitolites are so abundant at certain spots (as I learn from Mr. JUKES), that their entire disks and fragments, with fragments of Corallines (chiefly, I believe, the Corallina palmata of ELLIS), constitute the great mass of the dredgings. Among the Australian specimens, several attain a diameter of '45 inch, and a considerable proportion as much as '30 of an inch. Hence the Orbitolites of this type are among the largest forms of existing Foraminifera, being only surpassed, as far as I am aware, by the Cycloclypeus hereafter to be described. Of two speci- mens in my possession from the Feejee Islands, one measures '63 inch, and the other -53 inch in diameter ; but the average of the Polynesian specimens, like that of the Philippine, Red Sea, and yEgean, seems to be considerably lower than that of the Australian, as their diameter seldom exceeds '25 of an inch, and is usually not more than '10 or '12. 25. The disks formed on this plan, like the preceding, may be considered as typi- cally circular, although they are seldom or never exactly so in reality. They may be considered, too, as typically flat, with a slight concavity in the central part, from which, however, the nucleus often projects; but, as will hereafter appear, there is no constant relation either between the thickness and the diameter of different specimens, * Petrefacta, pi. 12. fig. 8. 202 t)R. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. or between the thickness of different parts of the same specimen, and the distance of these parts from its centre. The only remarkable departure from the ordinary form which I have met with, presents itself in certain Orbitolites from the Feejee Islands, of which several specimens in the Museum of the Royal College of Surgeons, and two in my own possession, exhibit a curious plication towards their margins ; the degree of this departure varies so much, however, in different individuals, the plica- tion being almost obsolete in some (see Plate V. figs. 2, 3), that it cannot be admit- ted to mark a specific diversity; and considering that these disks always grow on the surfaces of other bodies, it can scarcely, I think, be considered improbable, that the plications originate in the inflections of those surfaces*. These same specimens, moreover, also exhibit another curious abnormality ; namely, the projection of the upper and lower edges of the margin, so that a groove is left between them, the pro- jecting laminse being thin and foliaceous, and their cells very irregularly arranged. This peculiarity, again, being far from uniform in its degree, and being altogether wanting in specimens which in other respects precisely resemble those with plicated and foliated margins, must be considered merely in the light of an accidental variety ; but I cannot suggest any explanation of its occurrence, or of its limitation (so far as I am aware) to this particular locality. 26. The surface of the disk (Plate V. fig. 6) is marked out, as in the simpler type, by concentric zones of cells, the number of which bears a general (though not a constant) ratio to its diameter ; these cells are usually somewhat rectangular in shape and sometimes approach a square, but are more commonly nearly twice as long in the line of the radius of the disk as they are in the transverse direction, their long sides being parallel to each other. We shall hereafter see, however, that the form of the superficial cells is very subject to variation, and that it may be very dissimilar even in different zones of the same disk (^[ 48-52). The pores at the margin of the disk are disposed, as in the simpler type, between the projections formed by the convexities of the cells ; and each is usually surrounded by the pro- jecting annulus formerly noticed (^[ 19). The disposition of these pores, however, is far from regular, as they seldom form rows that seem exactly continuous with each other, either horizontally or vertically ; and the number of pores in each vertical row is by no means constant, even in different parts of the margin of the same disk-f". * I have elsewhere noticed the fact, that various species of Orbitoides are disposed to exhibit a like con- tortion ; and that the well-marked ephippial shape which some specimens present, is nothing else than an acci- dental variety (see Quart. Journ. of Geol. Soc., vol. vi. pp. 34, 35). t In following the description of the internal structure of this type of Orbitolite, it will be convenient for the reader to make frequent reference to the ideal representation which has been built up in Plate V. fig. 6, by the combination of materials furnished by a great number of preparations which are represented in separate figures accompanying the original Memoir in the Archives of the Royal Society ; these last of course furnish the real authority for every point in the description, the ideal figure, however, serving to display the relation of different parts to each other in a manner that no single preparation would possibly admit. GENUS ORBITOLITES -.GENERAL PLAN; COMPLEX TYPE. . 203 27. The disks of this complex type are not distinguished from those of the simple type already described, by any difference in the structure of the Nucleus ; and there is frequently nothing specially characteristic in the structure of the zones that imme- diately surround it. Each of the peripheral zones, however, consists of two super- ficial layers, an upper and a lower, and of an intermediate stratum ; these will now be described seriatim. 28. The superficial layers are formed of the (usually) oblong cells, whose contour is indicated by the surface-markings ; when they are laid open horizontally, by rub- bing away the thin shell which covers them in (Plate VI. fig. 3), it is seen that the floor of each cell has an aperture at either end ; but no communication can be traced, either through the side-walls between the contiguous cells of the same zone, or through the end-walls, between the cells of successive zones. Moreover, there is no such alternating arrangement of the cells of successive zones, as we have seen to prevail in the simpler type (^[ 17); and they altogether seem to be quite independent one of another. When this superficial layer is examined in a vertical section having a radial direction (Plate VI. fig. 7), it is seen that the floors of its cells (a, a) are formed by the expanded summits (d, d', d") of the irregular septa, which separate from each other the columnar cells of the intermediate stratum (cccc); and that the apertures at the two ends of the floor are the entrances to passages (e, e', e"), which lead obliquely downwards (the passages on either side of the partition between two successive cells of the superficial layer inclining towards each other) towards these cavities. It is observable, moreover, that just at the point at which the con- tiguous passages meet each other, there is always a round aperture (f,f,f) in the partition (g, g) which divides the contiguous cells of each zone ; and when, in a hori- zontal section, the superficial cells have been entirely ground away, so as to lay open the most superficial part of the intermediate stratum, this part is found to be traversed in each zone by a continuous circular canal (Plate VIII. fig. 3), with large rounded openings that lead into the columnar cells beneath. The meaning of this arrange- ment becomes obvious, when we examine the disposition of the animal substance which occupies these cavities ; for we find, as might have been anticipated, that the superficial cells are filled with segments of sarcode of corresponding shape (Plate IV. figs. 4, 7, aa) ; and that whilst these have no direct connexion with one another, each of them is connected by means of fleshy peduncles with the annular stolons bb that run along its extremities ; whilst from the under side of these annular stolons (fig. 4) descend the thick columns of sarcode (cc, c'c 1 ), which occupy the columnar cells of the intermediate stratum. The absence of any essential dependence of the segments of the superficial, and of those of the intermediate strata upon each other, seems indicated by the fact that there is no constant numerical relation between them, a circumstance which extremely perplexed me, until I had ascertained, by examination of the animal, that the passages (Plate VI. fig. 7, e, e 1 , e") debouch, not (as I had at first supposed) into the columnar cavities, but into the annular 204 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. which serves to bring the superficial and columnar segments of each zone into mutual communication. 29. As the description now given of the superficial layer applies equally to both surfaces, we may now proceed to the intermediate layer. When this is laid open by a horizontal section (Plate V. fig. 6), it is seen to consist of a series of concentric zones, the cells of which alternate with each other, like those of the simpler type (^[ 17-)- The cells are usually circular (or nearly so) in form ; but seem to differ consider- ably in size, even in different parts of the same zone. Their borders, however, very commonly present a funnel-like aspect; and thus we perceive that the diameter of the cavity is liable to vary, according to the part of it which the section happens to traverse, a fact which becomes more obvious when vertical sections are examined ; for it is then seen (Plate VI. figs. 7, 8, 9) that each columnar cell is narrowed by constrictions at intervals, so as to divide it imperfectly into a series of segments vertically superposed one upon another. The number of these segments varies according to the thickness of the disk ; so that it is anything but constant, either in different individuals, or in different parts of the same. Moreover, it may be often observed that the columns neither always pass from end to end in a straight line, nor maintain a complete isolation from each other (Plate VIII. figs. 1, 2) ; an inoscula- tion of two columns not being unfrequent, and more rarely a fusion of two columns into one. All these features of structure presented by the shell, are beautifully dis- played by the animal (Plate IV. fig. 4) ; the columns of sarcode (cc, cV) exhibiting the imperfect transverse segmental division, the not unfrequent inosculation, and the occasional fusion, which we have seen to exist in the cavities which they occupy. At their upper and lower extremities, they unite with the horizontal bands (bb 1 , bV), which pass continuously round, in each zone, between the intermediate and the super- ficial layers. 30. Save in the case of such accidental inosculations as those just noticed (which are indicated in vertical sections like that represented in Plate VI. fig. 7, by the irregularly disposed apertures h, h), no other lateral communication seems to exist between the contiguous cells of the same zone, than that which is established by the annular sto- lons just mentioned. The cells of the successive zones communicate with each other, however, as in the simple type previously described (^[ 17.) ; but with a curious modi- fication ; for whereas a horizontal section of the latter shows that each cell communi- cates with the two cells alternating with it in the interior zone (Plate V. fig. 1), a like section of the Orbitolite of complex type seerns to show that such a connexion exists with only one cell of the interior zone, by a passage running obliquely from one to the other, and extending continuously through several successive zones (Plate V. fig. 6, i, k). I was long perplexed by the want of constancy in the direction of these passages ; the very same section exhibiting opposite obliquities in contiguous parts (Plate VI. fig. 2). By the study of vertical sections, however, made tangentially instead of radially, so as to cross these connecting passages, I arrived at the explanation of this GENUS ORBITOL1TES: GENERAL PLAN; COMPLEX TYPE. 205 apparent anomaly, which is simply as follows. Each columnar cell really communi- cates with the two alternating columnar cells in the next interior zone; but by two distinct passages, instead of by the divarication of one ; and these passages are not upon the same plane, but those of different planes turn alternately to one side and to the other. This is well seen in the two tangential sections represented in Plate VI. figs. 8 & 9; of which 8 shows the back or central side of four contiguous columnar cells ad, bb', cc', dd', of the same zone, each of them perforated by a series of aper- tures, in which some degree of alternation is perceptible ; whilst 9 shows the front or peripheral side of four other columnar cells, in which it is seen that, by the sinuo- sity of the partition, the apertures of any vertical row, even when in a line with each other, open alternately into the cells on the right and on the left of the septum ; so that, (e.g.) the passages extending backwards from the row of apertures in the columnar cell bb 1 , fig. 8, will debouch alternately in cells aa' and bb', fig. 9, of the zone within. The same will of course be true of the pores which open on the margin, these being nothing else than the orifices of the inter-zonular passages just described, which, when another annulus is added, lead into its cells. This idea of the alterna- ting direction of the inter-zonular passages, seemed to furnish the solution of the appearances presented in Plate VI. fig. 2 ; for, as the disks are seldom perfectly flat, the section which traverses, at one part of the disk, the set of passages running in one direction, will traverse the other set of passages, where, by the flexure of the disk, the plane of section is slightly altered in regard to it. All doubt, however, as to the validity of this explanation, was removed by the examination of the animal substance filling the vertical columns ; for, as is shown in Plate IV. fig. 4, each column of sarcode in one zone (cc) does communicate with the two columns alternating with it in the next zone (cV) by two rows of peduncular stolons ; and the peduncles which pass from each pair of contiguous columns, to the single column of the next zone, incline towards one another, so as to enter it nearly in the same vertical line, though in dif- ferent horizontal planes. 31. That which has been already stated in regard to the partial deficiency of the inner wall in each of the concentric zones of the simple type (^[21.), holds good also in regard to the septa which divide the successive zones of the intermediate stratum in this more complex type; for the walls of the columnar cells close-in around them very imperfectly on their inner or central side, leaving large irregular vertical fissures (Plate VIII. fig. 1) which are applied to the vertical rows of orifices (Plate VIII. fig. 2) on the outer margin of the included zone. 32. The thickness of this intermediate stratum, and the number of vertical segments of which it consists, are found to vary considerably in different parts of the same disk ; being usually least near the nucleus, and gradually augmenting in successive zones as their distance from the centre increases (Plate VI. fig. 7) ; or ceasing to augment at a certain point, so that the outer part of the disk is flat; or even dimi- nishing again, so that the disk thins away towards its margin. It is specially worthy MDCCCLVI. 2 E 206 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. of note, that whatever differences of this -kind may exist, they are entirely due to the variable length of the columns of the intermediate stratum ; the depth of the cells of the superficial layers being- nearly constant, and no vertical multiplication of these ever taking place. The intermediate stratum, where it abuts on the nucleus, is often very thin ; the annular stolons that run beneath the superficial layers being in such near proximity to each other, that the intervening column of sarcode is very short, and consists of only a single vertical segment, Plate V. fig. 12. 33. The foregoing description applies in every particular to those specimens only, which present the structure of this type of Orbitolite in its most regular and charac- teristic development ; and the differences between this more complex form, and the simple form previously described, are such as at first sight to preclude the idea of their specific identity. Hence I am not in the least degree surprised, that Professor WILLIAMSON, by whom their respective plans of organization were first compared (loc. cit.), should have unhesitatingly regarded these two forms as specifically distinct. But when a large number of specimens of the more complex type are carefully examined and compared with each other, it becomes obvious that a vast amount of diversity in the arrangement of the cells of the shell, and of the segments of the animal, may present itself; and that one after another of the characters which at first seem most clearly marked and therefore most distinctive, may be shaded off (so to speak) in such a manner that a complete transition is established, sometimes even in a single disk, between the simple and complex types. Such a transition is exhibited by the specimen of which a vertical section is figured in Plate V. fig. 7; for it is obvious that the central portion of this disk (a b) is so exactly conformable to the simpler type, that if this growth had stopped at the twenty-third zone, it would have undoubtedly been regarded as an unusually large example of that form. So many variations present themselves in the development of the different parts of the more complex type, that it will be desirable to describe them under a distinct head (Sect. IV.) ; and when these shall have been duly considered, I think that all doubt as to the specific identity of the simpler and more complex forms will be done away. III. Physiology. 34. Growth. Of the mode in which the Nutritive process is carried on in Orbito- lites, our imperfect acquaintance with their living habits leaves us much in the dark ; nevertheless it is fair to reason by analogy from a comparison of their structure with that of other Foraminif'era whose habits of life are known ; especially as this analogy is sufficiently complete in the present instance, to justify a tolerably firm reliance upon it, and as the results to which it would lead are in harmony with the facts of observation. All the Orbitolites, whether of the simple or complex type, which have been collected in the living state, have been found growing on the surface of Sea- weeds or other marine plants (as Zostera), or of Zoophytes ; it may therefore be fairly presumed, that such is their ordinary habitat ; and hence it is scarcely conceivable GENUS ORBITOL1TES: PHYSIOLOGY; GROWTH. 207 that their attached surface should ever be invested by sarcode. Moreover, several of the spirit-specimens which I have submitted to decalcification, have proved to be so closely invested by a covering of vegetation, chiefly composed of Diatomacece, Desmidiece, and other minute Algae, that I cannot suppose even the free surface of their disk to be ordinarily covered by sarcode*. The analogy of other Rhizopods, however, would lead us to suppose that the sarcode projects from the marginal pores under the form of pseu dopodia, and that it is by the introduction of alimentary parti- cles (chiefly minute forms of vegetation) through their means into the mass of sarcode from which they are put forth, that the fleshy body pervading the entire disk is nourished. For although there is nothing like a digestive cavity in any part of it, or an alimentary tube passing from one portion to another, still less any vascular commu- nication between the segments, yet as the sarcode forms one soft homogeneous mass continuous throughout, the body as a whole will receive the benefit of any incorpora- tion of new matter with its substance, in whatever situation this may be made. That organic particles small enough to pass through the marginal pores, are thus introduced into the chambers of the disk, is proved by the curious fact, that the residuum left after the decalcification of large and therefore aged disks, whose animal contents have not been preserved, consists almost entirely of an assemblage of remains of minute Diatomacece, Desmidlece, &c., which have obviously been retained in the interior of their cavities, after the assimilation of the nutriment they were competent to afford. 35. The sarcode-body of the animal, growing at the expense of the nutriment thus appropriated, will gradually, it is probable, project itself through the marginal orifices, not merely in filamentous pseudopodia, but in quantity sufficient to form new segments on the outside of each pore ; and these segments, extending themselves laterally, will come into mutual connexion, and will thus form a complete annulus. It may be presumed to be by the calcification of the surface of this beaded ring of sarcode, that the formation of the shelly zone is accomplished ; and if the calcifying process commence on the segments, and extend from these along the surface of their connecting stolons, we can understand why the passages that are left for communi- cation with the exterior, should arise from the intermediate divisions of the annular canal, instead of from the segments themselves. 36. The addition of new zones usually takes place with the same regularity in the complex as in the simple type of structure ; but departures from this regularity, occasioned by a want of completeness of particular zones, are more frequent; and this is perhaps to be accounted for by the larger size of the disk, which will tend to produce a less intimate dependence of each part of the animal body upon every * I have found such an investment also on several dried specimens ; and until I had detached and examined this, I should have supposed from its aspect that it was the desiccated flesh of the animal. I have little doubt that the "greenish cuticle" described by Mr. CARTER as covering his Operculina arabica (Ann. of Nat. Hist, ser. 2. vol. x. pp. 168, 172) and supposed by MM. D'ARCHIAC and HAIME (op. cit. p. 52) to be specially con- cerned in the formation of the shell, is of the same nature. 2 E 2 208 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. other, and will thus favour the partial action of any cause (e. g. an excess of nutrient materials) which promotes a more rapid growth on one side than on the other. And this view is most remarkably borne out by the fact, which I shall more fully illus- trate in a subsequent memoir, that in another example of this group*, which, though normally growing upon the cyclical type, possesses a greater degree of segmental independence, such irregularities occur far more frequently ; so that, in fact, it is rare to meet with a disk whose increase has taken place with uniformity throughout. 37. Reparation of Injuries. Looking at that vegetative repetition of parts which pre-eminently characterizes the body of the Orbitolite, every one of the segments first budded-off from the nucleus, and subsequently from the margin of the pre-formed zones, being the precise repetition of every other, it may be expected from the analogy of similar organisms, that every one of these parts should be equally capable, both of repairing injuries done to itself, and of maintaining an independent existence when detached from the mass to which it originally belonged. And although no opportunity has yet presented itself, of subjecting such a conclusion to the test of experiments devised for the purpose, yet accident has furnished the means of verifying it, to a degree that could scarcely have been anticipated. For in the course of my examination of the large collections which have been placed at my disposal, I have met with several specimens, in which it is evident that, after larger or smaller por- tions of the disk had been broken away, a new growth has taken place along the fractured edge. Various examples of this are shown in Plate VIII. In the first that I happened to meet with, which is represented in fig. 6, the injury is evidently very slight, being confined to the loss of a few rows near the edge of the disk, for some- thing less than half its circumference (a b). This injury had obviously been sustained previously to the formation of the last two zones ; for these, whilst added to the uninjured part of the margin in the usual way, have followed the irregular contour of the broken edge; and whilst in the former case the cells present their normal con- formity to those of the margin they invest, in the latter, the cells, while obviously continuous with the preceding, are quite unconformable to those of the fractured margin, as is shown on a larger scale in fig. 7- Hence it seems to me probable, that the growth of these two rows along the fractured edge, has taken place, not from that edge itself, but by an extension of the sarcode about to form the new circle of the entire edge, from the points a and b. In fig. 9 is seen an example of a similar kind, in which a much larger portion of the disk has been broken away, so as to leave only an irregular fragment, including its centre and about an eighth of its margin. Here seven rows of cells have been formed since the injury ; and these, whilst produced conformably to those of the uninjured margin, present the most marked want of con- formity to those of the fractured margin, which, nevertheless, they completely surround. A careful examination of this specimen, indeed, seems to me to leave little room for doubt, that the growth of the innermost, or what 1 may call the reparative zone of * I refer to a genus hitherto undescribed, which I shall designate Cycloclypeus. GENUS ORBITOLJTES: PHYSIOLOGY; REPARATION OF INJURIES. 209 cells, took place, not from the broken edge, but from .the margin of the unbroken ; just as, to use a professional simile, an ulcerated surface 'skins-over' by an extension of the integument from its edges, not by the direct formation of skin upon the granu- lation-surface itself. All the six rows subsequently produced, are conformable to each other, and to the first or reparative row, from which they have obviously extended themselves after the normal manner. It is observable, however, that the breadth of these rows varies in different parts, being least where they invest the pro- jecting portions of the fractured edge, and greatest where they sink into its hollows. And thus it comes to pass, that the irregularities left in the shape of the disk, by the loss of a large part of its substance, are gradually compensated, so as to restore it to a form much more nearly corresponding to its typical symmetry. 38. Even a very small fragment appears thus to serve as the nucleus for a new disk. In fig. 8 is shown an example of this kind, in which the tendency to the repro- duction of the typical form, by the compensative reparation just "described, is very curiously marked. This specimen also presents the following very curious feature, that the new growth has taken place from the inner margin of the original fragment (aa), and not from its outer or growing margin, as in the cases previously noticed. Having carefully examined it in various modes, I cannot entertain the slightest doubt that such has been the case ; for the cells of the first new zone, as well as those of all the zones subsequently produced, are so manifestly conformable to those of the thinner and older portion of the fragment, and are so unconformable to those of the thicker and newer margin, that it seems obvious that the sarcode must have extended itself from the former part, along the fractured edge on each side, and have then enveloped the margin which had been left entire. This may have more readily taken place in the present instance, because at the part (aa) the fracture seems to have followed the course of one of the zones, instead of passing, as at the sides of this fragment, and in the instances previously cited, in such a direction as to cut the zones transversely. 39. The preceding instance clearly proves, that connexion with the central nucleus is not in the least degree requisite for the continued growth of the peripheral parts ; since these may be entirely detached from it, without any loss of vital activity. The same inference may be deduced from the examination of specimens, in which, the central portion of the disk having been broken-out, a growth of new zones seems to have taken place from without inwards, so as to fill up the void space thus left. In no other way can I account for the appearances presented by a specimen in my possession, in which the included portion is as evidently unconformable to that which surrounds it, as it is in the preceding case, but in which there is also an unfilled void, the shape of one part of which clearly indicates that it occupies the site of the original centre. The included portion, and not the peripheral, must therefore be the after-growth in this instance ; and if a little more time had elapsed, the whole of the central vacuity would probably have been filled up by it. 210 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. 40. In the specimen represented in fig. 5, the central portion appears to have been lost, with about a third of the peripheral ; and the new growth seems to have taken place at the same time, from the inner margin aaa of the fragment, and from its outer margin b b b, the two growths becoming continuous with each other along the broken edges ab, ab. For although the zones that lie internally to a a a are conform- able to those which surround them, yet there is a peculiar character about them (more apparent in the specimen than in the drawing) which indicates them to have been formed at a later period, and to have been contemporaneous with those which surround the zone b b b. Their actual continuity at the angles a a is unfortunately interrupted by an injury which the specimen seems subsequently to have received; yet its traces are sufficiently perceptible on one side, to justify the belief in its former existence. The specimen of which fig. 4 is a delineation, seems to have been the subject of several minor fractures and reparations ; but the course of its zones marks out an obvious separation between an earlier- and a later-formed portion, one having sprung from the other along the line ab. The incompleteness of the speci- men, however, prevents me from coming to any certain conclusion, whether the small inner portion is here the older, the large outer portion having grown in the first instance from its margin a b, and having gradually extended itself around it; or whether the outer portion is the residue of an unusually excentric disk, which, having lost its nucleus and the zones immediately surrounding it, has filled up the central space with an extension from its innermost zone, which is consequently the newest portion of the whole. It is interesting to find evidence in fossil specimens, that the same kind of reparation has taken place. Among the Orbitolites which I have examined from the Calcaire grossier of Paris, is a disk of which a large part had obviously been lost by fracture, but of which the original symmetry had been in great degree restored by a similar outgrowth from the zones formed from the uninjured margin, along the fractured edge. 41. This series of abnormal phenomena, then, not only confirms the conclusion that seemed fairly deducible from our previous examination of the normal mode of growth, with regard to the independent endowments of the component segments of the Orbitolite body, but also affords some additional information of much interest. For we see, in the first place, that the growth of the sarcode, and the addition of new parts, may take place in the direction of the centre, where a free edge is exposed at the inner margin of any zone, as well as in the peripheral direction from the normal outer margin. Secondly, the reparative nisus seems always to tend towards the pro- duction of a disk, whose shape shall approach the circular, whatever may be the form of the fragment which serves as its foundation; thus showing that, notwithstanding the repetition and independence of the separate parts of these organisms, each cluster, whether large or small, is an integer, having an archetypal symmetry to which it tends to conform, thus strongly reminding us of the laws of crystallization. And thirdly, the plan by which this recurrence to the discoidal form is provided for, GENUS ORB1TOLITES: PHYSIOLOGY; QUESTION OF INDIVIDUALITY. 211 seems partly to consist in the limitation of the new growth to the natural margins of the zones ; no such growth taking place from the edge of a fracture which has crossed the zones transversely, although it may proceed from the remains of a zone which has been broken off by a fracture that partly follows its course. 42. Question of Individuality. It has been frequently discussed, whether each of the composite forms of Forarninifera, such as Orbitolite or Nutnmulite, is to be regarded as a single individual, or as a colony or clutter of individuals. All occasion for this discussion would, I think, be removed by the adoption of philosophical views as to what really constitutes an individual, and as to the relationship between the parts which, having a common origin in one generative act, are multiplied by a process of gemmation. As I have elsewhere endeavoured to show*, the entire product of every generative act, whether developing itself into a body of high organization, distin- guished by the structural differentiation of its parts, or evolving itself as an almost homogeneous aggregate of equal and similar segments, must be regarded as hornolo- gically the same ; and the essential difference between the two, as living beings, lies in the functional relations of their respective parts. For whilst in the former there is so close an interdependence amongst them all, that no one can exist without the rest, and the life of the whole is (as it were) the product of the lives of the component parts, there may be in the latter such a mutual independence, that each part can con- tinue to live, grow, and reproduce itself when separated from the rest, so that the life of the whole is (so to speak) but the sum of that of its components. Now the term 'individual,' being commonly applied to the entire organism in the first case, and to only a small segment of it, perhaps, in the second, is obviously inappropriate either to one or to the other, except in so far as it expresses the fact of independent existence. But the limits of such individuality as this cannot be strictly denned, and they even differ widely in animals whose general plan of structure is the same-f~. Hence in regard to the Foraminifera, as in regard to Zoophytes, Composite Acalephae, &c., we are to regard the entire mass originating in a generative act, as a single organism; and the question in regard to the functional independence of its multiple segments, is one of degree in each particular type. Thus, as we have seen, this independence exists in the case of the Orbitolite to such a degree as to make each part entirely self-sustaining, and to prevent the existence of any definite limit to the growth of the whole ; yet it is quite possible that in a form so much more elevated as Nummulite, there may be, as maintained by MM. D'ARCHIAC and HAIME (op. cit. p. 69), such a degree of mutual dependence among the segments, and of unity in their aggregate life, that the latter predominates sufficiently to limit the growth of the organism to a tolerably determinate size|. * Principles of Comparative Physiology, chap. xi. sect. 1. t See also Mr. HUXLEY'S observations on this subject, in Philosophical Transactions, 1851, pp. 578, 580. I Whilst admitting the possibility of this view, I shall hereafter have occasion to question its correctness ; since the evidence on which it is based appears to me by no means satisfactory. In fact, when I come to 212 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. 43. Reproduction. The mode of Reproduction of the Foraminifera generally, is at present involved in the deepest obscurity ; and there is little probability that it will be fully elucidated by any other means, than continued observation of the animals in their living state, such as may probably be best carried out with regard to the species of our own seas by keeping them in Vivaria. In default of such observations, and as a guide to further inquiry, I think it as well to state what has fallen under my own notice. In many parts of the body of spirit-specimens of Orbitolite, especially, but not solely, in the superficial cells, I have found the sarcode broken-up as it were into little spherules, as represented in Plate IV. fig. 3 ; these spherules, however, do not seem to possess any peculiar investment, nor does their sarcode appear to have undergone any special change. Similar spherules are figured by Professor EHRENBERG (op. cit.) in several of the cells of his Sorites orbiculus ; and Professor SCHULTZE has recently (op. cit. pp. 26, 27) described bodies which seem to be of the same kind, though more opake (probably through having a denser envelope), as frequently pre- senting themselves in certain chambers of Rotalice, or even throughout the entire series. I feel much inclined to believe that these bodies are gemmules, which, like the zoospores of the Algae, are produced by a resolution of certain portions of the substance of the organism into independent particles, which, spontaneously detaching themselves, and escaping through the marginal pores of the disk, will go forth to lay the foundation for new disks elsewhere. Besides these, however, I have more rarely met with certain other bodies, appa- rently imbedded in the sarcode, which may be either gemmules in a later stage, or may possibly be true ova ; these, represented in Plate IV. fig. 11, seem to exhibit various stages of binary subdivision ; and they present a deep-red colour, even in spirit- specimens. I can scarcely imagine that these can be vegetable organisms that have been introduced through the marginal pores ; since they are much too large to pass through these, without a great alteration in form ; and this would seem to be incompatible with the firmness of their envelope. At g, fig. 1 1, is represented, under the same magnifying power with the foregoing, an object which I have detected in one of my vertical sections of the shell, where it occupies one of the superficial cells, the cover of which is deficient. Now it is quite possible that this cell may have been accidentally abraded, and that the object in question may have found its way into it ab externo ; its position and aspect, however, seem to me much more conformable to the idea, that it has been developed in the disk itself, and that it has burst through the lid of the cell by its own enlargement, in preparation for its final escape. And this view seems borne out by the fact, that I have frequently found a few cells open on different parts of the surface of disks which did not appear to have suffered any abrasion ; as if the rupture of their lids had taken place as an ordinary describe (in a future memoir) the structure and varieties of Nonionina, the nearest existing type to Nummulite, and in my belief generically identical with it, I shall have occasion to show, that there is not only no proof of the existence of such a limitation to its growth, but that there is strong evidence to the contrary. GENUS ORBITOLITES: REPRODUCTION; VARIATIONS IN SIZE. 213 phenomenon of their development, instead of being the result of accident. I do not wish to attach any weight to the interpretations I have here offered ; but I simply state the facts, and the explanations of them which have suggested themselves to my own mind ; merely adding, what I hope to present in more detail at a future opportunity, that bodies resembling the first or primordial cell, in which Foraminifera of all forms originate, are not unfrequently met-with in the chambers of many other species. IV. Variations. 44. Variations in Size. We have already seen that diversities both in the diameter and in the thickness of the disk, arise directly from the degree in which the animal substance (whereon the skeleton is modelled) has extended itself either horizontally or vertically, so as to multiply either the number of concentric rings, or the number of the superposed segments of which each ring consists. This, however, is not the only source of variation in size ; for a most extraordinary diversity presents itself in the dimensions of the individual components, by whose repetition the entire disk is made up. It is in the nucleus that I find this diversity most strongly marked, as will appear from a comparison of Plate VII. figs. 1 4, which exhibit parts of a grada- tional series of twelve, from the smallest to the largest forms I have met with, all of them accurately drawn, under the same magnifying power, from specimens in my possession*. The length of the entire nucleus of fig. 4 is about seven times that of the nucleus of fig. 1, and its breadth aboutybwr times as great ; the area of the former is therefore about twenty-eight times that of the latter ; and as it is also several times as thick, the whole of the cavity, which was occupied in the living state by animal substance, could scarcely have been less than a hundred times as large in the one as in the other. (Compare also figs. 5, 6, 10, 12, 13 of Plate IV.) There is not by any means the same amount of difference between the dimensions of the ordinary cells which are formed by concentric extensions of the nucleus ; nevertheless, it will be seen by a glance at the figures just referred-to, that these also exhibit marked diversities in size, the largest cells being usually found to spring from the largest nuclei, and vice versa. Moreover, the individual cells of the very same disk are occasionally found to differ amongst each other, as widely as do the cells of fig. 1 from those of fig. 4. 45. Similar differences present themselves in the thickness of individual cells ; as is of course best seen in the simple type of Orbitolite, in which the augmentation of thickness is not produced by the vertical superposition of multiple segments. A remarkable example of this kind is presented in the comparison of figs. 4 and 5 of Plate V. ; these being, like the figures in Plate VII., drawn under the same magni- fying power. I possess a series of vertical sections of different individuals, in which the same gradual transition is seen from the thin to the thick, as I have just stated * The entire series of figures is in the possession of the Royal Society. MDCCCLVI. 2 F 214 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. V to exist in regard to superficial area ; and which also proves that the relative thick- ness of the central and of the peripheral portions is equally liable to variation. 46. It seems obvious, from the foregoing considerations, that neither the absolute nor the relative dimensions of the individual parts of these composite fabrics, can, anymore than the dimensions of the entire disks, be taken as affording valid charac- ters for the discrimination of species ; and that such a wide range of variation exists among individuals, as would, if the extreme cases alone were known, seem fully to justify their separation under distinct specific designations. Thus, if the two extreme forms, figs. 1 and 4 (Plate VII.), had been the sole objects of comparison, most naturalists would undoubtedly have considered the strongly-marked difference in the size of their respective nuclei to entitle them to rank as separate species; and even if fig. 2 or fig. 3 had been brought into comparison with them, it might have been a question whether it should be associated with fig. 1 or with fig. 4, or should rank as a third species intermediate between them, or should be considered as a con- necting link specifically identical with both, and therefore establishing their specific identity with each other. The comparison of the entire series must be felt to remove all ground for hesitation on this point, since it is manifestly impossible to draw a line across any part of it, which should divide it into two or any larger number of groups, respectively characterized by constant and well-marked differences in size. And here again, therefore, we have evidence of the great importance of bringing into comparison a sufficiently large number of forms, to enable us to determine in some degree the measure of individual variation. 47. Variations in Shape. The very strong tendency which we have seen to pre- vail in the Orbitolite, not only to the maintenance of the circular type in the regular growth of the disk, but to its reproduction after accidental injuries, seems to prevent the occurrence of any considerable variation in its general form, except such as may be directly produced by external agencies. The circular sometimes gives place to an elliptical shape (Plate VII. fig. 4), especially in young specimens, whose form is more determined by that of the nucleus than is that of older individuals. And the occur- rence of such a variety in the recent type, makes me indisposed to admit that ellip- ticity of shape can be in itself a sufficient basis for the specific differentiation of any fossil form of this genus. Of the marked differences in the general aspect of the disks, which may arise from differences in the relative proportions between their thickness and their diameter, and in the relative thickness of their central and peripheral portions, mention has been already made (^[^j 17 and 32.). And the only other important departure from the typical shape which I have met with, either in the simple or in the more complex form, has been described under a former head (f 25.). 48. Variations in the Form of the Superficial Cells, and in the Markings of the Sur- face. The appearances presented on minute observation by the surface of the Orbi- tolite, are so far from being uniform, that to any one whose eye had not become GENUS ORBITOLITES: VARIATIONS IN SURFACE-MARKINGS. 215 familiarised with their variety by the examination of a considerable number of speci- mens, they would become sources of great perplexity. We have already seen that the cellular markings present two very distinct forms, the rounded (Plate V. fig. 1) and the oblong (Plate V. fig. 6); the first of these being specially characteristic of that simpler type of structure in which there is only a single layer of cells, but not being confined to it ; whilst the second is peculiar to the complex type, in which there are two superficial layers, distinct from the intermediate stratum. Now the occasional coexistence of both these plans of structure in a single individual (^[ 33.), sufficiently proves that the diversity of the surface-markings to which they respect- ively give rise, cannot be regarded as a basis for specific distinction ; and when these extremes of diversity are kept in view, it must be felt to be highly improbable that any modifications of either form should possess greater importance. That such modifications are mere individual varieties, is further evidenced by their gradational character, and by the fact that two or more of them may present themselves in the same disk. In my description of them, I shall limit myself to an account of those more remarkable and frequently-recurring varieties, which will serve, I think, as a key to any others that are likely to be met with. 49. Although each surface, in either of the two principal types, ordinarily shows a division into concentric zones, which are again transversely subdivided so as to mark the separation of the cells, yet sometimes the concentric zones are alone visible, and no transverse subdivision is indicated, save by the alternation of lights and shadows proceeding from a like alternation of solid substance and of hollow spaces beneath (Plate VII. fig. 7)- This predominance of the concentric divisions, which gives a very distinctive aspect to the disks which exhibit it, is usually most apparent in indi- viduals whose vertical section exhibits two planes of cells ; and it has seemed to rne to depend on the unusual freedom between the lateral communications, which I have noticed in certain individuals thus formed, so that the animal portion of each zone might be described as an annulus of sarcode, merely constricted at intervals. This peculiarly cyclical aspect of the surface (on whose occurrence in fossil specimens I believe the genus CycloUna to have been founded, ^[ 5.) may pass into either of the principal types previously noticed; thus in fig. 14 we observe the concentric zones, though still very strongly marked, breaking up (so to speak) into bands of rounded cells with slightly convex covers ; whilst in figs. 5, 6 they are subdivided by very defi- nite transverse lines into cells of remarkable squareness, which still retain the original flatness of their surfaces. 50. On the other hand, the appearance of concentric division is sometimes almost entirely wanting ; the surface of the disk exhibiting excentric circular markings, which resemble those of an engine-turned watch-case (Plate VII. fig. 8), and the boundaries of the cells being formed by the intersection of these with each other. This aspect, however, which seems due to an unusual freedom in the oblique com- munications between the cells in each zone and those alternating with them in the 2 F2 216 DR. CARPENTER'S RESEARCHES ON THE FORAMIMFERA. contiguous zones on either side, insensibly passes into the ordinary type ; and it is not uncommon- to meet with disks, especially fossil, which exhibit in one part the engine-turned aspect (Plate VII. fig. 8), and in another (fig. 14) that of concentric zones transversely subdivided. Indeed I have sometimes found that the very same disk might be made to present either of these aspects, according to the manner in which the light is made to impinge upon it and is reflected from it. 51. Although the rounded or ovoidal form of the superficial divisions is specially characteristic of the simple type of Orbitolites, yet it is by no means restricted to this; being frequently met-with in the thicker disks of the more complex type, and being almost constant in the fossil forms that abound in the early Tertiaries. Its occurrence, however, may always (I believe) be considered as indicating an incom- plete separation between the superficial cells and the columnar cells of the inter- mediate stratum (^[ 58.); so that the former present the shape of the latter, in place of the form which properly characterises them. The shape of the cell is sometimes marked out in unusual strength by the convexity of its lid or cover, as shown in Plate VII. fig. 15 ; and this feature is often so pronounced in the large fossil Orbito- lites of the Paris basin, as to become visible to the naked eye. A very marked diversity in its degree, however, as well as in the size of the cells, is to be noticed in the contiguous zones of another specimen (Plate VII. fig. 1.6) ; whence it is obvious that the convexity is a mere accidental variation, and is a character of no value whatever as regards the differentiation of species. The relation of the rounded to the square or oblong cells is made evident by the occurrence of intermediate links of transition. Thus, from such circular cells as are delineated in Plate VII. fig. 9, the passage is easy, through those shown in figs. 8 and 14, to those of fig. 6 (which are drawn under about twice the magnifying power), and thence to the square cells of the inner part of the portion of the disk figured in Plate VII. fig. 13. This last figure illustrates the important fact, that while the cells in one part of the surface of the disk are square, others in close proximity with them may be oblong; thus conducting us to the extreme form of this type, represented in Plate VII. fig. 12. 52. The foregoing considerations seem to render it obvious, that the diversities in the form of the superficial cells do not afford any ground whatever for the establish- ment of a corresponding multiplicity of specific types, but that they must rank as individual variations to which there is scarcely any definite limit. If the originals of Plate VII. figs. 5, 7, 9, 12 and 15, had happened to have presented themselves to the Systematist without any of the connecting forms, he might have been pardoned for describing them as distinct species characterised by well-marked differences in the form and arrangement of their cells ; but no such differentiation can be admitted in the face of the fact, that these are only extreme examples of variations, which show themselves in a minor degree between almost every two specimens brought into com- parison, and even between the different parts of the same disk. Moreover, when it is borne in mind, that the animal basis on which the calcareous skeleton is moulded GENUS OtlBlTOLITES: VARIATIONS IN SURFACE-MARKINGS. 217 is not a body of constant shape, provided with organs having 1 a fixed relation one to another, but is a mass of almost homogeneous sarcode, which in the living state is continually undergoing changes of form, one part extending itself into pseudopodia, whilst another undergoes a corresponding contraction, a strong a priori improba- bility is seen to exist, that, in animals of such organization, the form of the compo- nent segments should possess that value as a specific character, which it can only derive from constancy. 53. Besides those regular markings of the surface, which correspond to the divi- sion of the interior into cells, a peculiar aspect is frequently given to it by the deposit of calcareous thickenings, which are sometimes irregular, but which occa- sionally present an approach to symmetry. The most remarkable example I have met with, of this kind of addition, is delineated in Plate VII. fig. 11, in which it will be seen that the deposit has taken place in radial lines disposed with a certain degree of regularity. But in fig. 10, which represents a specimen whose surface is far less altered by these deposits, no such symmetry presents itself; and other specimens in my possession exhibit the means between these extremes. Hence we are justified in pronouncing this peculiarity to result from an accidental outgrowth, which is so variable in its degree as not to afford the least basis for specific differentiation. It is worthy of note, however, that it presents itself far more frequently, and also in a far more characteristic manner, in the Orbitolites of the Philippine Seas, than in those of the Australian or of any other provinces ; and this circumstance seems to render it probable, that the outgrowth is directly due to the influence of some exter- nal conditions, probably to an excess in the proportion of carbonate of lime in the waters inhabited by these particular specimens. 54. Variations in Mode of Grotvth. Although the cyclical mode of growth, when once established, is subsequently maintained with great regularity, and although in what may be considered the typical form, it commences from the 'nucleus' itself, yet there are numerous instances in which the typical regularity is more or less widely depart ed-from, so that the early increase seems to take place after an alto- gether different plan. The most marked antithesis to that regularly concentric type of growth, in which a complete annulus of cells is formed around the large circum- ambient segment of the ' nucleus ' (see ^[ 17, also Plate IV. fig. 5, and Plate VII. fig. 2), is presented by those forms in which this circumambient mass only gives origin to new cells at its extremity ; these in their turn bud-forth others, which extend and multiply themselves laterally as well as in advance ; and thus a kind of spiral is pro- duced, which opens-out very rapidly, the lateral portions of its mouth tending to grow-round and embrace the nucleus. Thus, starting from the central globular mass, 1, of Plate IX. fig. 4, we see that the circumambient mass 2 2, which nearly surrounds it, gives origin at one of its extremities to a smaller mass, 3, from which bud-off two cells, 4 4, which again give origin to four cells, constituting the row 5 5. The cells of the next row, 6 6, are more numerous, but are themselves exceeded by 218 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. those of the zone 7 7, which not only surrounds it, but extends further backwards upon the nucleus. So the cells of the zones 8 8, 9 9, 10 10, 1111, and 12 12, pro- gressively increase in number, and each zone extends itself further back upon the nucleus, until those of the last of these zones nearly meet upon its yet unenclosed margin. The cells of zone 13 13 do actually meet there, so as to form a complete circle; and zones 14 14 and 15 15 are formed with cyclical regularity, as would be any other zones subsequently produced. A still more complete spiral, in which twenty-two zones (commencing with the central segment) succeed one another before the first complete annul us is formed, is shown in Plate IX. fig. 2. Both figs. 2 and 4 (the latter of which is diagrammatized) represent the central portions of large disks, whose peripheral portions grow on the regular cyclical plan. 55. Now if these two plans of growth the one cyclical from the beginning, the other cyclical only after having been at first spiral were constantly presented in well- marked contrast with each other, there would be good ground for considering them (as Professor WILLIAMSON has done*) to be characteristic of distinct specific types. But this idea cannot be entertained, when a large number of individuals are examined. For it then becomes apparent, that the number of cases in which the nucleus is surrounded on all sides by the same number of zones, indicating that the concentric mode of growth has prevailed from the very first, are very few ; but that in by far the larger proportion of specimens, there is a slight excentricity of the nucleus, with a larger number of zones on one side than on the other, as in Plate IX. figs. 1,3; indicating that the first-formed zones have been incomplete circles, owing to a restriction of the gemmation of the nucleus to one part of its periphery. This is shown extremely well by decalcified specimens of the animal, no two of which, in fact, precisely resemble one another as to the mode in which the first zone of segments originates in the nucleus. Thus in the specimen represented in Plate IV. fig. 1, of which the nucleus is represented on a larger scale in fig. 12, the circumambient segment of the nucleus gives off only three stolons, at the end most remote from its connexion with the central mass ; and the first zone of segments is far from being entire, the cyclical type not being completely attained until two or three successive additions have been made. In fig. 13, eight stolons are given off from the nucleus; and from the half-zone which they form, an entire circle is next produced; thus affording a remarkable confirmation to the idea I have already suggested (^[ 37.), as to the capacity of a portion of a zone to give origin to a complete annulus, by the lateral extension of its bands of sarcode. In fig. 10, the nucleus gives off eleven stolons on one side, and there are indications of three or four on the other. In fig. 6, the stolons come forth from a still larger proportion of the periphery of the nucleus ; the zone which first surrounds it, however, is still incomplete in some parts, though the succeeding zone forms an entire circle. Finally, in the specimen represented in fig. 5, which is almost the exact counterpart of the disk represented in Plate VII. fig. 2, * Transactions of the Microscopical Society, 1st series, vol. iii. pp. 116, 119. GENUS ORBITOLITES: VARIATIONS IN MODE OF GROWTH. 219 the stolons pullulate from the entire circumference of the nucleus, and the annular zones of segments are complete from the first. The greater the limitation of the power of gemmation to one side of the nucleus, and the larger the number of incom- plete zones, the more will the early plan of growth approximate to the spiral type, such as is represented in Plate IX. figs. 2, 4. It is obvious that the existence of these intermediate gradations breaks-down that barrier between the extreme forms, which Professor WILLIAMSON had proposed to erect ; and shows that in this, as in many other particulars, differential characters, which at first sight appeared to be per- fectly satisfactory, lose all their force when carefully traced through a sufficiently extended series of specimens. 56. It is desirable to note, as bearing on the relations between Orbitolites and Orbiculina, that even in those forms of the first-named type, in which the spiral mode of early growth is most characteristically displayed, it never seems to proceed far beyond a single turn ; and further, that the later portion of this whorl merely surrounds the earlier, and does not cover it ; so that unless (as sometimes happens, Plate V. fig. 5) the nucleus should itself be thicker than the zones of cells which im- mediately surround it, there is no central protuberance. In Orbiculina, on the other hand, the early growth invariably takes place according to the spiral type ; this type is always maintained, until several turns have been made ; and the later whorls not only surround but cover-in the earlier, so as to give rise to the central knob or pro- tuberance. Some general remarks, which I have to make on the combination of the helical and cyclical types of growth, bearing upon certain fundamental questions of classification, will be more appropriately introduced in a subsequent Memoir, after the structure of Orbiculina shall have been compared with that of Orbitolites. 57. It is not, however, in the early mode of development alone, that striking diversities present themselves ; for numerous variations, some of them quite as remarkable, are seen in the course of the evolution of the several parts which are characteristic of the ' complex' type. Thus, in the first place, the intermediate stratum is sometimes entirely deficient in the zones immediately surrounding the nucleus ; so that the upper and lower annuli of sarcode are represented by only a single band, as is indicated by the singleness of the aperture through which it passes. In the specimen figured in Plate V. fig. 9, we see this to be the case only with the three zones nearest the centre ; in that represented in Plate V. fig. 10, the canal is single in the Jive inner zones ; whilst in that represented in Plate V. fig. 7 5 the canal is single for the first twenty-three zones. Whenever the annular canal is single, the upper and lower superficial cells also become continuous, and form a series of columnar cells in every respect similar to those of the simpler type (compare Plate V. fig. 5 with the portion a b of Plate V. fig. 7). If, then, the growth of either of these disks had been checked within the first zone in which its annular canal becomes double, it would have been accounted as belonging to the simpler type ; and the wide variation which here shows itself, in regard to the distance from the nucleus at which the more com- 220 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. plex type begins to be assumed, sufficiently explains the fact already noticed (^[ 22.), that although the disks of the 'simple' type are for the most part of minute size, yet that the diameter of many of them exceeds that of the smaller disks of the complex type. 58. Even when an intermediate stratum is formed by the separation of the annular canals, the superficial cells are not always clearly marked-off from its columnar cavities ; for instead of being separated by floors formed by the expanded summits of the zonal septa (^[ 28.), they sometimes open at once into the columnar cells of the intermediate substance, so as to be quite continuous with them. Of this we have an example in the three zones 6 8 of the specimen represented in Plate V. fig. 10, b c. This continuity of the superficial cells with the intermediate columns, is sometimes maintained throughout the disk, so that in no part of it are the former clearly marked off from the latter ; as is seen in the portion b d of fig. 7, Plate V. ; in which, however, the intermediate layer is much less regular than usual. This method of growth is so remarkably constant in the Fossil Orbitolites of the Eocene strata, whose intermediate layer is fully and very regularly developed (see Plate VI. figs. 10, 11, and Plate VIII. fig. 2), that it might be considered to be specifically characteristic of them, did we not occasionally find it to occur in certain zones of recent disks, which are elsewhere exactly conformable to what I have described as the regular type. Thus in the vertical section represented in Plate V. fig. 10, we see that whilst the superficial cells of the three zones b c are continuous with the columns of the intermediate stratum, a change then occurs in the relative places of their zonal septa, so that the cells of the former come to be, as it were, detached from the columns of the latter, and to have floors formed by the summits of the partitions by which these are divided. It is, as already remarked (^j"51.), where the superficial cells are continuous with the columnar cells of the intermediate substance, that they present the rounded or ovoidal shape, instead of the elongated straight-sided figure which is their characteristic form. And the former seems to give place to the latter, whenever the cells of the superficial layers are perfectly separated from those of the intermediate stratum, and are connected only with the annular passages. 59. The intermediate stratum, again, may be altogether wanting, notwithstanding that the two superficial layers are separated from each other by a horizontal partition. In this case, each layer has its own annular canal ; and its cells have sometimes such an arrangement as regards those of the other layer, that one of the connecting stolons from which each segment arises, will pass into the alternating cell of the upper layer, and the other into that of the lower. This arrangement may present itself as one of the modes of transition from the simpler to the more complex type, as is shown in Plate V. fig. 9; the columns being disposed to subdivide transversely when they attain a considerable length, and the annular canal becoming double ; whilst in zones more distant from the centre, the two layers are separated by the interposition of the intermediate stratum. Sometimes, however, the disk continues to increase and attains GENUS ORBITOLITES : VARIATIONS IN MODE OF GROWTH ; MONSTROSITIES. 221 a considerable size on this duplex type ; and its edge then presents two rows of rounded prominences with pores between them, those of the upper and lower rows alternating with each other, as is well seen in the vertical plate of the monstrosity represented in Plate IX. fig. 10. It is on a disk of this type that Professor EHREN- BERG has founded his genus Amphisorus, which I cannot regard as even specifically distinct from the ordinary Orbitolite. 60. The next variety to be noticed, consists in a complete absence of regularity in the disposition of the columnar cells of the intermediate stratum, so that they present an assemblage of indefinitely-shaped passages, communicating with each other in various directions. This variety is chiefly interesting, as showing how little import- ance is to be attached to smaller deviations of the same kind. The most remarkable example of it which I have met-with, is represented in Plate V. fig. 7, c d. In Plate V. figs. 1 1, 12 are represented two examples of irregularity in the disposition of the super- ficial and intermediate cells in the zones immediately surrounding the centre. 6 1 . Lastly, I have to mention, that the septa dividing the contiguous cells of the same zone are occasionally deficient, so that the interior of the zone is a continuous circular passage, with only slight indications of the normal divisions. In such a case, it is ob- vious that the ring of sarcode must have been everywhere of nearly uniform thickness, showing no division either into horizontal or into vertical segments ; and it may not be thought improbable that this is its first condition in every case, and that its seg- mental division is a subsequent process, so that the shelly investment, if formed pre- viously to the segmentation, will have the character of incompleteness just described. I cannot help suspecting, that the peculiar groove around the margin of the Feejee specimens formerly noticed (^[25.), is referable to a still greater incompleteness of the production of the calcareous investment around the newly-forming zone. 62. Monstrosities. Besides those departures from the normal type of growth, which have been described as variations or irregularities, there are certain others of rarer occurrence, which can only be regarded as ' monstrosities by excess ;' consisting in the production of one or more incomplete secondary disks by outgrowth from the first. In the specimen represented in Plate IX. fig. 8, the secondary disk forms a half-circle BD, of about the same diameter with the primary AC, and is superposed vertically upon the latter, the plane of junction passing through its centre. In the specimens represented in figs. 7 and 9, the secondary disk is relatively j smaller, extending only from the centre to the margin of the primary, B but still meeting it nearly at right angles. In the specimen of which two different aspects are shown in figs. 5, 6, it would seem difficult to say which is the primary and which the secondary disk, and it would be more correct to describe the entire structure as consisting of a single half-disk AB and of A_ two half-disks BC and BD, meeting each other at an acute angle CBD, neither of them being in the same plane with the single half-disk, but both of them meeting it at similarly obtuse angles ABC and ABD. The opening of one of the MDCCCLVI. 2 G 222 DR. CARPENTEll'S RESEARCHES ON THE FORAMIN1FERA. obtuse angles is shown in fig. 6, and that of the acute in fig. 5 ; and in each view it is seen, that the divergence takes place in a plane which passes through the common centre of all three. The specimen delineated in fig. 10 exhibits a multiple outgrowth of a nature resembling that shown in figs. Sand 9. For from the surface of the disk there rises a triradiate crest, formed by three vertical plates meeting one another at nearly equal angles, but all of them nearly perpendicular to the plane on which they rest. It is a very remarkable feature in this specimen, however, that the line in which the three vertical planes meet, is traceable at its base to the nucleus of the horizontal disk ; so that they all bear the same relation to it, as does the single outgrowth in the instances previously cited. Hence we may attribute all these monstrosities to an excess of productive power in the sarcode of the original nucleus, which has put forth its first extension, not merely in the horizontal, but also in the perpendicular direction ; the whole subsequent development of these outgrowths taking place after the normal plan, from the foundation thus laid. It is interesting to remark, that the presence of such outgrowths as those now described, is far more frequent in certain localities than it is in others. Among some hundreds of specimens which I have examined from the coast of Australia, I have only met with those represented in figs. 7, 8, 9, and two or three others ; the remarkable specimen delineated in figs. 5, 6, occurred with another less peculiar among a comparatively small number of Orbitolites collected by Mr. CUMING in the Philippine Seas; but in a small collection which I have inspected from the ^Egean Sea, the monstrosities of this kind (of which fig. 10 was the most remarkable) were so numerous, that I think I am scarcely wrong in assert- ing that one specimen out of every three or four presented some excess*. Among the fossil Orbitolites of the Paris basin, the presence of a completely-semicircular vor- tical plate is not at all uncommon. 63. There may be some doubt in the first instance, as to the light in which we are to regard the specimen represented in Plate VIII. fig. 10; whether as a ' monstrosity by excess,' or as the product of the fusion of two individuals : but I think this will be removed by a closer examination. For it is obvious, that the smaller disk, which is surrounded by the outer zones of the larger one, has been developed from a nucleus of its own ; and this nucleus does not appear to have any direct connexion with the periphery, still less with the centre, of the larger disk : on the other hand, when we consider the circumstances under which Orbitolites grow (^[ 34.), it is very easy to understand, that the smaller and younger individual, having attached itself in too near proximity to the larger and older one, should become imbedded therein (so to speak) by the extension of the newly-forming zones of the latter around its margin. * This is by no means a solitary case of the prevalence of monstrosities in particular localities. The collec- tion of Mr. BEAN of Scarborough contains a number of curiously- distorted specimens of the common Planorbis marginatus, -which have all been collected in one brook. Their peculiarities are by no means repetitions of each other ; and I am disposed, therefore, to regard them rather as resulting from the influence of external condi- tions, than as accidental varieties hereditarily propagated. GENUS ORBITOLITES: MONSTROSITIES; ESSENTIAL CHARACTERS. 223 In fact, the manner in which the outer zones of the larger disk envelope the smaller, precisely corresponds with that in which we have seen the new zones originating from the uninjured margin of a mutilated specimen, to extend themselves along its frac- tured edges (^[37.). V. Of the Essential Characters of Orbitolites, and of its relations to other Types of Structure. 64. If, now, we seek to determine the essential characters of that type of organi- zation which is known under the designation Orbitolites, we find them to lie in the presence of a series of annuli of sarcode (and of corresponding passages in the shelly disk) arranged concentrically round a nucleus ; each zone in the simpler type con- taining but a single annuhis, so constricted at intervals as to form a series of some- what columnar segments (occupying the cells of the shelly disk), connected with each other by narrow bands of sarcode ; whilst in the more complex type each zone contains two such annuli, including between them a portion of its series of columnar segments, so as to constitute an intermediate stratum, distinct from the superficial portions. In either case, the segments of successive zones freely communicate with each other by radiating bands of sarcode (also leaving passages in the shelly disk), whose normal direction is such as to connect each segment with the two segments that alternate with it in each of the adjacent zones. 65. Now the addition of new zones, each similar to the last, is a simple matter of growth ; but the passage from the simpler to the more complex plan marks an advance in development; and this advance essentially consists (here as elsewhere) in a progressive differentiation of parts. When, with the vertical extension of the columnar segments, the annular canal subdivides itself into two, the communica- tions between the successive zones no longer come-off, as before, from the annular canal, but from the intermediate portions of the columnar cells ; and instead of the two diverging passages from each cell being in the same plane, they lie in different planes, alternating with each other vertically. Up to this point, we observe little else than a multiplication of parts vertically, as well as horizontally, and a separation of connexions that were previously confluent. But in the highest stage of develop- ment, we find a marked alteration in plan ; for those portions of the columnar seg- ments, which lie between the two annular canals of each zone and the two surfaces of the disk, become completely differentiated from the portions that occupy the inter- mediate stratum, so as to form a peculiar set of superficial cells ; and these are so equally connected with two zones, as to make it impossible to say that they belong specially to either. 66. Now we have seen that development may be checked, while growth continues, at any period of its progress ; so that we find Orbitolites growing to a considerable size upon the very simplest plan, others still larger formed upon the duplex plan, the largest yet known (fossilized in the Paris basin) developed upon the multiple plan 2 G 2 224 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. without separation of the superficial cells, while the most complete, in regard alike to multiplication and to differentiation of parts, are only found among the disks at present existing; and it is interesting to observe, that some of them present this highest grade of development, while as yet of comparatively minute size. I am not acquainted with any other Animal body, in which so wide a range of developmental variation normally exists. The lower classes of the Vegetable Kingdom, however, especially the group of Fungi, afford abundant examples of it*. 67- The extreme freedom with which all the cavities of the shell mutually commu- nicate, is a very marked feature in the structure of this type; and shows that the several parts of the animal body are far more closely connected into one whole, than they are in most of the other Forarninifera whose general plan of conformation is more or less analogous. Indeed, if we were to imagine a discoidal mass of sarcode to be traversed by a reticulated calcareous skeleton, somewhat resembling that open areolar texture which forms the shell of the Echinida, and this network to possess something of that regularity of the disposition of its successively formed parts, which is presented to us in the spines of the Echinida, we should have no unapt repre- sentation of the relation of the shelly disk of the Orbitolite to the animal which it envelopes. There are certain Sponges which have a reticulated skeleton composed of mineral matter disposed in a mode not altogether dissimilar, whilst the constitu- tion of their soft bodies is essentially the same. And a remarkable connecting link between Orbitolites and Sponges, seems to be presented to us in the curious Thalas- sicolla discovered by Mr. HUXLEY -j~. The relations of Orbitolite to other Forami- nifera have already been partly touched-on, and will become more clear hereafter, when the types which most approximate to it shall have been themselves described. VI. Of the Species of Orbitolites. 68. It only remains to inquire, whether the diversities which have been described as existing among Orbitolites, afford any ground for assuming the existence of more than a single species. With regard to the recent forms, with which, so far as they are at present known, I have made myself fully acquainted, I can speak confidently ; since, as I have demonstrated, the Orbitolite with a single stratum of cells (O. mar- glnalls of LAMARCK, Sorites of EHRENBERG), that with a double stratum (Amphisorus of EHRENBERG), and that with multiple strata (Marginopora of QUOY and GAIMARD, Orbiculina Tonga of Professor WILLIAMSON), are fundamentally the same forms, deve- loped in three different modes. 69. Of the identity of all i\\e fossil species with the foregoing, I cannot speak with the same confidence; since there are some of which I can only judge from figures. Into the structure of that which is best known, however, and which has been com- monly accounted the type of the genus, viz. the Orbitolites complanatti of the Paris * See especially the' recent Memoirs of M. TULASNE, in the ' Annales des Sciences Naturelles.' . f Annals of Natural History, 2nd series, vol. viii. p. 433. GENUS ORBITOLITES: SPECIES. 225 basin, I have made investigations scarcely less minute and extended than into that of the recent forms ; and I have come to the conclusion, that it cannot be specially distinguished from the large Australian Orbitolite, to which it bears a very obvious general conformity. It is true that it differs from the typical forms of the latter in two important features of structure, which are, however, mutually connected ; namely, the direct continuity of the cells of the superficial layers with the columnar cells of the intermediate layers; and the rounded or ovoidal form of the superficial cells, which (as already stated) these always possess, as in the simpler type, unless they are disconnected with the columnar cells, and communicate only with the annular stolons (see ^[ 58.). But since this very peculiarity does present itself in certain existing individuals, whose development seems to have taken place upon a lower type, and since it occasionally shows itself in the course of the passage from the simplest to the most complex type, in such as ultimately attain the latter, there appears to me no room for questioning the specific identity of the O. complanata with the Australian forms, notwithstanding that I have never met, among the numerous specimens which I have examined of the former, with those elongated parallel-walled superficial cells, which constitute the most distinctive feature in the latter. It may be well, moreover, to bear in mind the remark I have already made, respecting the local prevalence of particular varieties of form ; since there is nothing more strange in the incompleteness of the type of development presented by the Paris-basin Orbi- tolite, than in that tendency to excessive development, which gives rise to the nume- rous monstrosities that are presented by the ^Egean specimens (^[ 62.), or in those radial deposits on the surface, which are so common among the Philippine forms (^[ 53.). My belief in the specific identity of this fossil with the recent types has been strongly confirmed by the circumstance, that among the Paris-basin forms I have found a minute specimen, which corresponds in every respect with the simple type of the existing species. 70. Of the other fossil species cited by LAMARCK, the O. macropora of the Maes- tricht beds, judging from the figure given of it by GOLDFUSS, is nothing else than an Ocbitolite of simple type, whose marginal cells have been laid open by attrition both above and below, as in Plate VII. figs. 8, 10. The O. concava and O. pileolus of LAMARCK are not distinguished in his definition by any other character than that drawn from form, which we have seen to be so variable as to be quite insufficient as a distinctive feature. It is quite possible, moreover, that they may belong to another type, nothing being said in the description of them, either of concentric lines, or of pores. If, as I believe, the O. concava of LAMARCK (figured by MICHELIN in his ' Icon. Zoophyt.,' pi. 7- fig. 9) be identical with the O. conica of M. D'ARCHIAC, I feel certain (from careful examination of its imperfectly-preserved internal structure) that, whatever it may prove to be, it is not an Orbitolite. So again, the O. lenticulata of LAMARCK, judging by the figure given of it by LAMOUROUX*, is not an Orbitolite, * Polypiers, pi. 72. figs. 13, 16. 226 DR. CARPENTER'S RESEARCHES ON THE FORAM1NIFERA. but probably a Lunulite. Of the species subsequently described by other authors, I entertain no doubt that the O. disculus of M. LEYMERIE *, as well as in all proba- bility the O. plana of M. D'ARCHIAC, is nothing else than a variety of O. complanata, more especially as its young is said to be like the O. macropora of Maastricht. The O.gensacica, O. set-u/ts, and O. socialis of M. LEyMERiE, are all undoubtedly Orbitoi- des, and all belong, I believe, to the same species ; though on this point I could not speak positively, without an examination of their internal structure. So, again, the O. mamillata, O. Fortisii (O. gigantea of D'ORBIGNY), O. papyracea, O. stellata, O. sella and O. radians (O. radiata of D'ORBIGNY) of M. D'ARCHIAC -f- are all probably Orbi- toides, as he has himself subsequently recognized in regard at least to some of them J. The O. elliptica of MICHELIN, so far as I can judge from the figure he gives of it (pi. /I, fig. 1 1), is certainly not entitled to rank as a distinct species, its elliptic form being utterly insufficient to separate it (^j 47.). And as I have already pointed out (^| 49.), unless some distinctive character be furnished by the internal conformation of the bodies which have been ranked in the genus Cyclolina by M. D'ORBIGNY, this also should take rank merely as a variety of Orbitolites . VII. Concluding Remarks. 71. It might be asked with some show of reason, what good purpose can possibly be answered by such a minute and prolix description of a type of animal structure, so mean and insignificant as that which has been occupying our attention. To such a question I would reply First, that I hold it to be a worthy labour to learn, and to place within reach of others, everything that can be learned respecting any form of Organized Being ; that such a complete acquaintance is the great desideratum in every department of Biological Science ; and that no works have ever exercised so beneficial an influence on its progress, as those admirable Monographs of single species, which, by thoroughly elucidating their structure and physiological history, have served as a basis for all subsequent inquiry into the nature of the Plants or Animals formed on a like plan: Secondly, that such an inquiry can scarcely be otherwise than of peculiar utility, as relating to a tribe of Animals whose nature and history are almost as unknown to us now, as those of Polypes were to the Naturalists of a century ago, when TREMBLEY wrote his immortal ' Me" moires pour servir a 1'Histoire d'un Genre de Polypes d'eau douce;' what little is known respecting them, being of a nature to mark them as distinct from every other type of living beings * Mem. Ge'ol. Soc. de France, 2 se'r., torn. v. pp. 190, 191. t Ibid. torn. ii. p. 178, and 2 ser., torn. iii. p. 4. J Description des Animaux Fossiles du groupe Nummulitique de 1'Inde, p. 350. For a recent description of a form of this reputed genus, which occurs in contiguity with typical Orbito- lites, and which seems to me to correspond in every respect with those recent specimens of which Plate VII. fig. 14 exhibits the external aspect, see Mr. CARTER'S Memoir on the Fossil Foraminifera of Scinde, in 'Ann. of Nat. Hist.,' 2nd ser., vol. xi. p. 174. GENUS ORBITOLITES: CONCLUDING REMARKS. 227 with which we have a tolerable acquaintance: and Thirdly, that certain general principles evolve themselves as results of these investigations, which are quite as applicable to every other department of Biological Science, as they are to the single case of the Orbitolite. 72. It has been shown that a very wide range of variation exists among Orbitolitex, not merely as regards external form, but also as to plan of development ; and not merely as to the shape and aspect of the entire organism, but also with respect to the size and configuration of its component parts. It would have been easy, by selecting only the most divergent types, from amongst the whole series of specimens which I have examined, to prefer an apparently substantial claim on behalf of these to be accounted as so many distinct species ; and I could thus have easily created an almost indefinite number of such species. But after having classified the specimens which could be arranged around these types, a large proportion would yet have remained, either presenting characters intermediate between those of two or more of them, or actually combining those characters in different parts of their fabric; thus showing that no lines of demarcation can be drawn across any part of the series, that shall defi- nitely separate it into any number of groups, each characterized by features entirely peculiar to itself. Thus, then, we see that the real relationship of the different types to each other, can only be determined by the careful comparison of a very large number of individuals obtained from as many different sources as possible ; a process which is too frequently neglected by Systematists, many of whom erect species, and even genera, without the least care to determine, by any such process, the real value of the distinctions by which they characterize them. 73. The right mode of proceeding in every other department of Natural History, must be that which has thus been proved to be the only reliable method in this ; and I venture, therefore, to lay down the following general Canons, deduced from the results of the preceding investigation; which will, I think, be found accordant with the experience of all soundly-judging Naturalists, although they have not yet, so far as I am aware, received a formal expression. 74. To become fully acquainted with the Natural History of any Species, it is requisite ; ( 1 ) to study not only its external conformation, but also its internal organ- ization; by which alone can the value of superficial resemblances or differences be duly estimated ; (2) to trace out its entire developmental history, so that the true relationship of individuals in different stages of evolution may be appreciated ; (3) to compare together a large number of specimens taken from any one locality, in order to ascertain what is the range of its variation under nearly identical condi- tions ; (4) to search out its whole Geographical distribution, and to bring into mutual comparison large numbers of specimens from the remotest regions, as well as from all the intermediate areas over which it spreads, in order to determine the range of its variation under the most diverse conditions ; and (5) to follow the same course of comparison throughout its Geological distribution, still relying only on the informa- 228 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. tion afforded by large numbers of specimens, collected (so far as may be possible) from different formations, and from different geographical areas. Until the whole of this process shall have been carefully and systematically gone through, no limitation of a species by a definition of any kind, can be regarded in any other light than as a provisional means of marking-out the existence of a particular type of structure, whose relationship to other types must be a matter of further investigation. 75. Let me subjoin such "pregnant instances," as shall prove the importance of each of the foregoing principles, from the result of the violation or neglect of it : (1) so long as external conformation was alone regarded, and no account was taken of internal organization, the Nautiloicl Forarninifera were placed among Cephalopods, and the Coralloid forms among Polypifera; to neither of which classes have they any kind of relationship ; (2) so long as developmental history was un- studied, the Hydroid Zoophytes and the Medusoid Acalephse were considered as entirely disconnected groups, belonging to two different Zoological classes, instead of (as in reality) different states of the very same organisms ; (3) so long as reliance is placed on the comparison of a. few individual specimens only, without any account being taken of the intermediate forms by which the more divergent types may be connected, so long are species multiplied to a most unwarrantable excess, as is found to be the case in almost every department of Zoology and Botany by those who devote themselves to a more extended comparison ; thus, nineteen species have been made from the common Potatoe, and many more from the Solanum nigrum ; so, multitudes of species have been instituted in various genera of California!! shells, by the late Mr. C. B. ADAMS, whose identity is established by a more extended comparison of individuals (as will be shown in a Report which is being prepared at the request of the British Association, by my brother, the Rev. P. P. CARPENTER) ; in fact, wherever this test, is conscientiously applied, its effect is to reduce the number of reputed species, sometimes in a most remarkable degree*. (4) In like manner it has been by comparing only a small number of specimens from remote geographical provinces, and by neglecting the intermediate varieties that present themselves even among sufficiently large collections from these, still more among specimens collected from intervening regions, that not only numerous errors of detail have been com- mitted, but general doctrines have been propounded, which the advance of Science has proved to be utterly untenable. As an example of the former kind, may be cited the facts mentioned by Dr. J. D. HOOKER (op. cit.), that of the New Zealand varieties of Oxalis corniculata, one of the most widely-diffused and most variable Flowering plants in the world, no less than seven or eight species have been made, neither of them supposed to be identical with any belonging to the European Flora ; whilst Pteris * I am most glad to find my views on this point in accordance with those of Dr. JOSEPH D. HOOKER (see his ' Introductory Essay on the Flora of New Zealand,' 2), who has been led to the conviction, that instead of affirming the existence of 100,000 species of known Plants, we ought not to reckon more than half that number. GENUS ORBITOL1TES: CONCLUDING REMARKS. 229 aquilina has a different name in almost every country in the world. It has been through reliance on such ignorant determinations, most of them proceeding on the notion that " the plants (or animals) of newly-discovered, isolated, or little-visited localities must necessarily be new," that the doctrine of the universal distinctness of the species of the New World from those of the Old, and of those of the Southern from those of the Northern Hemisphere, has attained a very wide currency amongst Naturalists, and is still obstinately persisted-in by some, in defiance of ample evidence to the contrary*. (5) Lastly, not only has the limitation of the comparison among Fossil types, to a small number of individuals, led to the excessive multiplication of species in the forms that are furnished by the same strata ; but the same habit of relying on minute differences, without attention to osculant characters, has given rise to that disposition to regard the species of successive formations as necessarily different, which is introducing the greatest confusion into geological and palaeonto- logical determinations of every kind. How an extended comparison of individual forms tends not only to reduce the number of reputed species, but to establish the continuity of the same specific types fiom one stratum to another, will be remarkably seen when the laborious researches of Dr. WRIGHT of Cheltenham on the Cidarites of the Liassic and Oolitic formations shall have been made public. 76. Another general consideration of some interest, appears to me naturally to connect itself with the foregoing history, namely, that the lower the general plan of organization of any being (that is, the greater the prevalence of ' vegetative' or ' irrelative' repetition in its different parts), the more is that plan liable to be modified by slight differences in external conditions, and the wider, therefore, may we expect its range of variation to be, if it be disposed to vary at all. In some instances, it is true, there appears (as in many higher forms of organization) to be an absolute incapacity for any such variation ; and a limitation of the geographical and geolo- gical distribution of the species results from its want of power to exist under any great diversity of external conditions. But when the same general type of organiza- tion is found to prevail extensively both in space and in time, it may, I think, be safely regarded as probable, that that type has within itself the power of accommo- dation to a considerable diversity of external conditions ; and hence that in the com- parison of individuals, differences of conformation should be considered as of less account towards the establishment of specific distinctions, than they are when there is an obvious restriction of the type to a limited Geographical area or a' particular Geological epoch. 77- In the foregoing communication, I have thought it right not only to make known the results of my researches, but so to develope my plan of investigation, that the value of those results may be duly estimated. In the memoirs which I trust to be * " Thus as long ago as 1814, Mr. ROBERT BROWN gave a list of 150 European plants common to Australia and Europe. The identity of many of these has repeatedly been called in question, but almost invariably erroneously ; added to which, more modern collectors have greatly increased the list." HOOKER, op. cit. p. 18. MDCCCLVI. 2 H 230 DR. CARPENTER'S RESEARCHES ON THE FORAMIN1FERA. enabled to forward from time to time, regarding those other typical forms of Fora- minifera of which I have made a special study, it is not my intention to do more than state the results ; hoping that they may be understood to have been attained by a method of inquiry as closely resembling that which I have here followed through its details, as the circumstances of each case may have admitted. I would have it borne in mind throughout, that, as has been admirably remarked by one of the most accom- plished Botanists of our time, " the Naturalist who has the true interest of science at heart, not only feels that the thrusting of an uncalled-for synonym into the nomen- clature of science is an exposure of his own ignorance, and deserves censure, but that a wider range of knowledge and a greater depth of study are required, to prove those dissimilar forms to be identical, which any superficial observer can separate by words and a name" (Dr. J. D. HOOKER, op. cit. p. 14, note). EXPLANATION OF THE PLATES. PLATE IV. Structure of the Animal of Orbitolite. Fig. 1. Entire Animal, from a small and simple disk, the shell having been removed by maceration in acid ; in the peripheral portion the segments of sarcode are wanting, and the structureless residuum of the shell is alone seen : magnified 40 diameters. Fig. 2. Appearance of a portion of Sarcode, highly magnified : 180 diam. Fig. 3. Portion of the body of one of the more complex forms (resembling fig. 4), in which the sarcode has broken up into little spheres (gemmules?); a, a, superficial segments ; b, b, annular band : 180 diam. Fig. 4. Portion of the body of one of the more complex forms, as seen in vertical section ; aa, a'a', upper and lower rows of superficial cells, each cell con- nected, at its two extremities, with the annular bands bb and b'b' of two zones ; from these annular bands spring the columnar segments cc, cV, those of the same zone occasionally passing into each other, and commu- nicating with those of the next zone by oblique peduncles alternately pass- ing towards one side and the other: 150 diam. (N.B. This figure is somewhat ideal, being made-up from several preparations ; but for every point which it represents, these preparations give warranty.) Fig. 5. Nucleus and first two annular zones, exhibiting the typical conformation ; a, the central segment ; bb, the circumambient segment, from the entire margin of which are given off peduncles of sarcode, which give origin to the first annular zone : 84 diam. GENUS ORBITOLITES: EXPLANATION OF PLATES. 231 Fig. 6. Nucleus and first two annular zones of another disk, showing a deficiency in the connexions of the first annular zone with the circumambient segment bb, for about a third of the circumference of the latter, between the points cc: 84 diam. Fig. 7- Portion of one of the surfaces of an animal of complex type (as in fig. 4) : aa, aa, rows of superficial cells, connected at their two extremities with the annular bands bb, bb: 150 diam. Fig. 8. Portion of fig. 1, enlarged, to show the ordinary mode in which the segments of each zone are connected by peduncles with the annular band of the pre- ceding zone, so as to alternate with its segments ; showing also the occasional interpolation of additional segments, a, a, whose peduncles come off from the segments of the preceding zone: 90 diam. Fig. 9. Portion of a section of the shelly disk enlarged, to show the corresponding appearances it presents ; each cell being ordinarily connected by a radial passage with the annular canal of the preceding zone, and thus with the two cells alternating with itself; but cells a, a, being sometimes interpo- lated, which open directly into the cells of the preceding zone : 90 diam. Fig. 10. Central and circumambient segments (a and bb) of a large disk, showing the origin of the segments of the first annular zone, c, c, from less than half the circumference of the nucleus : 84 diam. Fig. 11. Peculiar bodies (ova?) found in the substance of the sarcode in different parts, showing successive stages, a, b, c, d, e, f, of binary subdivision ; g, other bodies of somewhat larger size, found in one of the superficial cells of a vertical section : 130 diam. Fig. 12. Central portion of fig. 1, enlarged to the same scale as figs. 5, 6, 10 ; show- ing the central segment a, the circumambient segment bb, and the origin of the first annular zone in three peduncles proceeding from the end of the latter : 84 dfam. Fig. 13. Central portion of another disk, showing the origin of the first annular zone by eight peduncles from the circumambient cell : 84 diam. PLATE V. Structure of the Calcareous Disks of Orbitolite. Fig. 1. Ideal Representation of a Disk of the Simple Type, the details of the different parts rnade-up from actual specimens ; showing the natural surface, with the markings of the cells ; the natural margin, with the single row of pores between the protuberances of the cells ; a portion of the interior, as dis- played by a horizontal section, showing the central cell a, the circumam- bient cell b, b, the concentric zones of cells c, c, with the annular passages which connect together the cells of the same zone, and the radiating passages 2 H 2 232 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. which extend from the annular passages of each zone to the cells of the next, and, in the outermost zone, to the pores d, d; another portion as displayed by a vertical section e, e, in a radial direction, which lays open the colum- nar cells, but passes through the intercellular partitions, in alternate zones; and another portion ff, as displayed by a fracture in the course of one of the zones, laying open the entrances to the cells from its inner or central side. Figs. 2, 3. Two large recent disks from the Feejee islands, plicated towards the mar- gin, but one much more so than the other, and with a projection of the upper and lower edges, so as to leave a deep marginal furrow: enlarged 2 diam. Figs. 4, 5. Vertical sections of two disks, drawn under the same magnifying power, showing the marked difference in the size and proportions of their parts. In each disk we see the central cell, with the circumambient cell laid open on either side of it ; and the cells of successive zones, with their commu- nications : 30 diam. Fig. 6. Ideal Representation of a Disk of the Complex Type; the details of the different parts made-up from actual specimens : a, central cell ; b, circum- ambient cell ; c, c, concentric zones of oblong superficial cells, some of them laid open ; d, d, marginal pores, forming several rows ; d", d'", cor- responding pores of inner zones, once marginal, but now connecting them with surrounding zones; e, e, vertical section in a radial direction, show- ing the zones nearest the centre to be made up of simple columnar cells, but those of the remainder of the disk to be composed of two superficial layers and of an intermediate stratum ; /,/, floors of the superficial cells, with an aperture at each end of every one ; g, g, annular canals, running beneath these floors, with the large apertures leading to the columnar cells of the intermediate stratum ; g 1 , the same canals near the other surface of the disk ; g", the same canals laid open through the plane at which they give off the two passages into the superficial cells ; g"', the same canals, as out transversely by a vertical section ; h, passage of the horizontal section through the intermediate stratum, showing the summits of its columnar cells about to enter the annular canals ; i, i, i, and k, k, k, passage of the horizontal section through two different planes of the intermediate stra- tum, showing the connexion between the columnar cells of successive zones, by oblique passages running in different directions ; /, portion immediately surrounding the nucleus, formed upon the simple type, as in fig. 1. Fig. 7- Vertical section, taken in a radial direction, of a recent disk incompletely developed on the complex type ; showing at a, a single chamber of the cavity of the nucleus, the section having traversed the circumambient cell; from a to b, including twenty-three zones, the disk developed upon the simple GENUS ORBITOLITES: EXPLANATION OF PLATES. 233 type ; from b to c the annular canal double, and an irregular intermediate stratum interposed, from which, however, the superficial cells are not com- pletely differentiated ; and from c to d the imperfect separation of the super- ficial cells from the intermediate layer, and the extreme irregularity of the latter: 50 diam. Fig. 8. Tangential section, near the margin of the same disk, showing the same peculiarities : 50 diam. Fig. 9. Central portion of a disk, showing the nuclear cavity divided into two (the section having traversed the central cell and one side of the circumambient) ; the formation of the first three zones (a b) on the simple type ; the im- perfect separation of the superficial cells from the intermediate stratum in the next two zones (b c) ; and the regular development of the remainder (c d) on the complex type : 50 diam. Fig. 10. Vertical section of the central portion of a disk of complex type; showing four cavities in the nucleus ; the first five zones (a 6) constructed upon the simple type, with a single annular canal, but an incipient separation manifesting itself between the superficial cells and the intermediate stratum ; in the next three zones (b c) the annular canal double, but the superficial cells still partly continuous with the columnar ; the outer part (c d) framed according to the regular complex type: 48 diam. Fig. 1 1. Vertical section of a disk of complex type, showing an unusual development of vertically superposed cells immediately around, and even partly cover- ing, the nucleus, from which they arise by four layers of passages ; a pro- gressive diminution in the thickness of the disk, as far as the fifth zone (b), with a gradual approximation towards the regular type, on which the remainder (b c) is developed : 48 diam. Fig. 12. Vertical section of a disk of complex type, showing commencement of its first zone by a single layer of passages (a) from the nucleus ; the presence of two annular canals, with the absence of separation of the superficial cells, in the first zone (b) ; the incomplete separation of the superficial cells from the intermediate columnar portion in the second zone (c) ; and the normal conformation of the remainder (c d) : 48 diam. PLATE VI. Fig. I. Portion of a disk of simple type, from the immediate neighbourhood of the nucleus, showing the beak-like projections at the sides of the passages very remarkably developed: 100 diam. Fig. 2. Horizontal section of the intermediate layer, passing (in consequence of the flexure of the disk) through two different planes, and thus showing the two different directions of the oblique passages which connect the colum- 284 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. nar cells of different zones ; also showing a marked difference in the size of adjacent cells: 100 diam. Fig. 3. Horizontal section through one of the superficial layers, showing the elon- gated form of its cells, and the aperture at each end of their floor: 100 diam. Fig. 4. Nucleus and surrounding zones of the excentrically-developed disk shown in Plate IX. fig. 3 ; the beak-like projections at each side of the pores open- ing into the new zones (shown in fig. J) are here generally deficient: 100 diam. Fig. 5. Appearance of the thin wall covering-in the central cell of a large nucleus, as seen under a high magnifying power, showing the quasi-cellular mark- ings : 100 diam. Fig. 6. Horizontal section through the intermediate layer of a disk of complex type, showing an unusual irregularity in the communications of the cells: 100 diam. Fig. 7- Radial section of a recent disk, showing the complex development on the normal type ; aa, bb, upper and lower superficial layers ; c, c, c, c, interme- rnediate stratum ; dd, d'd', d"d", summits of partitions between successive zones, forming the floors of the superficial cells ; e, e', e", oblique passages through the floors of the cells of the superficial layer, leading towards the annular canals ; f, /',/",/'", annular canals of four zones, near the lower surface of the disk ; gg, partitions between the adjacent cells of the same zones ; hh, perforations in these, through which the columnar cells inos- culate with each other: 100 diam. Fig. 8. Tangential section of a fossil disk of complex type, showing four columnar cells, aa', bb', cc', dd', of the same zone, divided by sinuous partitions, with the orifices leading into the columnar cells of the next interior zone: 100 diam. Fig. 9. Similar tangential section, showing four cells of the next interior zone, aa 1 , bb', cc', dd', alternating with the preceding, and the entrance of the passages of the same vertical row, alternately into one and the other of the cells on the two sides of each of the sinuous vertical partitions: 100 diam. Fig. 10. Radial section of a fossil disk, showing the incompleteness of the separation of the superficial cells from the intermediate layer, and the irregularity in the arrangement of its cells, with their numerous lateral inosculations (compare Plate V. fig. 7, b d) : 48 diarn. Fig. 1 1. Radial section of a thicker fossil disk, showing a more regular arrangement of the columnar cells of the intermediate layer (still, however, with numerous passages for lateral inosculation), and the complete continuity of the superficial cells with these: 48 diam. GENUS ORBITOLITES: EXPLANATION OF PLATES. 235 PLATE VII. Figs. 1 4. A series of specimens, drawn under the same magnifying power (35 diam.), illustrating the variations which present themselves in regard to the Size of the nucleus. The remaining figures represent various appearances exhibited by the Surfaces of Disks, simple and complex : all magnified 35 diam., except fig. 6, which is magnified 70 diam. Figs. 5 7. Portions of surfaces of simple disks, in which the concentric circles an: strongly marked, and the transverse divisions of the cells are compara- tively obscure. Fig. 8. Simple disk, slightly concave, the marginal cells laid open by abrasion, the surface marked rather by ' engine-turned' or excentric, than by concentric- circles. Fig. 9. Portion of the surface of a complex disk, having small, round, and flattened superficial cells. Fig. 10. Simple disk (from Australia), slightly concave, the marginal cells laid open by abrasion, the surface irregularly thickened by calcareous deposits. Fig. 11. Flattened simple disk (from the Philippines), with abundant superficial depo- sits, arranged with considerable regularity in a radiating direction. Fig. 12. Portion of the surface of a large complex disk, with very narrow oblong superficial cells. Fig. 13. Portion of the surface of a complex disk, showing a marked difference in the proportions of the straight-sided cells, even in adjacent zones. Fig. 14. Surface of simple disk (Cyclolmai of D'Orbigny), with an excentric nucleus, surrounded by strongly-marked concentric circles. Fig. 15. Portion of the surface of a simple disk, showing unusually large and pro- tuberant ovoidal cells. Fig. 16. Portion of the surface of a large fossil disk of complex type, showing varieties in the size and form of the superficial cells. PLATE VIII. Fig. 1. Inner surface of a zone of a thick fossil disk, showing a tolerably regular arrangement of the columnar cells, the segmental constrictions of these, and the fissures leading into those of the next row : 35 diam. Fig. 2. Outer surface of a similar zone, showing the continuity of the cells of the superficial layer with the vertical columns of the intermediate (the latter not unfrequently dividing, however, at their extremities, so as to form two cells at the end of each column), and the pores leading to the oblique 236 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. passages, directed alternately to one side and to the other, which lead to the cavities of the columnar cells : 35 diam. Fig. 3. Horizontal section of the intermediate layer, immediately beneath the floors of the superficial cells, laying open the annular canals, with the summits of the columnar cells: 100 diam. Figs. 4 9. Various examples of the Reparation of disks after fracture, and of the growth of new disks from detached fragments. For description, see ^[37 40 of Memoir. Fig. 6 is magnified 6 diam. ; the rest 35 diam. Fig. 10. Monstrosity formed by the inclusion of a young disk within the outer zones of an older one, that seems to have been brought into contact with it by the progressive increase of its own diameter: 25 diam. PLATE IX. Figs. 1 4. Various departures from the typical mode of concentric development, presented in their early state, both by simple and complex disks ; some of them even passing towards a spiral mode of evolution. Figs. 1 and 3 are magnified 35 diam. ; Figs. 2 and 4 are enlarged 90 diam. The remaining figures show various forms of Monstrosity, produced by excess of growth from the nucleus. Fig. 5, 6. Two views of a disk from the Philippine shores, having a single plane on one side, and two planes, meeting at an acute angle, on the other : 6 diarn. Figs. 7, 8, 9. Small Australian disks, with vertical crests: 35 diam. Fig. 10. Disk from the ^Egean Sea, with a tri-radiate crest: 35 diam. 10 IS : ee laA. ' Phd Trans. &DCCC1M. Plate VI . "fflO 1 i< ' - ijgj , PUL. Trans. MDCCCLJI. Plate W. 'Vy//'" PlaaLL. L? [ 551 ] XIV. Researches on the Foraminifera Supplemental Memoir. On an Abyssal type of the Genus Orbitolites ; a Study in the Theory of Descent. By WILLIAM B. CARPENTER, C.B., M.D., LL.D., F.R.S. Received May 31, Bead June 14, 1883. [PLATES 37,38.] INTRODUCTION. THE subject of this communication is a type of the genus Orbitolites first obtained in the deep-sea dredgings of H.M. Surveying Ship "Porcupine," off the north-west of Ireland, in 1869,* and subsequently brought up from various depths in other parts of the North Atlantic and also in the Mediterranean, which presents many points of general scientific interest ; the first of these being the completeness of the transition which it establishes between the Milioline and the Orbitoline plans of growth, and the full confirmation it thus affords of the validity of the principles on which my Classification of the FORAMINIFERA is founded. In the Monograph of the genus Orbitolites (1855), which constituted the First Series of my " Researches on the Foraminifera," t I embodied the results of a careful and thorough investigation of the structure and relations of all the forms under which that type was then known to me : and I showed that while the most highly developed and most specialized of these forms exhibit the cyclical plan of growth almost from the very commencement, a complete zone of sub-segments being formed by gemmation from the entire periphery of the " circumambient segment " of the central " nuclear mass,"J and the whole disk being made up of a succession of similar concentric zones, there are other forms in which the primary gemmation takes place from only one side of that mass, so as to impart to the early extension of the composite structure a more or less spiral direction, which only gives place to the cyclical after repeated gemmations. The transition from the one plan of growth to the other I showed to be made by the progressive widening-out of the spire, and the increase in the number of the * See Proc. Roy. Soc., vol. 18, p. 397. t Phil. Trans., 1856, pp. 181-236. J Ibid., Plate IV., figs. 5 and 6. Ibid., Plate IX., figs. 1 to 4. 4 B 2 552 DE.-W. B. CAEPENTEE ON OEBITOLITES TENUISSIMA. sub-segments (formed by the division of the principal segments) at every new stage of gemmation ; so that at last the aloe of the spire, extending themselves on either side round the nuclear mass, meet and complete the circlet, around which new zones are then successively budded forth, as in the forms that are cyclical from their com- mencement. I did not at that time feel justified in calling in question the validity of the order Cyclostegues, which had been instituted by M. D'OKBIGNY, for the reception of this and other types characterised by the cyclical plan of growth ; but in my Second Series (presented in the following year), which contained the results of a similar investigation of the geims Orbiculina, I showed that the latter always begins life on the spiral plan of growth, which may or may not give place subsequently to the cyclical, and that the marginal portions of a full-grown cyclical Orbiculina cannot be distinguished from similar portions of an Orbitolite. From this fact I drew the conclusion* that although Orbitolites and Orbiculina had been placed by M. D'ORBIGNY in two distinct orders, Cyclostegues, and Helicostegues, " the relation- ship between them must be extremely close ; " and ventured further to affirm that no Classification can have any claim to be considered as natural, in which they shall be widely separated." To this point I reverted in the Concluding Summary appended to my Fourth Memoir,t in which I showed how completely the results of my researches were opposed to the principles on which the Classification of M. D'ORBIGNY had been framed, indicated the line of " descent with modification " by Avhich a division of the primary segments that form the simply-chambered shell of a Peneroplis into sub- segments would give origin to the chamberlets of the spiral Orbiculina, and pointed out how gradational the transition is from the latter to the cyclical Orbitolites. When I subsequently undertook, in conjunction with my friends, W. K. PARKER and T. RUPERT JONES, to frame an entirely new Classification of FORAMINIFERA on the basis of the principles I had laid down, I felt no difficulty in assenting to their view that the pedigree of this series might be traced yet further back, viz. : to those simplest forms of the Milioline type, whose shell is a flattened nautiloid spire altogether destitute of partitions thus belonging to that monothalamous section which all previous Systematists had ranked as fundamentally distinct from the polythalamous. "From the undivided spiral of Cornuspira" I pointed out ("Introduction to the Study of the Foraminifera," p. 67), "to the regular scarcely-divided spiral of certain ' spiroloculine ' forms of Miliola, the transition is almost insensible ; and from the 'spiroloculine' we pass by easy steps to all the other forms of .the Milioline type." Again, a subdivision of the widely-expanded spire of Cornuspira into segmental chambers gives us Peneroplis, with its septal planes perforated by a row of separate pores ; while from this, it was again pointed out, the spiral Orbiculina might * Phil. Trans., 1856, p. 552. t Ibid., 1860, p. 571. DE. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 553 be derived by a further division of the segments of the sarcodic body into sub- segments, with a corresponding division of the primary chambers of the shell into chamberlets. In the new specific type of Orbitolites I have now to describe, the whole transition which I thus hypothetical ly indicated, is actually presented during the successive stages of its growth. For it begins life as a Cornuspira, taking-on that ' spiroloculine ' condition which marks the passage towards the Milioline type : its shell forming a continuous spiral tube, with sh'ght interruptions at the points at which its successive extensions commence ; while its sarcodic body consists of a continuous coil, with slight constrictions at intervals. The second stage consists in the opening-out of its spire, and in the division of its cavity at regular intervals by transverse septa, traversed by separate pores, exactly as in Peneroplis. The third stage is marked by the subdivision of the ' peneropline ' chambers into chamberlets, as in the early forms of Orbiculina. And the fourth consists in the exchange of the spiral for the cyclical plan of growth, which is characteristic of Orbitolites ; a circular disk of progressively increasing diameter being formed by the addition of successive annular zones around the entire periphery. This increase in diameter is not here accompanied (as it is in most of the other forms of the Orbitoline type) by a corresponding augmentation in thickness ; and as the extraordinary tenuity of these disks affords an easily recognisable and (as I believe) a constant differential character of the species, I proposed in 1870* to designate it Orbitolites tenuissima. ORBITOLITES TENUISSIMA. Carpenter, 1870. The disks of 0. tenuissima are usually almost perfectly flat (Plate 37, fig. 1), and exhibit a remarkable regularity of structure. The diameter of the largest complete specimen I have seen is not above 0'25 inch ; but it is obvious from the size and curvature of the fragments which the dredges frequently contained, that they must have belonged to disks whose diameter was at least 0'6 inch, these larger specimens having come to pieces in their rough removal from the soft and tranquil ooze on which they had previously lain. This fragility depends in part upon the extreme tenuity of the disks, their thickness rarely exceeding one three-hundredth of an inch ; and in part on the slightness of the connexion which (as I shall presently show) exists between the successive zones.t When either surface of the disk of 0. tenuissima is viewed by reflected light under a low magnifying power, its concentric zones are seen to be crossed by radial lines (Plate 37, figs. 1, 2) resembling those which pass between the septal bands * Proc. Roy. Soc., vol. 19, p. 176. f In Cyclodypeus, the marginal portions of the disk, though of even greater tenuity, have not by any means the same fragility ; partly because its vitreous shell-substance is much firmer than the porcellanous shell-substance of Orbitolites, and partly because a layer of it is usually continued from each new zone over the whole surface of the previously formed disk. See Pbil. Trans., 1856, p. 558. 554 DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. of Peneroplis. But when a portion of the disk is viewed under a higher power by transmitted light, -which, through the extreme tenuity of its superficial lamellae, brings its internal structure into distinct view (Plate 38, fig. 5), these lines are seen not to be mere surface-markings, as in Peneroplis, but to be the indications of internal shelly partitions, which divide each flattened annular chamber into a series of narrow chamberlets, resembling those which I formerly described as constituting the two superficial layers of the "complex" type of Orbitolites.* Here, however, these chamberlets form but a single plane, as in the " simple " type formerly described ; and the pores by which the last-formed annulus opens at the margin of the disk are arranged in single series (Plate 37, figs. 4, 5). It is worthy of note that these pores are not round, like those of ordinary Oi'bitolites, whether of the " simple " or of the " complex " type ;t but are more or less elongated in the plane of the disk a peculiarity obviously related to its extreme compression. Similar pores are seen upon the edge of any zone from which the zone external to it has been detached by fracture ; and it is obvious that they constitute the channels of commu- nication between the cavitary system of each zone and that of the zones internal and external to it ; while the marginal series brings the cavitary system of the peripheral zone (and, through it, that of every interior zone backwards to the spiroloculine " nucleus ") into relation with the surrounding medium. When a portion of the thin shelly lamella forming either surface of the disk has been removed by dilute acid, so as to lay open the cavity beneath (Plate 37, fig. 2), it is seen that each zone of chamberlets lies between two concentric rings of shell, a, a, b, 6 ; and that the radiating partitions, c, while springing from the inner shell- ring, do not extend to the outer, so that a continuous gallery is there left, into which all the chamberlets open at their peripheral extremities. And when we examine the disk by transmitted light (Plate 38, fig. 5), we see it to be from this gallery not from the chamberlets that the pores of the shell- ring which incloses it proceed. Whilst the structure of the concentric zones forming the peripheral portion of the disk thus corresponds in all its essential characters with that of the ordinary " simple " type described in my former Memoir, the structure of the central portion of the disk is altogether different. The spheroidal " primordial chamber " (Plate 38, fig. 3, a) is extremely minute, not exceeding 1-1 000th inch in diameter, and from this proceeds a compressed shelly tube, which forms a nautiloid spiral around it (Plate 38, figs. 3, 5) each successive turn slightly increasing in breadth, so as closely to resemble the first-formed part of the spire of Cornuspira. The continuity of its cavity, however, is interrupted, usually at about every two-thirds of a turn, by a thickening of its wall (Plate 38, fig. 3, b), which seems to have been formed as a sort of foreshadowing of a septum at each addition to its length ; and thus, as long as the growth of the shell proceeds upon the same plan, it is a ' spiroloculine ' Miliola. * Phil. Trans., 1856, p. 202, Plate V., fig. 6, c, c; Plate VII., fig. 12. t Ibid., Plate V., figs. 1 and 6, d, d. DR. W. B. CARPENTER ON ORBITOLITES TENTJISSIMA. 555 But after making from six to eight turns (the number varying in different individuals) the spire begins to open out in the horizontal plane (Plate 38, fig. 5, a) without any vertical enlargement, and a complete septum is formed at the next break, marking off the first principal chamber from the previously-formed spiral tube. This septum is traversed, as in Peneroplis, by a variable number (four in the specimen here figured) of passages, which would show themselves as pores upon its external surface ; but these, instead of opening into another single undivided chamber, lead into as many chamberlets, which are formed by the subdivision of Lhe next principal chamber, b, by radial partitions, exactly as in Orbiculina. This chamber, in the individual here figured, is not separated by a completely-formed septum from the succeeding chamber, c, and the latter is undivided save by a single radial partition ; but this is a mere individual variation, which is of interest, however, as showing that the subdivision of the chambers into chamberlets is a secondary, not a primitive formation. The septum which closes-in the chamber c is traversed by 13 pores, which open into as many chamberlets formed by the subdivision of the next principal chamber ; the separation of these chamberlets by radial partitions being complete for about four-fifths of the length of the chamber (that is, of the distance between its inner and its outer septum), but deficient for the outer fifth, so as to leave the continuous gallery d, d, into which all the chamberlets open at their outer ends. This chamber, it will be observed, extends itself on either side at d', d', so as to enclose a portion of the spiroloculine " nucleus ; " and this extension is still more marked in the next chamber, whose two alee, ei ', e , reach the ends of the transverse diameter of the original spire. The septum which separates this chamber from the preceding has the number of its pores increased to 30 ; and these open outwards into as many chamberlets in the next-formed chamber. As new chambers are successively added, the backward extension of their alse is carried further and further, until (in the individual here figured, Plate 37, fig. 1) those of the ninth chamber meet at the back of the spiroloculine " nucleus," so as to enclose it all round, and the tenth chamber forms a complete ring of chamberlets, whose derivation from the undivided chamber of the ' peneropline ' type is made obvious by the previous transition. With each increase in the length of the septal plane, there is a proportionate increase in the number of pores by which it is traversed, the distance between them having a very uniform average ; and the number of these pores determines the number of chamberlets in the next annulus, which has thus no definite relation to that of the chamberlets in either of the last-formed or in the subsequently-formed annulus. The breadth of the zones (and, consequently, the length of their chamberlets) has a range of variation from l-180th to l-80th of an inch, its general average being l-120th inch ; so that a disk having a diameter of 0'6 inch (or a radius of 0'3) would be made up of about forty such concentric zones. A very narrow zone is occasionally seen to intervene between two zones of ordinary breadth ; but, as I have always found this to originate 556 DE. W. B. CABPENTEE ON OEBITOLITES TENUISS1MA. in a fractured portion of the preceding zone which then formed the margin of the disk, it would seem to have been a special reparative addition. The whole cavitaiy system, from the primordial chamber to the marginal annular gallery, is occupied by a continuous sarcodic body of a dark olive-green hue (Plate 38, fig. 1). Although this body may be said to consist at any one moment of a multitude of sub-segments, connected together by annular and radiating stolon-processes, yet, from what we know of its semi-fluid condition in the living animal, we may pretty confidently surmise that this subdivision is by no means permanent, but that an interchange is continually taking place between the protoplasmic contents of the inner and the outer portions of the cavitary system, so that what occupies the central spire at any moment may be transferred in no long time to the marginal annulus, and vice versd. The extreme tenuity of this sarcodic body, and the transparence of the shelly laminae that invest it, have enabled me very distinctly to recognise, by light trans- mitted through the disk, the presence of nucleus-like bodies (Plate 38, fig. 2) of about l-1750th of an inch in diameter, imbedded in its substance. As might be expected from the consideration just stated, these corpuscles are very irregularly distributed. In the specimen here figured (Plate 38, fig. 1), two of the outer half- whorls of the ' spiroloculine ' centre (shown on a larger scale at b, b, b', b', fig. 4), are crowded with them ; while in a single chamberlet, c, of one of the interior annuli, there are as many as five. Elsewhere, on the other hand, they present themselves with less frequency, only one or two occurring in any single chamberlet (d, d, d), and a large proportion of the chamberlets being entirely destitute of them. The finding of these corpuscles in the highly composite sarcode-body of Orbitolites is an interesting extension of the discovery, of Dr. E. HERTWIG, of corpuscles regarded by him as nuclei, in what I long since characterised as the " reticularian " type of KHIZOPODA, of which the ordinary FORAMINIFERA are the testaceous forms. This discovery was first made in the fresh- water Monothalamous Mikrogromia,* and subsequently extended by him to various marine Polythalamia, such as Spiroloculina, Globigerina, and RotaliaJ and by F. E. SCHULTZE to Quinqueloculina, Lagena, Polystomella, and Planorbulina.% What is the function of these corpuscles in that indefinite extension of the protoplasmic body, and the multiplication of its segments, which is so remarkable a character of this type, is not yet apparent ; but that they do not become the centres of distinct cells separated from each other by any limiting membrane, or even of permanent segments or sub-segments, may be regarded as certain. If the nuclear character of these corpuscles be admitted, the entire composite organism thus seems to present a most interesting link of connexion between the unicellular and multiceUular types ; the absolute continuity of its protoplasmic substance entitling it to rank with * Archiv fur Mikrosk. Anat., Bd. x., Supplementheft (1874) p. 1. t Jenaische Zeitschrift, Bd. x., 1876, p. 41, &c. J Archiv fiir Mikrosk. Anat., Bd. xiii., 1877, p. 9. DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 557 the former, whilst in the multiplication of its nuclei it obviously tends towards the latter. The growth of this beautiful organism doubtless takes place after the completion of the first annulus, in the manner described in my former Memoir ( 35). The sarcodic body, when enlarged by the nutriment it has appropriated, will project itself through the marginal pores, in quantity sufficient to form, by the coalescence of its separate protrusions, a continuous belt of sarcode ; and in the substance of this a set of radial calcareous partitions will be deposited, commencing between the pores of the margin of the previous shell-ring, while two horizontal lamellae are formed on the superficial planes, to constitute, as it were, the floor and ceiling of the new circle of chamber] ets. These horizontal lamellae overlap but very slightly the margin of the previous annulus (Plate 37, fig. 3, a, a) ; and their adhesion to it is generally so weak that the annul! readily come apart. As the subdivision of the annulus into chamberlets does not extend to its outer portion, a passage is left (seen in vertical section at b, b, b, fig. 3), which is occupied by a continuous ring of sarcode, as shown in Plate 38, fig. 1 ; and from this ring proceed the stolons which pass outwards through the pores of the septum that closes it in. The homogeneousness of the protoplasmic substance by which the entire cavitary system is occupied, is shown (as in the types formerly described) by the completeness with which the effects of injuries are repaired, and the plan of the original fabric restored (see Phil. Trans., 1856, Plate VIII., figs. 4-9). Not only is the loss of any part of the disk repaired by the formation of a new and continuous annulus along the broken as well as the unbroken margin, so that the next and all succeeding zones follow the new contour (as shown in Plate 37, fig. 6) ; but a new and entire annulus of chamberlets may form itself around the whole circumference of a mere marginal fragment (tig. 7), by the enclosure of which in a subsequent succession of annuli, the discoidal form characteristic of the type is completely and characteristically restored. Owing to the transparence of these attenuated disks, I have been able to assure myself that every part of the margin of this fragment, whether broken or unbroken, peripheral, central, or lateral, has contributed to the formation of the first new complete annulus, by which the foundation was laid of the subsequent regular series of concentric zones ; thus clearly indicating that a sarcodic extension took place from every chamberlet laid open by the fracture, as well as from the normal pores of the last-formed septal plane, and that these extensions coalesced to form a continuous ring, as in the formation of the ordinary succession of concentric annuli. This perfect reproduction of a form of peculiar regularity, in a type of animal organisa- tion so low that its body-substance does not show any advance upon the primitive proto- plasmic condition, is, in itself, a matter of great interest. But the interest is much enhanced by the consideration that this organism begins life, and forms its first shelly envelope, upon a plan altogether different ; exchanging this for its later mode of growth, by a transition so rapid as to manifest the almost sudden attainment of a MDCCCLXXXIII. 4 C 558 DE. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. much more specialised character. And when this transition has been once made, there appears no disposition whatever, in the reparation of injuries, to a reversion to the earlier plan. Now, this is a "pregnant instance" of the following "law of formation," sagaciously laid down long since by Sir JAMES PAGET : " When, in an adult animal, a part is reproduced after injury or removal, it is made in conformity, not with that condition which was proper to it when it was first formed, or in its infantile life, but with that which is proper according to the time of life in which it is reproduced ; proper, because like that which the same part had, at the same time of life, in members of former generations." And the study of this humble Orbitolite will be found, not only in this, but in other particulars, to justify the profound remark made by the same philosophic Pathologist,* long before the promulgation of the doctrine of "evolution," that, " if we are ever to escape from the obscurities and uncertainties of our art, it must be through the study of those highest laws of our science which are expressed in the simplest terms in the lives of the lowest orders of creation." Geographical, Bathy metrical, and Geological distribution. So far as is at present known, Orbitolites tenuissima inhabits only the North Atlantic Ocean and the seas in communication with it. The first complete specimens were obtained in the " Porcupine " dredgings of 1869, at depths of from 630 to 1,443 fathoms, between the north-west of Ireland and Eockall Bank. In the " Porcupine " expedition of 1870, however, it was brought up from a bottom of only 64 fathoms in Setubal Bay, on the coast of Portugal, and afterwards from a shallow bottom within the Mediterranean, near Carthagena. That it is an inhabitant of other parts of the Mediterranean I then inferred from the fact that I had detected fragments of it in the Foraminiferal dredgings, made at 250 fathoms by EDWARD FORBES and Lieut, (now Admiral) SPRATT in the JSgean, in 1842; and it is stated by Dr. J. GWYN JEFFREYS, in his " Report on the Biology of the ' Valorous ' Cruise," that it has been dredged by the Marquis DU MONTEROSATO at from 100 to 200 fathoms' depth, off the coast of Sicily. That it might extend far to the north, would be expected from its capability of bearing the low temperature of 37 Fahr., which prevails over the deep bottom from which it was first brought up ; and this expectation was verified by its presenting itself in one of the "Valorous" dredgings in Baffin's Bay (lat. 62 6' N., depth 1,350 fathoms, temperature 34 6' Fahr.), as well as at two stations in the North Atlantic, No. 12, depth 1,450 fathoms, and No. 13, depth 690 fathoms, both in the parallel of 56. It has been only once brought up, however, in the "Challenger" expedition, viz,, at Station 44, off Cape Hatteras, from a bottom of 1,700 fathoms' depth, over which creeps (there is strong reason to believe) an underflow of cold water from the Arctic basin. Several specimens have (I am informed) been since found in a * " Lectures on Surgical Pathology," 1849 ; Lect. VII. General Considerations on Repair and Reproduction. DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 559 dredging taken by the French exploring-ship " Travailleur " in the Bay of Biscay (Fosse de Cap Breton), at a depth of 1,200 fathoms. It would seem, therefore, that Orbitolites tenuissima has its proper home on the sea- bottom of the deeper parts of the North Atlantic, where the temperature ranges from 37 to 35 Fahr. ; but that it also is capable of living, not only in much shallower, but also in much warmer, waters. For the temperature of the Mediter- ranean and ^Egean, even at depths below 100 fathoms, is never less than 54, while on the shallow bottom of Setubal Bay, and on the shore-slope near Carthagena, the summer temperature must be considerably higher. Looking to the singular retention, in this beautiful Orbitoline, of the Milioline type from which its derivation may now be confidently affirmed, the probability seems strong that it was a very early form ; and if identical with COSTA'S Pavonina italica* as the imperfect account given by him of that type would seem to indicate, it probably inhabited the Mediterranean during the greater part of the Tertiary period. Its persistence in the abyssal depths of the North Atlantic harmonizes well with the idea of its antiquity ; those depths having been found, by the recent exploration of them, to be inhabited by many " survivals " of the Cretaceous and even earlier Faunae. It may be remarked, finally, that the considerable diameter attained by these very fragile discs, seems a proof of the extreme tranquillity of the deep-sea bottom ; since they could not otherwise have gone on growing and extending themselves, without showing more frequent marks of injury and reparation than I have observed in them. Relation to other Orbitoline Types. Having been requested by the late Sir C. WYVILLE THOMSON to prepare a Report on the Orbitolites collected in the " Challenger " Expedition, I have carefully studied the remarkable gatherings made of them on and near the summit of the Fiji reef, and also at a depth of 18 fathoms on its slope. The result of that examination now enables me to indicate with great probability the successive stages of the evolution of that highly specialised " complex " type, the derivation of which from a Milioline ancestry would have seemed but for the completeness of the series of intermediate forms almost inconceivable. And I can now also mark out, with more distinctness than formerly, the types of this Genus, which, in virtue of their constancy and definiteness, are entitled to rank as distinct species. The first of these is the O. marginalis of Lamarck, known to him only by small Mediterranean specimens of no more than two millims. (about 0'08 inch) in diameter, but attaining on the Fijian reef a diameter of 0'2 inch, and presenting a much more characteristic aspect than is discernible in the dwarfed Mediterranean form. The well- developed "cycloline" disks of this beautiful form of the "simple" type (fig. I., 1), * See his " Paleontologia del Regiio di Napoli," part ii., in 'Afcti dell' Accad. Pontan,' vol. vii., p. 178, plate xvi., figs. 26-28. 4 c 2 560 DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. which I formerly* differentiated only by the singleness of its row of marginal pores (fig. I., 3), I now find to be uniformly characterised by the marked eccentricity of their primordial chamber, and by the spiral direction of their early growth (fig. L, 2), which I formerly supposed to be only occasional variations. In fact, the first formed portion of these disks, like that of a young Orbiculina adunca (loc. cit., Plate XXVIII., fig. 2), exactly resembles what a Peneropline shell would be, if its chambers, as they widen out, were to undergo division into chamberlets ; thus corresponding in every essential particular with the " orbiculine " stage of O. tenuissima. But while we have seen that this stage, in the last-named species, is preceded by a spiroloculine coil, representing a true " milioline " stage, it has no other predecessor in 0. marginalis than what I formerly designated as the " nucleus," consisting of a flask-shaped " primordial chamber," from the neck of which proceeds a " circumambient chamber " that passes Fig. I. Orliitolites marginalis. almost completely round it (as in fig. II., 3). The morphological import of this arrangement becomes clearer when we compare the sarcodic bodies of the two types ; for it is then obvious that the "circumambient segment," which springs from the " primordial segment," and then, after making a single coil around it, gives off the first " peneropline " segment, really represents the multiple spiroloculine coil of 0. tenuis- sima ; this early generalised " milioline " stage being (as it were) abbreviated with the advance towards specialisation, as we see in numberless cases elsewhere. The "Challenger" collection especially that of the 18 fathoms' dredging includes a very large number of thin flat disks, attaining a diameter of about 0'32 inch, whose surface often presents rather an " engine-turned " than an annular aspect, and which are specially characterised by the possession of a complete double row of marginal Phil. Trans., 1856, p. 215, Plate VII., fig. 14. DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 561 pores (fig. II., 2). Both these peculiarities were noticed in my former Memoir (pp. 215, 221), but were treated as merely varietal modifications. I now find, however, that they accompany one another very constantly ; and that the type is so well differentiated by them as to be fully entitled to rank as a distinct species, which I designate O. duplex.* Notwithstanding the difference in the surface-aspect of its disks, and the doubling of their marginal pores, the sarcodic body of this species conforms in every essential particular to that of the preceding. For each of its concentric annul i consists of a single cord (fig. II., 4, c c'), that passes through a Fig. ll.Orlitolites duplex. continuous circular gallery in the median plane of the disk, and carries a double series of columnar sub-segments (a a, b b'), which occupy chamberlets (fig. II., 1) that extend in vertical series to the two surfaces of the disk. But each annular cord, instead of giving off (as in 0. marginalis) a single stolon-process to initiate a sub-segment of the succeeding annulus, gives off two such processes between each pair of its own sub-segments (fig. II., 4, d d, d'd'}; and these have separate passages through the septal plane one above and the other below the annular canal, as shown in fig. II., 2, * This species, as intimated in my former Memoir, appears to be the type described by Prof. EHRENBERG (Abliandl. der Konig. Akad. der Wissenschaften zu Berlin, 1839) as a BRYOZOON, under the designation AmpMsorus Hemprichii. As his conception of the generic characters of this type was fundamentally erroneous, and as he gave no diagnosis of the single species he created, I have not thought it necessary to preserve his specific designation. 562 DR. W. B. CARPENTER ON ORBITOLTTES TENUISSIMA. opening on its external side in a double series. This arrangement, as will presently appear, is the first step in the evolution of the " complex " form of the Orbitoline type. Another advance upon 0. marginalia is seen in the more rapid approach of O. duplex to the cyclical plan, shown in the abbreviation of the early spiral stage. For the "nucleus" of O. duplex has but a slight eccentricity, and its circumambient segment (fig. II., 3, b, l>), instead of putting forth but a single stolon-process, gives off several ;* so that, as each of these originates a new sub-segment, a crescentic row of sub-segments is at once constituted. The row formed next in succession to this almost entirely encircles the milioline nucleus, and the third row generally completes the annulus, all further increase in the disk taking place on the cyclical plan. In O. duplex, therefore, we have such an abbreviation, not only of the " milioline " but also of the " orbiculine " stage, that the proper " orbitoline " type is attained at a relatively earlier period. Fig. lIl.OrMtoJitea complcmata. We come, lastly, to that "complex" form the 0. complanata of LAMARCK in which the special peculiarities of the Orbitoline type are most fully displayed. Its disks (fig. III., 4) attain not only a much larger diameter, but a relatively greater thickness, than those of either of the " simple " species ; the annulations which mark their surfaces are as complete in their central as in their peripheral portions ; their superficial * Phil. Trans., 1856, Plate IV., fig. 13. DR. W. B. CARPENTER ON ORBITOL1TES TENUISSIMA. 563 chamberlets have an elongated form (fig. III., 3, a), and their margins exhibit, even in the smallest (or youngest) specimens, multiple series of pores (fig. III., 1), in- dicative of that complicated arrangement of the cavitary system which I described minutely in my former Memoir. The meaning of that arrangement is best understood by an examination of the sarcodic body left after the decalcification of the disks, which are modelled, as it were, upon it. The accompanying representation (copied from Plate IV., fig. 4 of my Fig. IV. Portion of sarcodic body of Orbitolites complanata : a a', b b', the upper and lower annuli of two concentric zones ; c c, the upper layer of superficial sub-segments, and d d, the lower layer, connected with the annular cords of both zones ; e e and e'e', intermediate columnar sub-segments of the two zones. former Memoir) shows two annular cords, a a, b b', in each annular zone, instead of the single cord of 0. duplex ; and between these two cords is interposed a series of columnar sub-segments, e e, e' e, whose bases and summits (so to speak) are brought into continuity by them. It is of the interposed shell-substance that lodges these columnar sub-segments, that the thickness of the disk (fig. III., 2) is chiefly made up ; and this is obviously in relation with the length of the columns. Between each annular cord and the nearest surface of the disk, is a series of sub-segments, c c, d d, which occupy the elongated chamberlets whose partitions are marked externally by radial lines that cross the several annuli (fig. III., 3, a), as in O. tenuissima. These partitions, however, being complete, the chamberlets have no lateral communication with each other; neither do they communicate by means of radial passages with those of the annuli internal and external to them. But each has a passage at either end through its own floor, which allows a stolon-process to pass from the sub-segment which it lodges to the annular cord beneath ; each sub-segment being, therefore, in connexion with the two annular cords, and forming, as it were, a bridge between one and another, as shown in fig. IV. Except through the intermediation of these sub-segments, the annular cords of the successive zones have no connexion with each other ; but the intermediate columnar sub-segments of each annulus communicate with those of the next by 564 DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. oblique stolon-processes, that pass off alternately at regular intervals from the two sides of each column, traversing the annular septa ; and the orifices of the passages in the last-formed septum, through which these stolon-processes extend themselves outwards, are seen as multiple series of pores on the margin of the disk (fig. III., 1). The vertical section of the calcareous disk given in fig. III., 2, shows the separation of the two superficial planes of chamberlets by the interposition of the shelly fabric that gives lodgment to the intermediate sarcodic columns ; while at 3 is shown diagrammatically, on a larger scale, the cavitary system of the disk, with the commu- nication between its several parts. At a are seen the chamberlets of the superficial planes, which are completely closed in when not abraded ; and these are shown in vertical section, above and below, at b, while at c are seen their floors, each having a pore at either end, which communicates with the annular canal beneath. The annular canals are seen at d in vertical section, and at d' and d" as laid open in horizontal section ; the former showing how they cross the tops of the cylindrical chamberlets of the intermediate stratum, and the latter (taken a little nearer the surface) showing the manner in which they open into the pores leading to the superficial chamberlets. In the lower part of the figure, the intermediate stratum is traversed by two horizontal sections in slightly different planes, cutting across the cylindrical chamberlets, and showing the two series of oblique stolon-passages by which the chamberlets of successive annuli communicate with each other. The nuclear mass which occupies the centre of the disk consists, as in 0. duplex, of a " primordial segment," surrounded by a " circumambient segment," and this last (fig. V., 6, 6) puts forth a set of stolon-processes from its entire periphery, each of Fier. V. which gives origin to a columnar sub-segment ; so that a complete annulus is at once constituted, thus establishing the cyclical plan of growth from the very first. The collection of specimens of 0. complanata made on the Fiji reef contains disks of all sizes ranging from 0'04 inch to nearly 1 inch; and even in the smallest of them, whose nucleus is surrounded by only two or three annuli, the immediate assumption of the completed plan is marked by the multiplicity of the series of marginal pores. But while this may, I think, be unquestionably regarded as the typical condition of the species, the collection also includes an abundance of disks whose peripheral portion is characteristically " complex," whilst their central portion is no less characteristically " simple ; " the passage from the one plan of growth to the DR. W. B. CARPENTEE ON ORBITOLITES TENUISSIMA. 565 other taking place at no fixed stage, but being made sometimes earlier, sometimes later, for the most part gradually, but sometimes abruptly, as I indicated in my former Memoir ( 57, 58). I can now trace out more distinctly than before the successive phases of this transition ; and can show how exactly the fundamental characters of 0. marginalis and 0. duplex are reproduced in what may be called the childhood and youth of those " sub-typical " examples of O. complanata, which, instead of beginning life on the " complex " plan, only attain the more elevated type in adult age. These phases are exhibited in the following vertical section (fig. VI.) taken in the radial direction, Fig. VI. d' which, though representing them somewhat diagramatically, is true to nature in every essential particular. Of the successive zones traversed by the sectional plane between the circumambient chamber, c c', which passes round the primordial chamber, p, the first five, m, m\ m?, m s , m*, are formed exactly on the type of those of 0. marginalis ; each chamberlet being connected laterally with the other chamberlets of its own zone by a single annular canal, ac, and with the chamberlets of the zones internal and external to it by the radial stolon-passages, r r. These are succeeded by three zones, d, d l , cP, formed upon the " duplex " plan ; each chamberlet having, as in the preceding case, but a single annular canal, ac, but communicating with the chamberlet of the annulus external to it by two oblique radial stolon -passages, as shown at r. Thus, then, if the growth of this disk had been checked at the fifth zone, m*, its margin would have presented the single row of pores characteristic of 0. marginalis ; and if at the eighth zone, d 2 , it would have shown the double row characteristic of 0. duplex. But in the next zone, e, the annular canals ac, a'c', are duplicated, each of them sending off a stolon-passage into the next annulus. In this and the succeeding zone, e, however, there is no separation between the superficial portions, s s', of the chamberlets, and their median portions, m m ; and this continuity, here transitory, shows itself as the typical character of the chamberlets of the fossil 0. complanata of the Paris basin. But in the existing 0. complanata, a separation comes to be effected, as shown in the succeeding zones, f, f l , f~, f 3 , by horizontal extensions of the septa i i, that are interposed between the median portions of the chamberlets, so as to form the floors of the superficial layers ; while at the same time there is a shifting of their relative positions, so that the superficial chamberlets, s s, s' s', instead of lying over or under the median portions, m m, alternate with them, and are entirely MDCCCLXXX1IT. 4 566 DE. W. B. CARPENTER ON ORBITOLITES TENUISSlMA. cut off from any other communication with them than that which is afforded by the annular canals, with which each superficial chamberlet communicates at either end, by a passage which thus traced out is seen to be homologous with one of the double radial stolon-passages of 0. duplex, and therefore with the single radial passage of O. marginalis. The septa, i i, which divide the median portions, m m, of the successive annuli, are traversed by numerous passages, which, from the lateral obliquity of their direction (fig. III., B,f,f), scarcely show themselves in a radial sec- tion, although they debouch at the edge of the last annulus as marginal pores, mp. Notwithstanding this progressive complication in the structure of the shelly disks, there is no appearance of any corresponding specialisation in the character of the sarcode body: that of the typically " complex " form showing no other advance upon the very simplest, than is marked by the duplication of the sarcodic annuli, by the separa- tion of the superficial from the intermediate columnar sub-segments, and by the multi- plication of the oblique stolon-processes which connect these last with each other, this multiplication being obviously in relation with the increasing length of the interposed columns, which shows itself in the thickening of the disk. The most marked increase in the complication of the animal body obviously consists in the duplication of the sarcodic annuli ; and this may be readily conceived as a longitudinal splitting of each cord into two, with a persistence of adhesion at intervals, so that the two semi-annuli, when carried apart from one another by the interposition of the intermediate stratum, remain connected by the vertical sarcodic columns which traverse that stratum. The sub-segments which occupy the upper and under layers of surface -chamberlets are clearly shown, by their relation to the sarcodic annuli, not to be new productions, but to be homologous with the upper and under halves of the sub-segments that occupy the columnar chamberlets of the " simple " type ; that homology, however, being so masked in the typically " complex " form by the displacement they have undergone, that it could not have been certainly recognised, but for the occurrence of those sub- typical forms which enable the passage from the most "simple" to the most "com- plex " to be contimiously traced-out. I have been unable, after the most careful examination of the sarcodic bodies of 0. duplex and 0. complanata, to discover any indication that this progressive complica- tion in the disposition of their parts, is accompanied by any such structural modification as might lead to the suspicion of differentiation of function. On the contrary, I find their substance to be everywhere of the same elementary character, consisting of a homogeneous protoplasm, that contains a large number of spherules of from 6"^ooth to aoouih of an inch in diameter, sometimes crowded closely together, in other instances more dispersed, as shown in fig. 3, Plate IV. of my former Memoir (Phil. Trans., 1856). These spherules, when subjected to pressure, break up into a number of pellucid corpuscles, which are usually of from 7-5 ^ro^th to aooooth of an inch in diameter. The absence of these spherules is a marked feature of difference in the protoplasmic DE. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 567 body of 0. tenuissima; on the other hand, I have not met in the higher types with those nuclear (?) bodies which I have recognised in the abyssal species (see p. 556). The homogeneousness of the entire sarcodic body, even in the largest and most com- plicated forms of O. complanata, appears to be further indicated by the fact, that in specimens taken alive and preserved in spirit, the peripheral portion of the cavitary system is invariably found empty ; the sarcodic body, corrugated by the action of the spirit, being drawn together towards the central portion of the disk through very narrow passages of communication, which could only happen with a substance of which every part is free to move upon every other. Looking, also, to the manner in which the entire organism receives its nourishment through the marginal pores, and to the entire absence of any special means for the distribution of that nourishment, I think it may be fairly assumed that such a protoplasmic circulation goes on throughout life, as must produce a continual change in the substance of every individual sub-segment. Additional evidence of this homogeneousness is afforded by the two following facts : first, that in specimens which live under conditions peculiarly favourable to their enlargement, out-growths of irregular shape, but always possessing a regular internal structure, are put forth from any part of the disk (see " Challenger " Report, Plate VII.) ; and second, that, as in 0. tenuissima, every part seems equally capable of reproducing the entire disk on its characteristic plan. Evolutionary History of the Orbitoline Type. Thus by the combined study of O. tenuissima and of sub-typical examples of 0. complanata, we are enabled to work out the whole evolutionary history of the Orbi- toline type, from its simplest to its most complex form. For there can, I think, be no reasonable doubt, that the succession here presented to us in the consecutive phases of two lives, has been the genetic history of this type ; w T hich, originating in the simplest "jelly-speck " that could form a shelly chamber, first assumed the form of a spirally- coiled undivided tube (Cornuspira, fig. VII., 1) ; then of a spire interrupted at inter- vals by imperfect partitions (Spiroloculina , 2) ; then of a flattened spire crossed by complete septa traversed by stolon-passages (Peneroplis, 3) ; then of a progressively widening spire, whose chambers are divided into chamberlets (Orbiculina, 4) ; then of a chamberletted disk of one storey, commencing as an orbiculine spire, but subsequently increasing by annular additions (Orbitolites tenuissima and 0. marginalis, 5) ; then of a chamberletted disk, whose origin still shows in its slight eccentricity a trace of the primordial spire, and whose single storey has, so to speak, two rows of windows (Orbitolites duplex, 6) ; and lastly, of a " complex " disk, whose growth is cyclical from the beginning, and whose upper and lower superficial planes are separated by the interposition of an intermediate columnar structure between the duplicated annular stolons (Orbitolites complanata, 7). This last would seem to be the culmina- tion of the type, which, while attaining a considerable size, has never shown, so far 4 D 2 5G8 DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. as is at present known, any tendency to pass into a higher form. Indeed, the typical forms of the existing 0. complanata are in one small particular (as already shown, p. 565) more specialised than the fossil forms that were so remarkably abund- ant in the Middle Tertiary epoch. It is a remarkable feature of this case, however, that all the forms through which the highest Orbitoline type is thus shown to have passed, continue to hold their ground at the present time, as the characteristic representatives of less specialised groups. There being every reason to regard Cornuspirce, Peneroplides, and Orbiculince as distinct races, propagating themselves genetically without any essential modifi- Fig. VIT. cation, it can scarcely be supposed that every one of them is a " potential " 0. tenuis- sima. So, again, as we find O. marginalia and 0. duplex living and propagating under the very same conditions as 0. complanata, I cannot regard these " simple " forms of the Orbitoline type, each of which has its characteristic plan of structure and limit of growth, as potentially " complex ; " notwithstanding the exact repetition of their plans in the early stages of certain examples of the higher type. For I have never observed in the largest and best developed examples of 0. marginalis and O. duplex the least tendency to assume the " complex " form ; on the other hand, I have frequently found their last formed annuli deficient in internal partitions, as if their productive power had exhausted itself. It would seem, therefore, more just to DE. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 569 regard those sub-typical examples of 0. complanata, which exhibit the transition I have described from the " simple " to the "complex " plan of structure, not as advanced forms of either of the two " simple " species, but as retarded forms of the highest ; and a clue to the conditions of that retardation can, I think, be found in the marked inferiority I have invariably observed in the size of the original nuclear mass of these individuals. I drew attention in my former Memoir ( 44) to the remarkable range of dimension which this mass exhibits, when a considerable number of specimens are examined ; and showed that the cavity of the " primordial " and " circumambient " chambers in one individual must have been more than a hundred times as large as that of another. Now the result of the far more extended comparison of specimens which the " Challenger " collection has enabled me to make, is, that while the " nucleus " of the typical 0. complanata (in which the cyclical plan of growth, and the "complex" structure, show themselves from the very first) is always many times larger than that of either 0. marginalia or 0. duplex, the " nucleus " of its sub-typical forms always bears a very close accordance in size to that of 0. duplex, which it resembles also in the one-sided pullulation of the first sub-segments from the circumambient segment, rendering the earlier zones more or less incomplete, and the position of the " nucleus " slightly eccentric. Whether these forms genetically propagate themselves as a race, perpetuating an earlier stage of the evolution of the perfected type, or are merely individuals which have begun life as " starvelings " that do not inherit the characteristic vigour of the type, I have no adequate ground for even surmising ; being only able to affirm this, that as there is no kind of constancy in the stage of growth at which the " simple " plan gives place to the " complex," there is nothing to justify a specific differ- entiation of this sub-typical variety. That its peculiarity may depend upon conditions less favourable to the full development of the type, seems to be indicated by the fact that, whilst the largest and most typical specimens of 0. complanata were found in the rock-pools on the summit of the Fiji reef, where they would have the highest tempera- ture and the greatest abundance of food, the sub-typical specimens presented themselves chiefly in the collection made by the dredge at 18 fathoms' depth. Tlieory of Descent. I propose, in the last place, briefly to examine the bearing of the remarkable case of " descent with modification," which I have thus detailed, upon the general " Theory of Descent " and of the " Origin of Species." Those who find in " natural selection " or the " survival of the fittest " an all-sufficient explanation of the " origin of species," seem to have entirely forgotten that before " natural selection " can operate, there must be a range of varietal forms to select from ; and that the fundamental question is (as Mr. DARWIN himself clearly saw, at any rate in his later years), what gives rise to variations ? No exercise of " natural selection " could produce the successive changes presented in the evolu,- 570 DR. W. B. CARPENTER ON ORBITOLTTES TENUISSIMA. tionary history of the typical Orbitolites, from Cornuspira to Spiroloculina, from Spiroloculina to Peneroplis, from Peneroplis to Orbiculina, from Orbiculina to the " simple " forms of Orbitolites, and from the " simple " to the " complex " forms of the last-named type. And as all these earlier forms still flourish under conditions which (so far as can be ascertained) are precisely the same, there is no ground to believe that any one of them is better fitted to survive than another. They all imbibe their nourishment in the same mode ; and no one type has more power of going in search of it than another. That they are all dependent on essentially the same conditions of temperature and depth of water, is shown by their occurrence in the same marine areas. That they all equally serve as food to larger Marine Animals, can scarcely be doubted ; and it is hardly conceivable that any of their devourers would discriminate (for example) between the disks of a large 0. marginalis, a middle-sized 0. duplex, and a small 0. complanata, which even the trained eye of the Naturalist cannot distinguish without the assistance of a magnifying-glass. To me, therefore, it appears that the doctrine of " natural selection " can give no account of either the origin or the perpetuation of those several types of Foraminiferal structure which form the ascending series that culminates in Orbitolites complanata. On the other hand, there seems traceable throughout that series a plan so definite and obvious, as to exclude the notion of " casual " or " aimless " variation. Between the simple spirally-coiled sarcodic cord of a young Cornuspira, and the discoidal body of an Orbitolite, with its thousands of sub-segments disposed with the most perfect symmetry, and connected together in most regiilar and uniform modes, who (in the absence of the intervening links) would have suspected any genetic relation who would have ventured to construct a pedigree ? And yet we find the gradations from the one to the other to be not only most complete, but often significant of further progress ; many of the changes being such as seem to have no meaning except as anticipations of greater changes to come. Thus, the slight constrictions that show themselves in the first spiral coil of 0. tenuissima (Plate 38, fig. 3) are what constitute the essential difference between the spire of Cornuspira and that of Spiroloculina ; marking an imperfect septal division of the spire into chambers, which cannot be conceived to affect in any way the physiological condition of the contained animal, but which foreshadows the complete septal division that marks the assumption of the Peneropline stage. Again, the incipient widening-out of the body, previously to the formation of the first complete septum, prepares the way for that great lateral exten- sion which characterises the next or Orbiculine stage ; this extension being obviously related, on the one hand, to the division of the chamber-segments of the body into chamberletted sub-segments, and, on the other, to the extension of the zonal chambers round the " nucleus," so as to complete them into annuli, from which all subsequent increase shall take place on the cyclical plan. In 0. marginalis, the first spiral stage is abbreviated by the drawing-together DR. W. B. CARPENTER ON ORBITOLITES TENtflSSIMA. (as it were) of the " spiroloculine " coil into a single Milioline turn of greater thickness ; but the Orbiculine or second spiral stage is fully retained. In 0. duplex, the abbreviated Milioline centre is still retained, but the succeeding Orbiculine spiral is almost entirely dropped out, quickly giving place to the cyclical plan. And in the typical O. complanata the Milioline centre is immediately surrounded by a complete annulus, so that nothing remains of the original spire save the one turn of the circumambient segment. So, in the passage from the " simple " to the " complex " type, we have a remarkable anticipatory step in 0. duplex, which can scarcely be supposed itself to derive any advantage from the substitution of a double for a single row of communications between the annuli, since 0. marginalis nourishes equally well with its single row ; but which forms, so to speak, a stepping-stone to a higher grade. Everything in this history, then, shows a well-marked progressive tendency along a definite line towards a highly specialised type of structure in the Calcareous fabric ; and this without any corresponding departure from the original homogeneity of the Animal body which forms that fabric. And as being, so far as I know, altogether unique in these peculiarities, I venture to offer this study of a humble protoplasmic organism, brought up from an ocean-depth of nearly two miles, to the consideration of those who believe with Sir JAMES PAGET, that " the highest laws of our [biological] science are expressed in the simplest terms in the lives of the lowest orders of Creation." EXPLANATIONS OF PLATES. PLATE 37. Structure of Calcareous Disk of OrTjitolites tenuissima. Fig. 1. Surface of young disk, showing its eccentric spiroloculine "nucleus," giving origin to successive zones of orbiculine chamberlets, which gradually increase in breadth with the opening-out of the spire, until they extend completely round the nucleus ; after which the successive additions are made on the cyclical plan, as concentric annuli. Magnified 25 diameters. Fig. 2. A portion of three peripheral annuli, enlarged to 64 diameters, and partially laid open by the removal of the superficial lamella, so as to show the two annular septa, aa, bb, the chamberlets, c, separated by radial partitions, and the annular gallery, d, into which all the chamberlets open at their peripheral extremities. Fig. 3. Vertical section of three annuli of the disk, taken in the radial direction, so as to traverse the chamberlets lengthways ; a, a, junctions of two annuli, with 572 DE. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. the annuli external to them ; b, b, b, annular galleries traversing the septa between the chamberlets. At a, a, are seen the openings through which the sarcodic cords that occupy the annular galleries send radial extensions into the chamberlets of the succeeding annuli. Magnified 64 diameters. Fig. 4. Internal aspect of a small portion of an annulus detached by fracture ; showing the entrances to the chamberlets of that annulus through the septal plane. Magnified 64 diameters. Fig. 5. External or peripheral aspect of a portion of a marginal annulus, showing the passages through its septal plane, as marginal pores elongated in the plane of the disk. Magnified 64 diameters. Fig. 6. Portion of a disk, whose remainder, with the " nucleus," has been lost by injury previously to the formation of the last two annuli, which have extended themselves along the fractured margin, and into the nuclear space. Magnified 15 diameters. Fig. 7. Incipient production of an entirely new disk, with regularly concentric annuli, from a fragment of the peripheral portion of an old one. Magnified 15 diameters. PLATE 38. Structure of Sarcodic Body and Calcareous Disk of Orbitolites tenuissima. Fig. 1. Sarcodic body of the central portion of the disk; showing the primordial seg- ment giving off the spiroloculine coil, the sixth turn of which, a, begins to open out into a peneropline form, afterwards becoming divided into rows of orbiculine sub-segments, which are connected together laterally by the con- tinuity of the sarcodic body through the gallery at the outer end of each row, and radially by the stolon-processes that pass through the septal pas- sages, from the gallery of the inner row into the chamberlets of the outer. Nuclear (?) corpuscles are seen irregularly distributed through the sarcodic substance. Magnified 75 diameters. Fig. 2. Nuclear (?) bodies, as seen under a power of 450 diameters. Fig. 3. Section of first-formed portion of the disk, laying open the primordial chamber, a, and the spiroloculine chambers, partially divided as at b, which coil round it. Magnified 125 diameters. Fig. 4. Portion of the sarcodic body shown in fig. 1, enlarged to 125 diameters, to show the distribution of the nuclear (?) corpuscles : a, expanded extremity of the last spiroloculine coil ; b, b, b', b', portions of preceding coils, crowded with nuclear (?) corpuscles ; c, orbiculine sub-segment, with five corpuscles ; d, d, d, d, orbiculine sub-segments, each with one or with two corpuscles. Fig. 5. Central portion of the calcareous disk, as seen by transmitted light ; a, ex- panded chamber formed by the termination of the spiroloculine coil, and DR. W. B. CARPENTER ON ORBITOLITES TENUISSIMA. 573 closed -in by a peneropline septum traversed by four passages ; b, second chamber, divided by radial partition into orbiculine chamberlets ; c, third chamber, not here separated from the second by a septum, and having only one radial partition ; d, d, fourth chamber, having at d', d', lateral extensions which begin to enclose the spiroloculine coil ; e, e, fifth chamber, with lateral extensions, e, e, proceeding still further backwards ; these chambers, and those that succeed them, divided by radial partitions into orbiculine chamberlets. Magnified 75 diameters. MDCCCLXXXIIJ. 4 E m~penier. Phil.'lrans. 1883. Plate 37. ORBITOLITES TENUISSIMA. Phil. Trans. 1883. / J Zale38. West.Newraaj.SC" ORBITOLITES TENUISSIMA [ 535 ] XXV. Researches on Foraminifera. Fourth and concluding Series. By WILLIAM B. CARPENTER, M.D., F.E.S., F.G.S., F.L.S. Eeceived June 14, Bead June 14, 1860. 172. I HAVE now to bring to a close my account of the structure of those typical forms of Foraminifera which it has been my object to elucidate, by a description of four remarkable generic types ; of which the first, Polystomella, has long been known, but has been hitherto very imperfectly comprehended; the second, Calcarina, has never been carefully studied ; the third, Tinoporus, has only been very imperfectly known in one of its forms ; and the fourth, Carpenteria, is altogether new. Each of these will be found to present features of interest peculiarly its own ; Polystomella being remarkable for the very symmetrical distribution of its canal-system, whose existence and whose relation to the multiple fossa; upon its surface have hitherto been altogether overlooked ; Calcarina being distinguished by the extraordinary development of its " intermediate" or rather " supplemental skeleton," and by the amplification of the canal-system for its nutrition ; Tinoporus presenting us with a type of structure that is intermediate between the Eotaline group (to which it is allied in the character of its individual chambers) and the Orbitoline (to which it approximates in its mode of growth), and that helps us greatly in the interpretation of the structure of the fossil Orbitoides; and, lastly, Carpenteria furnishing us with a connecting link of the most striking significance between Foraminifera and Sponges. Genus POLYSTOMELLA. 173. History. Of the minute shells to which the generic name Polystomella is now assigned, one species, now known as P. crispa, seems to have early attracted the atten- tion of conchological observers and collectors, on account both of its beauty and of the frequency of its occurrence ; having been described and figured more than a century ago by PLANCUS and GUALTIERI, and adopted by LINNJEUS under the designation Nautilus into his ' Systema Naturae.' By this designation it continued to be known from the time of LINNAEUS to that of LAMARCK ; having been described and figured by WALKER, SOLDANI, FICIITEL and MOLL, MONTAGUE, DILLWYN, and many other writers of the latter part of the last and the early part of the present century. Its dissimilarity to Nautilus was first clearly pointed out in 1822 by LAMARCK; who conferred upon it the generic distinction Polystomella, apparently under the impression that the numerous pits on its surface are really multiple mouths of passages leading directly to its chambered cavity. His definition of the genus, contained in the First Edition of his ' Animaux sans Ver- tebres ' (torn. vii. p. 625), is as follows : " Coquille disco'ide, multiloculaire, a tours 536 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEBA. contigus, non apparens au-dehors, et rayonnee a 1'exterieur par des sillons ou des cotes qui traversent la direction des tours. Ouverture composee de plusieurs trous diverse- ment disposes." The genus was soon afterwards adopted by M. D'OBBIGNY in his first systematic classification of Foraminifera * ; and he seems also to have adopted LAMARCK'S idea of the structure of this type, which he defines as folloAVsf : "Les Polystomelles se distinguent de tous les genres de Nautiloidees par ce caractere singulier, que les ouver- tures du bord de la derniere loge reparaissent en fossettes, plus ou moins allongees, sur toutes les autres; les dernieres seulement ouvertes, les autres fermees. II en resulte qu'exterieurement ce genre se distingue de suite par . ce grand nombre de petits exca- vations transversales, qu'on remarque sur toutes les especes." He elsewhere adds to this description : " L'animal fait sortir des filamens non seulement par des ouvertures du dessus de la derniere loge, mais encore par des pores des cotes des dernieres" J. The same definition is more concisely repeated by M. D'OKBIGNY in his latest publication on the subject, notwithstanding that the true structure of Polystomella crispa had in the mean time been elucidated by Professor WILLIAMSON in his admirable memoir on that species ; the genus Polystomella being still defined (in the ' Cours Elementaire de Pale- ontologie et de Geologic Stratigraphiques,' torn. ii. p. 197) " Coquille nautiloide, pourvue de nombreuses ouvertures sur la derniere loge et sur les cotes de la coquille ; une cavite simple au loges." 174. I have already referred, in my General Introduction, to the elaborate inquiry of Professor WILLIAMSON into the organization of Polystomella crispa^, as having not only established several important facts in regard to its minute structure, but as having fur- nished the starting-point for all future investigations of the same kind : it now becomes necessary that I should somewhat fully recapitulate the most important features of his description, in order that it may be seen in what points my own inquiries upon a more developed form of the same group have evolved results supplementary to his. To the accuracy of his descriptions I can bear the fullest testimony ; having not only had the opportunity, through the kindness of Professor WILLIAMSON, of examining the prepara- tions still in his possession both of the shell and of the animal of Polystomella crispa, but having carried out an independent investigation upon specimens collected by myself. 175. On the exterior of each of the segmental divisions of the shell of P. crispa, strong transverse crenulations present themselves, which are deepest near the convex margin of the preceding septal band, where they terminate somewhat abruptly, and usually dis- appear before reaching the concave margin of the subsequently-formed chamber. The depressions between the elevated ridges, which are the " fossettes " of M. D'ORBIGNY, often present the appearance of orifices; but this appearance is fallacious, since at no * "Tableau Methodique de la Classe des Cephalopodes," in Anuales des Sciences Naturelles, 1826. t Voyage dans I'Amerique Meridionale, torn. v. p. 29. J Foraminiferes Fossiles de Vienna, p. 121. Transactions of the Microscopical Society, First Series, vol. ii. p. 159. GENUS POLYSTOMELLA : HISTOET. 537 period in the growth of the shell is there any passage through these " fossettes " to the cavity of the chamber, the only communication possessed by any chamber either with contiguous chambers or (in the case of the outermost chamber alone) with the exterior, being afforded by a variable number of minute orifices (corresponding with the septal pores of Peneroplis} which are to be found near the inner margin of each septal plane, close to its junction with the preceding convolution. Corresponding to the elevated ridges of the crenulations, we find a series of grooves on the internal surface, which shallow towards the anterior or concave margin of each segment, and deepen towards the posterior or convex margin ; and for a short distance from the posterior septum each groove is converted into a tube by a narrow lamella given off internally from the septum. These tubes, however, establish no communication between the contiguous chambers; for they are culs de sac, closed-in by the lamella of the septum which formed the boundary of the previously-formed chamber. In the living state they are occupied (as can be shown by examination of the decalcified body) by a set of processes of sarcode, which extend backwards for a short distance from both the outer or lateral margins of each segment of the sarcode-body, and then terminate abruptly. From the neigh- bourhood of the inner arch of each segment, on the other hand, there proceeds a series of threads of sarcode much slenderer than the " retral processes " just described, which unite each segment to the two contiguous segments before and behind, passing through the row of pores already mentioned as visible along the inner margin of the septum The shell is described by Professor WILLIAMSOX as " crowded with myriads of minute foramina," and as also covered over with small pointed tubercles, which, from the rounded forms of their bases, and their great transparency, may be easily mistaken for apertures in the shell, especially in the " fossettes," where these tubercles are often very large. He further pointed out that the umbilical region is occupied by a solid mass of shelly substance, into which the decalcified animal does not appear to extend, and the surface of which is often marked with small depressed pits, the orifices of vertical internal passages, through which pseudopodia are probably protruded. 176. It is obvious from the foregoing account that \iPolystomella crispa is to be taken as the type of the genus, the generic definition given by M. D'OKBIGNY is based on an entire misapprehension of its true structure ; the only considerable departure from the general type of Helicostegue structure being the substitution of a series of isolated pores for the ordinary single orifice of communication between the successive chambers (a difference which in the case of Peneroplis and Dendritina we have seen to have not even a specific value), and the supposed lateral orifices having no real existence. 177. Subsequently to Professor WILLIAMSON'S memoir, an elaborate account of the characters of the genus Polystomella, and especially of a species designated P. strigilata (which seems to me only one of the multiform varieties of P. crispa), has been given by Professor MAX SCHULTZE, in the excellent treatise ' Uber den Organismus der Poly- thalamien,' to which I have already referred. He had the advantage of being able to study this species in the living state ; and he has thus been enabled to give a beautiful MDCCCLX. 4 B 538 DE. CAEPENTEE'S EESEAECHES ON THE FOBAMINIFEEA. figure, not merely of the shell, but also of the pseudopodia protruded from various parts of its surface ; as well as to make preparations of the sarcode-body of the animal, by dissolving away the shell in dilute acid. He does not seem, however, to have had the advantage of a full knowledge of Professor WILLIAMSON'S memoir; his acquaintance with it being apparently limited to the abstract of it contained in ' L'Institut' (No. 787) ; and I find in his description of the shell a confirmation of the belief I have already had occasion to express, that he has not availed himself as fully as is desirable of the mode of examining the intimate structure of these minute objects by the preparation of very thin sections. In every point, in fact, in which he differs from Professor WILLIAMSON, I am satisfied that the truth lies with the latter ; and this not merely on account of the entire coincidence between the results of my own inquiries into the structure of Poly- stomella crispa and those of my accomplished predecessor, but also because our views are in every respect borne out by the structure of the much larger and more highly developed form of Polystomella which I am presently to describe. One point in Pro- fessor SCHULTZE'S description, however, requires special notice. He states that each of the crenulated prominences which are seen on the surface of the lateral walls of the chambers is traversed longitudinally by a wedge-shaped fissure, that is narrowest as it approaches the septal band, near which it penetrates the cavity of the chamber, whilst it becomes shallower as it widens out at the part where the crenulation merges in the smooth wall of the shell. I expect to be able to show that the supposed " fissures " of Professor SCHULTZE no more communicate with the cavity of the chambers, than do the "fossettes" of M. D'ORBIGNY; but that they are really the outlets of the canal-system, whose existence in Polystomella has not been discovered either by Prof. SCHULTZE or by Prof. WILLIAMSON, but which attains an extraordinary development in the type which has specially fallen under my observation. 178. The specimens of Polystomella, of which I have now to give an account, were chiefly collected by Mr. JUKES in his Australian dredgings ; I have met with the same form, however, in Mr. CUMING'S Philippine collection ; and I have reason to believe it to be generally diffused through the Indian and Polynesian seas. It seems to be the P. craticulata of FICHTEL and MOLL. The empty shells are occasionally the subjects of that very curious infiltration of silicate of iron, to which attention was first directed by Professor EHRENBERG as a peculiar mode of fossilization of Foraminifera, causing in- ternal " casts " of their chambers to be preserved long after their shells have been destroyed, in his memoir ' tlber den Gruensand und seine Erlauterung des organischen Lebens"*, and which was soon afterwards shown by Professor BAILEY f to be taking place at the present time over certain parts of the ocean-bottom. I have recently been enabled, through the kindness of Mr. W. K. PARKER, to examine a number of most perfect and beautiful " casts " which he has obtained, not merely of fragments, but of the entire animal of this type of Polystomella, by treating with dilute acid shells which * Abhandlungen der Konigl. Akad. der Wissenschaften, Berlin, 1855. t Quarterly Journal of Microscopical Science, vol.Y. p. 83. GENUS POLTSTOMELLA : EXTERNAL CHARACTERS. 539 have been thus infiltrated. These casts represent not merely the segments of the sarcode-body with their connecting stolons, but also those prolongations of the body which occupied the canal-system ; and as they preserve with the greatest exactitude the natural forms and relative positions of these parts, they really afford more precise and satisfactory information than that which could have been derived from an examination of the sarcode-body of the animal itself, since its softness and friability are such as greatly to interfere with the due appreciation of its characters, when it is deprived of the support afforded by the shell. 179. External Characters. This type of Polystomella (Plate XVII. fig. 1, , b) is distinguished from the others already noticed, not only by its comparatively large dimensions, the diameter of some of the specimens in my possession exceeding one- sixth of an inch, but by the considerable proportion of its two lateral surfaces occu- pied by that solid calcareous nucleus which is confined in other species to the umbilical region. The diameter of this nucleus is usually about three-fifths of the whole diameter of the specimen ; so that it covers and conceals all the earlier convolutions, meeting at its outer margin the chambers of the last formed whorl (as is made evident by vertical sections, Plate XVII. fig. 2), which are consequently the only chambers that show themselves externally, although the last formed whorl does not itself extend far over the preceding. I have not unfrequently found this central nucleus, however, to be sufficiently transparent (after its surface has been cleaned by a short immersion in dilute acid) to allow of the inner convolutions being discerned through it, when the microscope is focused down to their surface, and a strong light is directed upon this ; and it then becomes obvious that, if the solid nucleus were removed, the form of the shell would be bi-concave instead of bi-convex, the thickness of each whorl (i. e. the distance between its two lateral surfaces) being greater than that of the pre- ceding, and the later whorls not extending themselves over those previously formed. The septa are marked externally (as in most other Foraminifera of the nautiloid type) by bands which indicate their junction with the outer walls of the chambers : these bands are meridional (so to speak) in their direction, extending from the margin of the nucleus on one side to that of the nucleus on the other side ; they are not usually (in adult specimens at least) either elevated above or depressed below the surface of the walls of the chambers on either side of them ; but they are distinguished by their difference of texture, their substance being much more transparent and glistening than that of which those walls are composed. 180. The surface of the central nucleus is marked at pretty regular intervals with minute punctations (fig. 1, J), each of which occupies the centre of a little dimple or depression; and rows of similar punctations are very commonly seen to extend from the nucleus on either side, in a direction corresponding to that of the septal bands (fig. 1, a), two such rows usually intervening between each septal band and that which precedes or follows it (Plate XVIII. fig. 1, hh, h'h'). In the older portion of the last formed whorl, it is sometimes to be observed that these punctations with their surrounding dimples 4B 2 540 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. constitute the only interruption to the general uniformity of the surface, the septal bands not being clearly distinguishable ; and this disposition is commonly found to prevail on the surface of the inner whorls, when it is exposed by the removal of the outer (M', if). In the newer portion of the last formed whorl, on the other hand, we may observe that instead of each punctation having a separate dimple of its own, the corresponding punctations of the two rows lie in a succession of furrows that pass transversely between the septa (gg', gg'). In the most recently formed portions of specimens that have not attained their full growth, we find these furrows to be deeper towards the posterior than towards the anterior margin of each interseptal space ; and in the deepest portion of each of these furrows, which obviously correspond with the "fossettes" of P. crispa (although much less pronounced), a minute punctation may be brought into view by careful examination, a corresponding row of punctations being also traceable on the other side of the septum*. These varieties of superficial aspect may present themselves on different parts of one and the same specimen ; and it will appear from the explanations which I shall presently have to furnish, that they are occasioned by differences in the degree in which the proper external wall of the chambers is thickened by an exogenous deposit upon its surface, continuous with that of which the central nucleus is composed (^f 185). 181. Internal Structure. By the fortunate contingency already adverted to, I have been enabled to study the internal structure of this remarkable type, not merely by examining thin sections of the shell taken in various directions, and by comparing the appearances they present with those obtained by laying open its interior by fracture ; but also by submitting to microscopic examination siliceous " casts " of its cavities and channels, which appear to represent with the utmost fidelity the forms and connexions of the various parts of the sarcode-body which occupied those cavities and channels in the living state of this organism. Between the results of these two modes of study I have found the conformity to be so exact, that the account of the structure of the animal which I should have given from examination of the shell alone, has not needed to be modified in the slightest particular by the information more recently furnished by these " casts ; " and, in point of fact, the ideal representation in Plate XVIII. fig. 1, which was entirely based on the former source of information, has not needed the least altera- tion to bring it into accordance with the exact delineations of the latter class of objects which are given in figs. 12, 13f. 182. In the general shape and proportions of its segments, this type of Poiystomella * In order to distinguish the orifices of these pimctations, it is advantageous to remove from the surface of the shell that opacity which it derives from abrasion, and to get rid of the fine particles of calcareous matter which often choke up and obscure its pores. This is readily effected by immersing it for a short time in water so slightly acidulated with nitric or hydrochloric acid as only to exert a very feeble degree of solvent power. f I think it well to state this circumstance, as it may increase the confidence accorded to my descriptions of other types, of which the shells alone have been submitted to examination. GENTJS POLYSTOMELLA : INTERNAL STKTJCTUEE. 541 differs remarkably from most of the other nautiloid Foraminifera ; the breadth of each of the later whorls being many times exceeded by what may be termed its thickness, that is, by the distance between its two lateral surfaces. Thus the segments come to have somewhat of the form and arrangement which the carpels of an orange would exhibit, if, instead of lying in a single circle round a central axis, they were disposed in a succession of whorls, with a progressive increase in their dimensions. This comparison may be conveniently carried a little further. For as each carpel of the orange has its own investing membrane, so that the partitions between the adjacent carpels are double, so each segment of Polystomella has its own proper shelly investment, causing the septa which separate the adjacent segments to be double, as was originally pointed out by Professor WILLIAMSON, and as I have shown to be the case also in the higher types of Foraminifera generally. But further, as the separate carpels of the orange are collectively invested by a general integument, which also to a certain degree dips down between them, and which fills up what would otherwise be void spaces about the two poles of the spheroid, so shall we find that the proper walls of the spirally arranged segments of Polystomella are strengthened and consolidated by a secondary calcareous deposit upon their external surface, corresponding to that " intermediate skeleton," of which less developed examples have already been furnished by Cycloclypeus, Hetero- stegina, Operculina, and Amphistegina, its most distinctive peculiarity in Polystomella being its extraordinary thickness on the two lateral surfaces of all but the last formed whorl. 183. The spire of Polystomella, like that of other nautiloid Foraminifera, commences in a central cell, the dimensions of which are extremely variable ; the difference between the extremes of its size being, in fact, not less remarkable than that which I have shown to present itself in Orbitolites (First Series, ^[ 44). Thus in Plate XVII. fig. 3, which represents a section of the five inner whorls of a full-grown specimen, taken through the equatorial plane, we trace a progressive diminution in the size of the chambers as we approach the central cell, which is itself no larger than the chambers in nearest proximity to it. In fig. 4, on the other hand, which represents a correspond- ing section of the inner portion of another specimen, drawn under the same magnifying power, we see that not only is the size of the earlier whorls and of their component chambers considerably greater, but that the central cell alone occupies about the same space as the first 2^ whorls of the specimen represented in fig. 3. The average seems to be intermediate between these two extremes. The breadth of the successive whorls increases much more gradually than in most other nautiloid Foraminifera, in this respect resembling Nummulites rather than the recent forms described in former memoirs; and there is no tendency whatever, even in the oldest and most developed specimens, to that rapid opening-out of the spire, which we have seen to be so marked a feature of the older specimens of Heterostegina, Peneroplis, Operculina, and Amphi- stegina. The largest number of whorls I have met with in any individual (that, namely, to be counted in the specimen whose inner portion is represented in fig. 3) is eleven : 542 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. the earlier four or five of these completely invest the preceding, their chambers extend- ing on either side to the centre of the spire, as is partly shown in the vertical section (Plate XVII. fig. 2) ; but as new whorls are added around these, the chambers cease to be thus prolonged over the preceding whorls, which would consequently be apparent externally if not concealed by the nucleus. The distance between the successive septa remains nearly the same after the spire has made two or three turns ; and thus the size of the segments, as seen in an equatorial section, remains pretty much the same through- out all the later growth of the shell, while the number of chambers in the successive convolutions increases nearly in proportion to the length of those convolutions. 184. In these particulars, therefore, Polystomella corresponds rather with the Cyclo- stegue than with the ordinary Helicosttyue Foraminifera ; and this correspondence is further borne out by the existence of an obvious relation in the position of the chambers of successive whorls, which exists in the former, but which cannot be traced in those forms of the latter to which our attention has hitherto been directed. It has been shown that, in OrUtolites and Cycloclypeus, the chambers of each concentric zone normally alternate in position with those of the zones which adjoin it internally and externally (\^[ 17, 18, 100) ; and that this relation arises out of the mode of their communication with each other. In Peneroplis, Operculina, and Amphistegina, on the other hand, the position of the chambers of each successive whorl appears to be per- fectly independent of that of the chambers in the whorl which preceded it. Now in the Polystomella whose structure we are considering, such an alternating arrangement appears to be the normal one, as is shown in the relation of the chambers b, b, Plate XVII. fig. 8, to a, a, and c, c ; so that lines drawn from the centre of the spire through the septa of one convolution would pass through the middle of the chambers of the next, and would again meet the septa of the convolution beyond. This arrangement is shown in Plate XVII. fig. 7, as it presents itself in an actual section, and more diagrammatically in Plate XVIII. fig. 1 ; in both the vertical plane of section lays open the chambers a, a\ a?, whilst it traverses the septa b, b\ b 3 , b 3 of the alternating whorls. It is, however, by no means constant; being very liable to be disturbed by that interpolation of additional chambers, which is required for the augmentation of their number in successive whorls. We shall presently see (^[ 189) that although this relation does not depend, as in OrUto- lites and Cycloclypeus, upon direct communications between the chambers of successive rows, it is manifestly connected with the peculiar disposition of the canal-system, which here acquires a remarkable development and importance. 185. Although, however, there is but little progressive increase in the dimensions of the successive chambers, and of the segments of the sarcode-body which occupy them, as seen in sections taken through the equatorial plane, it is made obvious by sections made at right angles to this (Plate XVII. fig. 2), that a rapid augmentation takes place in what may be termed the meridional direction ; the distance between the two lateral surfaces of each whorl being considerably greater than between those of the preceding, so that the chambered portion of the shell progressively increases in thickness from the centre GENUS POLYSTOMELLA : INTERNAL STRTJCTTJRE. 543 towards the circumference. The conical hollow thus left on each side in the central portion of the shell, is entirely filled up by the solid nucleus already adverted to : the calcareous deposit, however, of which the nucleus is composed is by no means limited to it, but extends over the whole outer surface of each whorl, except where (in well- preserved specimens) the portion last formed is as yet unconsolidated by it. For a careful examination of sections taken in different directions, makes it clear that whilst the internal portion of the spiral lamina that forms the outer wall of each chamber is continuous with the nearest lamella of the adjacent septum on either side (Plate XVII. fig. 10), the substance of the external portion is no less continuous with that of the calcareous nucleus. The additional deposit is obviously homologous with that which forms the " intermediate skeleton " in Cycloclypeus (^[ 99), though less differentiated from the proper walls of the chambers than we have seen it to be in that type, or than we shall find it to be in Calcarina (^[202). The whole thickness of the spiral lamina is generally traversed by minvite tubuli, passing in a radial direction from one surface towards the other ; but these have by no means either the closeness or the regularity which distinguishes the tubular structure in Cycloclypeus and Operculina, and the shell- substance is in many parts so destitute of tubuli as to be of almost glassy transparence. The fun-owing of the external surface (^[ 180) is seen in vertical sections (Plate XVII. fig. 2) not to be produced by mere superficial excavations, but to proceed from a plicated arrangement of the spiral lamina ; and this is related to the prolongation of the posterior margin of each segment into a series of "retral processes" (Plate XVIII. fig. 12, a), corresponding to those described by Professor WILLIAMSON in Polystomella crispa (^[ 175). They are, however, much less elongated in this type, simply giving a crenu- lated margin to that angle of the segment, which contrasts remarkably with the smooth unbroken aspect of its anterior border #, V. The spiral lamina which forms the outer wall of the chamber, being modelled (so to speak) upon the surface of these retral pro- cesses, presents internally a corresponding series of grooves, which are deepest towards the posterior margin, and become rapidly shallower in passing towards the anterior margin, of each chamber, as is shown at , a 1 , 3 , figs. 1 and 11, Plate XVIII. : these grooves are not, however, as in P. crispa, completed into tubes for part of their length by an additional lamella of shell given off from the septum (^f 175); but they are sometimes shown, in sections which happen to traverse them (Plate XVII. fig. 10), to be extended into csecal prolongations (#, a) by backward inflexions of the septa at their junction with the spiral lamina. 186. The communication between the successive segments of the same whorl is esta- blished by a number of minute processes or stolons of sarcode (c, figs. 12, 13, Plate XVIII.), which pass at regular intervals between their internal margins through a series of pores which can be distinguished along the inner border of each septum (fig. 1, c, c', c") close to its junction with the preceding convolution. I have not detected in any instance, either in sections of the shell, or in the siliceous casts which so exactly represent the sarcode-body, any other communications between the chambers or their contained seg- 544 DB. CAKPENTEE'S EESEAECHES ON THE FOEAMINTFEEA. ments ; and I am therefore strongly disposed to believe that Professor MAX. SCHULTZE must have been misled by appearances when he stated (op. cit. p. 65) that various other parts of the septal plane are marked by similar pores, more particularly as his figures of the decalcified body do not show that any other threads or stolons of sarcode pass from one of its segments to another, than those just described. 187. So far, then, the structure of this comparatively gigantic type of Polystomella accords very closely with that of the more delicate species so well described by Professor WILLIAMSON. I have now, however, to give an account of a remarkable feature in its organization, namely, its highly developed canal-system ; which, though not entirely wanting in P. crispa, is so imperfectly presented there, that Professor WILLIAMSON may well be excused for having overlooked it, especially when it is borne in mind that at that period the existence of such a system in Foraminifera was altogether unknown. The general arrangement of this canal-system may be most readily apprehended from an ex- amination of the delineations of the internal casts given in Plate XVIII. figs. 12, 13 ; for the infiltrating substance which has penetrated the chambers has also found its way not only into the main trunks, but also into the minute ramifications of this system, and has thus given just that representation of their distribution and relations, which is afforded in regard to the blood-vessels of the higher animals by a well-injected and clearly dissected anatomical preparation. We observe, in the first place, that in each of what may be termed the two polar regions of the spheroidal body, there is a continuous spiral canal (fig. 12, d, d\ d 2 ), which overlies the extremities of the segments. These two spiral canals (which, although so widely removed from each other, are obviously homologous with the two spiral canals of Operculina, ^f 159) communicate with each other by a very regularly disposed series of canals which pass in a meridional direction between the adjacent external margins of the segments (e, e\ e*). And each of these meridional canals gives off, in its course from one polar region to the other, a uniform succession of pairs of short passages (/,/',/ 2 ) that diverge from each other widely, one series inclining backwards over the uniform anterior margin of the segment next behind it, whilst the other series passes forwards in the intervals between the " retral processes" of the segment next in front of it. The passages which thus diverge from the meridional canals of the outer whorl speedily debouch at its surface ; but if we examine into the termination of those appertaining to the inner whorls (which is best seen in such frag- ments as the one represented in fig. 13), we find that they become continuous with the stolons of the whorl which surrounds them, as is shown at c, c 1 , fig. 12. Further, it may be perceived that each of the meridional canals receives branches from the canal- system of the segment internal to it ; this point, however, can be more clearly made out in sections of the shell. 188. The spiral canals are frequently brought into view for part of their course, by sections of the shell that pass through it in a direction parallel to the equatorial plane, but at no great distance from one of the lateral surfaces. Such a section, passing over the chambers of the inner whorls, is shown in Plate XVII. fig. 5 ; where we see the GENUS POLTSTOMELLA : CANAL-SYSTEM. 545 spiral canal giving off the meridional canals, and these again sending off their diverging branches. Towards its centre, the spiral canal communicates with an irregular set of lacunae, which are excavated in the solid nucleus. In Plate XVII. fig. 6, are shown portions of two whorls (, a, and b, b) of the spiral canal, as shown in a section passing near the lateral surface of the outer convolution; proceeding from the outer side of these, in like manner, we see the meridional canals (c, c, d, c/, d) with their first pairs of diverging branches (see also fig. 12) ; whilst another set of canals (d, d, d) is seen to proceed from the inner side of the spiral canals, tending more or less obliquely towards the lateral surface ; and the portion of the section which passes through the solid cal- careous nucleus is seen to be perforated by numerous apertures (e, e) of corresponding diameter, disposed at pretty regular intervals. The relation of these to the canal- system is clearly evidenced by vertical sections, such as those represented in figs. 2 and 11 ; in which we see at a, a, the orifices of the spiral canals transversely or obliquely divided, and their connexion with the meridional canals b, b ; and which further show that the solid calcareous nucleus is itself traversed by straight canals, c, c, c, which spring from the successive convolutions of the spiral canal, and pass directly, without branching or inosculation, to the external surface. That this remarkable portion of the canal-system does not fully show itself in the " casts " represented in Plate XVIII. fig. 12, is easily understood, when it is remembered that the whole substance traversed by the straight canals having been removed, their long and slender casts would be left entirely without support ; and the points at which these have been broken off from the cast of the spiral canal are in fact to be seen on a careful examination, as there represented. 189. It is shown by the comparison of vertical and horizontal sections of the shell with fragments obtained by fracture (Plate XVII. figs. 2, 7, 10, Plate XVIII. fig. 11), that the meridional canals are in reality spaces left by the divergence of the two layers of which each septum is composed, in the immediate neighbourhood of its junction with the spiral lamina which constitutes the external wall of the chamber (Plate XVII. fig. 7, c, and fig. 10, b; Plate XVIII. fig. 11, b, b) ; and that they are thus homologous with the arches of the interseptal system of canals that connect together the spiral canals of Operculina (^[ 157), presenting, however, a much greater uniformity and constancy in their disposition. The diverging branches given off from these (Plate XVII. fig. 8, d, d\ d 2 ) consequently pass at once into the spiral lamina, through which they run obliquely towards the external surface of the convolution, usually increasing in diameter as they proceed. The divergence of the branches of each meridional canal causes those proceeding from adjacent canals to approach one another ; and when the spiral lamina has attained its full development, they not unfrequently open at its sur- face into the same depression, this being midway between the septa from which they respectively sprang ; and it appears to be from the correspondence of these junctions with the intervals between the segments of the succeeding whorls (as seen at e, e'), that the alternating arrangement of the chambers of consecutive whorls arises, of which MDCCCLX. 4 C 546 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. mention has already been made (*f[ 184), the prolongations of sarcode which occupy the diverging branches there passing into the stolons which connect the adjacent segments. It will be easily understood, however, that the position of the external orifices of these diverging branches will depend upon the thickness of the spiral lamina which they have to traverse before gaining its surface. In the newest portion of a shell which has not yet attained its full growth, we find that lamina comparatively thin ; its surface is distinctly marked by the septal bands (Plate XVIII. fig. 1, gg*, gg') ; and the external walls of the chambers present an alternation of ridges and furrows, passing directly across from one septal band to another, the ridges corresponding to the grooves of the internal surface that receive the "retral processes" (^[ 185), and the furrows with the internal ridges that separate these grooves. Into these furrows, which represent the deeper "fossettes" of P. crispa, the diverging pairs of branches from each meridional canal open by minute pores on either side of the septal band, as is shown in fig. 1, and as will be readily understood from the relation of the parts as displayed in fig. 12. The subsequent formation of a calcareous deposit, continuous with that which solidifies the umbilical portion of the shell, upon the external surface of the spiral lamina (^f 185), renders the septal bands less distinct, and obliterates the ridges and furrows of the intervening surface, as shown in the portion h h' of fig. 1 ; and at the same time it carries the orifices of the diverging branches from the neighbourhood of the septa into closer proximity with those of the branches proceeding from the adjacent meridional canals. As the diverging branches enlarge greatly in diameter with their augmentation in length, their superficial orifices become more and more conspicuous; each is surrounded by a little pit or depression of its own (fig. l,n', H'); and the rows of these depressions, when the spiral lamina has acquired its full thickness, constitute the only markings which it presents, the septal bands being completely obliterated, as is best seen on the surface of one of the interior whorls, exposed by the removal of that which covered it. It is obvious, therefore, that these depressions, which are related only to the distribution of the canal-system, are essentially different in character and position from the " fossettes " of P. crispa, which intervene between the ridges that cover-in the retral processes ; but they have this in common, that the orifices of the diverging branches are to be found in both of them ; and the removal of the superficial portion of the spiral lamina, even when thickest (which may easily be accomplished by the assistance of dilute acid), brings back these orifices, in P. craticulata, to the immediate neighbourhood of the septal bands, which then again become apparent*. 190. The meridional canals are further connected with the older and more internal portions of the organism, as well as with the newer and more superficial ; this connexion being established by a series of branches that pass between the two layers of the septa * This arrangement of the orifices of the diverging branches of the meridional canals is beautifully shown (as Mr. W. K. PABKEE has pointed out to me) in the Nautilus striato-punctatus of FICHTEL and MOLL, which Mr. PAEKEB considers to be the Geoponus stella-lorealis of EIIEENBEBG, but to be really a Nonionine form of Polystomella. See Ann. of Nat. Hist. ser. 3. vol. v. p. 103. GENUS POLTSTOMELLA : CANAL-SYSTEM. 547 in a radial direction (Plate XVII. fig. 2, d, d', Plate XVIII. fig. I,/ 1 ,/ 1 ), from the meridional canals of each convolution to the stolons which unite the segments of that convolution. These, which may be distinguished as the converging branches, are, however, much less regular in their distribution than those which pass outwards from the meridional canals to the stolons of the succeeding whorl. 191. Thus, then, it becomes apparent that by means of the two spiral canals, the number of convolutions of which equals that of the whorls of the shell, the very numerous meridional canals, of which there is one for every segment of each whorl, the vast multiplication of pairs of diverging branches, of which each meridional canal sends off a number equal to that of the connecting stolons between the segments, and the very considerable aggregation of converging branches, which probably do not fall far short of the preceding except in being single whilst they are in pairs, a very complete system of intercommunications is maintained between the external surface and even the innermost portions of the shell. That these passages are occupied, in the living animal, by prolongations of the sarcode-body, there can scarcely, I think, be any reasonable doubt ; and when we look to the remarkable development of what has been elsewhere termed the " intermediate skeleton," but which may here be more appropriately termed the " supplemental skeleton," namely, the secondary calcareous deposit which not only forms the solid nucleus, but spreads itself over the entire surface, adding considerably to the thickness of the spiral lamina, it cannot be deemed improbable that the special purpose of the canal-system is the formation and nutrition of this supplemental skeleton, which has obviously no direct relation to the segments of the animal body contained within the chambers. Through the trumpet-shaped diverging branches which open in such numbers upon the surface of those chambers, and the straight canals which arise from the nucleus, there will be abundant opportunity for the sarcode-body to extend itself over the whole exterior of the shell, and thus to form any additional deposit upon its surface. 192. Having carefully re-examined P. crispa with a view of ascertaining to what extent the canal-system is developed in it, I am enabled to state that I have found it to possess a canal-system distributed on the same plan with that of the organism we have been considering, but much more limited in its extent, as might be expected from the much inferior development of the " supplemental skeleton." In the vertical section repre- sented in Plate XVII. fig. 9, we see the solid umbilical protuberances a, a', traversed by straight canals, which terminate in the pits on their surface that were originally noticed by Professor WILLIAMSON, and supposed by him to give exit to pseudopodia. At b, b, b', b', we see the transverse sections of the spiral canals, from which the straight canals just mentioned may be presumed to proceed ; and at c c, dc 1 , we see plain indications of a system of interseptal canals, resembling those of Operculina in their aspect and distri- bution, but clearly homologous, in their relations to the spiral canals, with the meri- dional canals of P. craticulata. I have not been able to detect any indication that diverging branches are given off from them, to pass through the spiral lamina to the 4c2 548 DE. CAEPENTEB'S EESEAECHES ON THE FOEAMINIFEEA. external surface ; and the absence of any such distribution accords with the absence in P. crispa of any of that supplemental calcareous deposit which so remarkably changes the aspect of the general surface in P. craticulata. Genus CALCAEINA. The type next to be described is one in which the supplemental skeleton and the canal-system both attain a more remarkable development than in any Foraminiferous organism that has yet fallen under my notice ; and their mutual relation here becomes so obvious, that no reasonable doubt can be entertained in regard to it. 193. History. The generic name by which this type is now known, and which indi- cates its resemblance in form to the rowel of a spur, was first conferred upon it by M. D'OEBIGNY in 1825 ; the organism itself, however, was previously well known, both in the recent and in the fossil state, having early attracted the attention of the collectors of minute Testacea through the singularity of its shape. It seems to have first received the name of Nautilus Spengleri from GMELIN, its specific designation having been con- ferred on it in compliment to SPENGLEE, who was among the earliest to direct attention to it ; and under this name it was described and well figured in several of its varieties in the 'Testacea Microscopica ' of FICHTEL and MOLL (p. 84, plate 14. figs, d-i, plate 15. figs, i-k), who refer to the authors cited below* for previous notices of it. SPEXGLEE'S specimens were from Amboyna and Coromandel ; SCHEOTEE found the species in the Adriatic ; FICHTEL and MOLL obtained their specimens from the Indian Ocean and from the Red Sea ; and D'ORBIGNY received his from Madagascar, the Isle of France, Rawack, the Marianne group, Cayenne, and Martinique. The specimens on which my own descriptions will be founded were partly collected by Mr. CUMING in the Philippine Seas, and partly obtained by him from the Mediterranean, in the neighbourhood of Malta. This type may be said, therefore, to have a wide distribution through the seas of the warmer regions of the globe. 194. The fossil specimens of this type appear to have been first 'noticed in the cre- taceous beds of Maastricht by FAUJAsf. They were described and figured by LAMAECK^ under the designation of Siderolites calcitrapaides ; but he totally misunderstood the nature of the organism, which he grouped with the Corals, instead of among Polytha- lamia. The genus Siderolina has been adopted by M. D'OEBIGNY, who seems to have been entirely ignorant of the generic if not specific identity of the Maastricht fossils with his recent Calcarina calcar, as to which no doubt whatever is entertained either by myself or by Messrs. W. K. PAEKEE and T. RUPEET JONES . It is stated by these observers (loc. cit.), that another variety of the same with shorter spines, occurring in * LINNJETJS, Syst. Nat. xiii. ; GMELIN, Syst. Nat. p. 3371, No. 10 ; SPENGLEB, Schrift. dan. Gesellsdi. Kopenh. vol. i. p. 373, pi. 2, fig. 9, a, b, c ; SCHBOTEB, Einleit. Conch.-Kennt. vol. i. p. 756 ; Neue Literat. u. Beytr. z. Naturg. vol. i. p. 309, pi. 1, figs. 3-6 ; SCHEEIBEES, Conch.-Kennt. vol. i. p. 5, No. 10. t Hist. Nat. de la Montagne de St. Pierre, a Maastricht. t Syst. des Anim. sans Vertebres, 1801, p. 376 ; and Tableau Encycl. et Method, pi. 470, fig. 4, a-Tc. Ann. of Nat. Hist. 3rd Series, vol. iii. p. 480. GENUS CALCABINA: EXTEENAL CHAEACTEES. 549 the Eocene Tertiaries of France, has been described by D'ORBiGNY* under the name of Eotalia armata, and by DssHAYEsf under that of Calcarina rarispina. 195. The following is the character of the genus Calcarina given by M. D'ORBIGNY in his ' Foraminiferes Fossiles de Vienne ' (p. 160). " Coquille libre, spirale, deprimee, tres rugueuse, formee d'une spire enroulee sur le cote, entierement apparente en dessus, embrassante en dessous, composee de loges prolongees en appendices marginaux, souvent tres allonges, representant, dans leur ensemble, comme la molette d'un eperon. Ouver- ture en fente longitudinale contre 1'avant dernier tour de spire. Semblable aux Rotalines par son enroulement spiral, par ses deux cotes a peu pres aussi convexes, par 1'emplace- ment de son ouverture, ce genre s'en distingue par les tres longs prolongements exterieurs de son pourtour, appendices singuliers qui font supposer, dans 1'animal, un mode de secretion tout particulier." OfSiderolina he says (p. 116) : " Coquille libre, equilaterale, orbiculaire, encroutee en dehors, composee d'une spire embrassante a tous les ages, ayant des appendices allonges au pourtour, interrompant, dans 1'interieur, la suite des loges. Ouverture contre le retour de la spire toujours masquee a la derniere loge. Les Siderolina se distinguent des Nummulina par les appendices du pourtour, qui, dans 1'interieur de la coquille, viennent interrompre la suite des loges obligees de passer de chaque cote." No mention is made of Calcarina in M. D'ORBIGNY'S most recent sum- mary of the classification of the Foraminifera (Cours Elementaire de Paleontologie), this genus not being regarded by him as occurring in the fossil state ; but he repeats his characterization of Siderolince as Nummulites the turns of whose spire are inter- rupted by testaceous appendages. It will presently appear that the conception which M. D'ORBIGNY has formed of the structure of this type is erroneous in so many particulars, that without the assistance of figures and models it could not be recognized by the generic characters which he has assigned either to Calcarina or to Siderolina. 196. A very correct general description of the structure of Calcarina has recently been given by Messrs. W. K. PARKER and T. R. JONES, in their valuable series of papers on the " Nomenclature of the ForaminiferaJ" ; but their information in regard to it having been almost entirely based on external characters, and their description being altogether deficient in those minuter details which it is my special purpose to record, a full account both of its internal structure and of the principal varieties of its external configuration will not be the less valuable to those who may have occasion to study this type. 197. External Characters. The comparison of the form of this organism to that of the rowel of a spur sufficiently well characterizes the general aspect which it ordinarily presents, as shown in Plate XIX. figs. 1-4. The feature which most distinguishes it from other Helicostegue Foraminifera (only some Eotalice and Polystomellce presenting a slight approximation to it in this particular) is the presence of a set of rays or spines, variable alike in number, length, and direction, which diverge from the central disk. * Ann. des Sci. Nat. torn. vii. p. 273, No. 22 ; and Modeles, No. 70. t Lyell's Manual of Geology, 5th ed. p. 228, fig. 236. | Annals of Nat. Hist., Ser. 3. vol. v. p. 174. 550 DE. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. Pig. I. These spines are usually few in young specimens, and become numerous with age ; but this rule is by no means constant, since full-grown specimens are occasionally met with having no more than four or five very short spines (Fig. III. H). The spines are usually either cylindrical or somewhat club-shaped, the latter form being the more common; but we occasionally see them showing a tendency to bifurcation at their extremities ; and they not unfrequently appear as if two, three, or even four spines had coalesced to form one, this being indicated not only by its unusual size, but by the mutual divergence of its components as they extend themselves from the central disk (Fig. I. B, and Fig. II. F). A somewhat remarkable contrast in the relative development of the disk and of the spines is pre- sented by the general aggregate of the Philippine and Mediterranean specimens placed in my hands by Mr. CUMIEXJ; as is shown in comparing Figs. II. and III. It is in the former (Fig. II.) that we meet with the greatest number as well as the greatest relative length of the spines ; and that the bifurca- tion or trifurcation of their extremities presents itself. In full-grown specimens of the Philippine Calcarina, we commonly find the spines diverging from the Two specimens of Philippine variety of Calcarina, distinguished by unusual exuberance of spinous outgrowths. Kg. II. Kg. III. Outline-representations of various specimens of Philippine variety of Calcarina. 6 H Outline-representations of various specimens of Maltese variety of Calcarina. GENUS CALCAEINA: EXTEENAL CHAEACTEES. 551 margin of the central disk in such abundance, that very little of that margin is left free. The length of their spines, moreover, at different ages, varies pretty constantly with the diameter of the disk, the average proportion being about two-thirds ; though we occa- sionally meet with specimens in which the disk is unusually large and the length of the spines does not equal more than half its diameter, and others in which the disk is unusually small and the length of the spines is equal to its whole width. In the Mediterranean Calcarince (Fig. III.), on the other hand, I have seldom met with more than nine spines ; and any excess beyond that number is only presented by specimens in which the disk is very large, so that wide intervals present themselves along its margin between the bases of the spines (Plate XIX. fig. 2). Their spines are nearly always simply clavate in form, any tendency to bifurcation or trifurcation at their extremities being rare ; and they show but little disposition to increase in length with the enlarge- ment of the central disk, being often not only relatively but even absolutely shorter in old specimens; as if the spines had entirely ceased to grow, and the disk had (as it were) included their basal portions within itself. In some instances, indeed, they scarcely show themselves enough to attract attention ; being little more than tubercular projections from the margin of the disk (Fig. III. H). Notwithstanding, however, this strongly marked difference in general physiognomy, it becomes obvious, on the com- parison of a sufficient number of individuals, that no line of specific distinction can be fairly drawn on such a basis between the Philippine and the Mediterranean forms ; since among the Philippine we meet with not a few specimens, in which the spines are as few and simple as they are in the great bulk of the Mediterranean (Fig. III. B, c) ; and specimens not unfrequently present themselves among the Mediterranean (Plate XIX. fig. 2), in which there is not merely an addition to the ordinary number of the spines, but a manifest disposition in many of them to subdivide near their extremities, thus showing an obvious approximation to the Philippine type. Although the spines usually radiate nearly in the equatorial plane, yet it becomes obvious, when their connexion with the central disk is examined, that they originate at different levels (Fig. III. E, F) ; this will presently be found to depend on the fact that the form of the spire is not nautiloid but turbinoid. Besides the ordinary radiating spines, an extraordinary growth of short pointed spines is sometimes seen, either partially or completely covering one or both surfaces of the central disk (Fig. II. G) ; and examples occasionally present themselves (Fig. I. A), in which the development of these seems to have altogether superseded that of the ordinary radiating spines. 198. It is remarkable that among the very young specimens of this type, a yet greater variety shows itself than among those further advanced in life. In Plate XIX. figs. 5, 6, and 7, are shown what may be considered the ordinary or normal aspect of the very young Calcarina; whilst in figs. 8-11, and Plate XX. fig. 6, we have representations (under higher magnifying powers) of examples of what may be termed the hispid con- dition, which so frequently presents itself in small Calcarince as to give rise to a question whether they should not be made to constitute a distinct species. I have satisfied 552 DE. CAEPENTEE'S EESEAECHES ON THE FOBAMINIFEEA. myself, however, by the comparison of a large number of specimens, that so continuous a gradation presents itself between the smoothest and the most hispid specimens, as to render the attempt to separate them specifically altogether futile ; and Mr. W. K. PAKKEB, who has examined a yet larger number of specimens, fully confirms this conclusion. Moreover the internal structure of these hispid specimens, as shown in horizontal section (Plate XX. fig. 8), shows no departure whatever from the ordinary type. 199. Returning, now, to the external aspect of the fully-developed Calcarina, we have to notice that each surface of the disk is very commonly elevated, especially in its central portion, into rounded tubercles, more or less closely set together (Plate XIX. figs. 3, 4). These are sometimes large and prominent, and present the semitransparent appearance which is common among the like tubercles of Ojoerculina (^[ 146); more commonly, however, they are less conspicuous either as to size, prominence, or distinctive aspect ; and sometimes they are almost or altogether wanting (figs. 1, 2). Yet it is seldom, if ever, that they are absent from both surfaces of the disk ; and it is more common to find them deficient on the surface nearest to the apex of the spire, than on that on which its last-formed chambers are visible. And even when no prominent tubercles are present, a distinction may be generally made out by careful examination in the parts of the surface corresponding to their usual situation. For whilst the exterior of the disk is marked, more or less conspicuously, with minute punctations (figs. 1, 3, 4), these are not to be seen on its tuberculated prominences ; and the like deficiency is generally to be noted in circular spots of the surface, even when it is not raised into tubercles. Occasionally, though rarely, not even this mark of differentiation is seen, the punctations being uniformly distributed over the surface, which is in that case always the one nearest the apex of the spire. 200. The surface of the spines is marked, more or less conspicuously, by a longitu- dinal furrowing (figs. 1, 3, 4), not unlike that of the " marginal cord " of Operculina (^[156). The furrows maintain a general parallelism, but there are frequent inoscu- lations between them ; and punctations marking the orifices of deeper canals are often to be noticed at the bottom of the furrows. 201. One important feature of the external aspect of the disk has yet to be mentioned ; namely, the indication of a spire, which more or less distinctly presents itself on one of its surfaces (Fig. II. u, Fig. III. D, G, H). This indication is sometimes limited to two or three chambers ; but more commonly about half a turn may be distinguished, the spire becoming absorbed (as it were) into the solid mass of the disk, as we trace it backwards. The walls of the last-formed chambers, where entirely disengaged from the disk, are extremely thin (Plate XIX. fig. 4), so that it is rare to find them perfect ; and an opening formed by the fracture of the wall of the newest chamber has been mistaken by MM. FICHTEL and MOLL, and apparently by M. D'OEBIGNY also, for the true aperture of the shell, which, as will presently appear, is of an entirely different character, and not easily to be distinguished. The prominent surface of the walls of the conspicuous chambers GENUS CALCAKINA : INTEENAL STKTJCTTJBE. 553 of the newest whorl is covered with punctations resembling those of the general surface of the disk ; but they are more minute and more closely set together, and they are dis- tributed with great uniformity, no unpunctated spaces being anywhere visible. 202. Internal Structure. When the internal structure of this organism is examined by means of thin sections taken in different directions, the apparent anomalies of its conformation are found to be dependent simply upon the extraordinary development of its " supplemental skeleton;" its general plan of structure being much simpler than the peculiarities of its aspect would seem to indicate. The spire, as laid open by vertical section (Plate XX. fig. 1), is turbinoid ; consisting usually of about five whorls (, a 1 , a 2 , a 3 , a 4 ), that start as usual from a central cell, and progressively increase in size, each whorl being applied merely to the surface of the preceding, and not investing it in any degree, the chambers being altogether destitute of alar prolongations. The aspect of the spire as seen in equatorial section is shown in Plate XX. fig. 4 ; this section, being taken in such a plane as to cut through the outer whorls a 4 , a 3 , a 2 , passes entirely over the surface of the two inner whorls a 1 and a. The disposition of the chambers, as indi- cated by such sections, is ideally shown in Plate XIX. fig. 12. The turns of the spire are separated from each other by the interposition of a thick layer of solid shell-sub- stance ; and this is quite distinct from the proper walls of the chambers, as may be well seen in Plate XX. fig. 4, where the walls of the newest chambers are shown at b to be entirely destitute of any such addition, whilst at b } in the preceding part of the same whorl we observe them encrusted by a thin additional layer d, and, as we trace this layer backwards to d l and d 2 , we perceive that it progressively augments in thickness, until it acquires its maximum at d 3 , just where it is covered by the subsequent whorl. This distinction between the proper walls of the chambers and the " supplemental skeleton " can be traced to the very centre of the spire. The septa are entirely formed by the infolding of the proper walls of the chambers, which are there flattened against each other so as to form two layers, which are usually in contact, but which sometimes diverge (especially near the external margin of the chambers) to give passage to canals. There does not appear, however, to be any regular " interseptal system " as in Oper- culina and Polystomella. The communication between the adjacent chambers of the same whorl is effected, as in Polystomella, through series of pores (fig. 1, c) disposed at pretty regular intervals along the inner margin of the septum*. 203. That the spines entirely originate from, and are strictly appendages of, the "supplemental skeleton" is well seen in fig. 4, which shows their connexion with its * Messrs. PARKEE and JONES say (Ann. of Nat. Hist. Ser. 3. vol. v. p. 175) of the aperture of Calcarina, that " in well-preserved specimens of the typical forms, the real aperture, which is essentially a slit, as in the true Botalite, becomes bridged over by delicate bars of shell-matter." I cannot but think that they have allowed themselves in this statement to be somewhat influenced by a foregone conclusion that " Calcarina is a subgenus of Botalia." The examination of numerous sections of the character represented in fig. 5 leaves no doubt in my own mind that the aperture of Calcarina is essentially such as I have above described ; and that if it occasionally in perfect specimens has the character of a continuous fissure (which I would by no means take upon myself to deny), such is an aberrant rather than the ordinary form. MDCCCLX. 4 D 554 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. successive convolutions. Thus the spine e is one of the oldest, being traceable inwards to the earlier whorls ; whilst those marked e\ e 2 , e 3 , e 4 , e 5 are obviously of progressively later production, their respective origins being further and further removed from the centre of the spire. It is, moreover, to be observed that each spine receives an augmen- tation in thickness as the convolution from which it sprang is encircled by others ; this augmentation, however, is not marked (as in the spines of Echini) by lines of demarca- tion between the earlier and the later formations ; and there is every reason to believe that the growth of the spines, both in length and in diameter, is continuous rather than interrupted. Although it might seem, from the examination of such sections only as are taken in the direction of the equatorial plane or in one parallel to it, as if the course of the spire must be seriously interrupted by the radiation of the spines (which some- times appear to be so interposed between consecutive chambers as completely to separate them); yet the fact is that owing to the turbinoid form of the spire, a spine projecting from an earlier whorl is very little in the way of even the next convolution ; for as this passes by the spine on a different level, its chambers are but slightly encroached-on, and this only upon the side which looks towards the apex of the spire, as will be readily understood by examining the relation of the last half-convolution, visible in such a specimen as the one delineated in Plate XIX. fig. 4, to the pre-formed spines, or by an inspection of the ideal represented in Plate XIX. fig. 12. 204. The canal-system of Calcarina presents a development so extraordinary in itself, and so obviously related to that of the " supplemental skeleton," as to throw great light upon its special functional destination. We do not here observe any such peculiar but limited distribution of systematically arranged passages, as that which constitutes so remarkable a feature in Polystomella ; but every portion of the supplemental skeleton, with the exception of certain solid cones presently to be noticed, is traversed by canals which run very close together, with frequent inosculations, and which thus form a con- tinuous network with long narrow meshes, that commences from the parietes of the chambers and extends itself to the very extremities of the spines (Plate XX. figs. 1, 2, 4). The proper walls of the chambers, as already stated, are uniformly perforated, like those of the chambers of Rotalice, by foramina of considerable size (averaging about 3-^5-0 th of an inch in diameter) ; with these the canals of the supplemental skeleton do not seem to be directly continuous, for they are of about double the diameter and lie further apart from one another; but immediately round the proper walls of the chambers (as shown in Plate XX. fig. 1) there seem to be irregular lacunar spaces, into which their foramina open externally, and from . which the passages of the canal-system originate. How numerous and closely-set these passages are, is shown in Plate XX. fig. 3, which is taken (under a much higher magnifying power than the rest of the figures) from a section that passes through the supplemental skeleton just outside the walls of one of the chambers, in such a direction as to cut through the passages transversely or obliquely. These passages run in different directions; some proceeding directly towards the external margin of the convolution, and being GENTS CALCAEINA: INTEENAL STEUCTTJEE. 555 continued into the spines where these are given off from it ; whilst others pass not less directly towards the two convex surfaces of the disk. In the earlier whorls of the spire, as shown in Plate XX. figs. 2 and 7, indications of spiral canals, commencing (as in Polystomella, ^[188) in a central lacunar system, are frequently traceable; and sometimes we may make out a general distribution of the canal-system of the earlier whorls (fig. 7) that strongly reminds us of that prevailing in Polystomella (^f 189), the canals of the spines originating (as seen at a) in diverging branches which radiate outwards through the spaces left between the two layers of the septa. But this arrangement soon seems to be merged, as it were, in the much more copious distribution of passages that arise from the lacunae round the proper walls of the chambers. The canals which pass towards the two surfaces of the disk soon lose the general uniformity of arrange- ment which they elsewhere present ; for they become crowded together in some situa- tions and separated in others, so as to leave a number of columnar spaces untraversed, whilst the intercolumnar spaces are copiously penetrated by them, as is seen at e, e, e, in Plate XX. figs. 1 and 5, the one figure showing the solid columns divided longitudi- nally, and the other showing them as they are cut transversely. The varying appear- ances of the external surface, as described in ^f 199, will now be understood. When, as commonly happens, the summits of the columns rise above the general level of the surface, they will show themselves as rounded tubercles. But when they are not thus elevated, they will merely be distinguished as circular spots surrounded by the puncta- tions which are the orifices of the canals. In the spines, on the other hand, the canals form a longitudinally inosculating system (Plate XX. fig. 2), of which the branches near the surface usually reach it so obliquely as to pass along it for some distance as open furrows (^[ 200), the punctations seen in which are the orifices of branches that strike the surface at a greater angle. 205. It is obvious from the foregoing description, that the statement of M. D'OEBIGNT as to the prolongation of the chambers into the spines is altogether erroneous ; and it is further obvious that the nutrition of the spines must be provided for either through the investment of their surface by external prolongations of the sarcode-body, or through the penetration of its substance by prolongations of the sarcode-body conveyed into it by the canal-system, or through both methods jointly. That the sarcode-body is con- tinued in the form of pseudopodial prolongations into the canal-system can scarcely be doubted, when it is borne in mind that such prolongations are known to pass through the pores which are scattered through the chamber-walls of Eotalia, and to extend themselves through the surrounding medium. After having made their way through the proper walls of the chambers of Calcarina, they will probably coalesce again in the lacunar cavities on the exterior of these, just as they coalesce to form a continuous layer of sarcode over the chamber-walls oiEotalice or Polystomellce ; and from the aggregation of sarcode in those cavities a new set of pseudopodial prolongations will take their departure through the canal-system of the " supplemental skeleton," just as a secondary set of pseudopodial filaments of sarcode are often seen to diverge from the little agglo- 4 D2 556 DE. CAKPENTEE'S EESEARCHES ON THE FOKAMINIFEKA. merations formed by the reunion of some of those that primarily issue from the pores of the shell. The analogy of other Foraminifera, moreover, renders it very probable that the prolongations of the sarcode-body which reach the surface through the canal-system, will reunite upon it so as to form a continuous investment over the whole ; and that this will be especially the case on the spines, appears to be indicated by the provision there is in the furrowing of the surface, for conveying the prolongations of the sarcode- body to every portion of their exterior. 206. Thus, then, by interpreting the structural phenomena presented by Calcarina according to the analogy of the facts which have been determined by observation of the living animals of the allied type Eotalia, we seem almost indubitably led to the con- clusion that the canal-system is specially destined for the formation and maintenance of the " supplemental skeleton ; " serving to convey prolongations of the sarcode-body from the segments which occupy the chambers, through the thick layer of solid shell-substance that is secondarily formed around those chambers, and through the prolongations of that layer which constitute the spines; and further, that it maintains a tolerably direct communication between even the innermost chambers of the spire and the external surface, a connexion which would have been cut off by the interposition of the "supplemental skeleton," had not this been provided with some such system of inter- vening canals. 207. One more fact remains to be noticed, which is of much interest as showing that the growth of the spire and that of the " supplemental skeleton" are to a certain extent independent of each other: I refer to the departure from the regular form that frequently shows itself in the later turns of the spire, which (so to speak) often " run Kg. IV. D E Abnormal specimens of Calcarina. wild" in a variety of strange modes, examples of which are so well represented in Fig. IV. A-F, that it is unnecessary to refer to them in other than these general terms. GENUS TINOPOEUS: HISTOEY. 557 The extension of spines from the whole surface of the disk, in the mode represented in Fig. I. A, may in like manner be regarded as a sort of " running-wild " of the supple- mental skeleton. Genus TINOPORUS. 208. A more remarkable example could scarcely be adduced, of the necessity of a thorough examination of the internal structure of the skeletons of Foraminifera as a guide to the determination of their true affinities, and of the danger of relying upon external characters alone, than that presented by the type I am next to describe under the name Tinoporus ; certain forms of which present so strong a superficial resemblance to many specimens of Calcarina, as to be unhesitatingly associated with them by every one who has not had his attention directed to the minute features of difference they present; whilst the plan on which it is constructed will be found to be essentially diverse. 209. History. My reason for adopting this name is as follows. In the ' Conchyliogie Systematique ' of DENTS DE MONTFORT (Paris, 1808), there is described and figured * under the name of Tinoporus baculatus a small polythalamous body, which he seems to have distinguished from the other varieties of Nautilus (Calcarina) Spengleri figured by MM. FICHTEL and MOLL, partly by the fewness of its spines, and partly by the difference of its structure as displayed in vertical section. And although his figure and description are alike inaccurate (the former, as has been pointed out by Messrs. W. K. PARKER and T. R. JONES f, being partly drawn from specimens of Calcarina), yet as I can scarcely doubt that he had before him a specimen of the type I am about to describe, I think it right to retain the distinctive designation he conferred upon it. The following is DE MONTFORT'S characterization of the genus : " Coquille libre, univalve cloisonnee et cellulee, spiree et lenticulaire ; tet granule exterieurement ; bouche semi-lunaire, placee vers la circonference et sur un des cotes ; dos carene, arme de quatre pointes au plus ; les deux centres bornbes et releves." Of the species T. baculatus, which served as his type of the genus, and of which he gives the fourth variety of the Nautilus Spengleri of FICHTEL and MOLL as a synonym, he says : " Cette coquille, qui pour nous est la tete (type 1) d'un genre nouveau et assez nombreux, ressembleroit a la nummulie ou camerine tuberculee et criblee, si elle n'etoit armee de trois pointes obtuses. Ces pointes sont interieurement sillonees et tuberculees a la maniere de quelques tubipores : la bouche de la coquille, placee sur un des cotes, est tres remarquable, en ce qu'elle est petite et formee en demi-lune : la spire est cachee et interieure. Les auteurs allemands que nous avons cites dans notre synonymie, y compterent au moins quatre-vingts cellules. Nous avons fendu cette coquille a demi par le milieu, afin de faire apercevoir la con- struction de I'interieur, qui, cellulee sur divers plans, nous conduit naturellement aux nummulies, mais elle en differe par ses bras ou pointes, qui sont constans, quoique leur * Toine i. p. 147. t Aiiuals of Natural History, Series 3. vol. vi. p. 34. 558 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. interieur presente une organisation qui n'est point cellulee mais tubulee. La couleur du tinopore bacule est blanche, flambee et teintee de jaune ; la coquille est entierement opaque. L'individu qui a servi de sujet a notre description venoit de la mer des Indes orientales ; on le trouva dans le sable dont etoit remplie une coquille du genre casque ; on rencontre encore les tinopores parmi d'autres coquilles microscopiques, sur les plages du golfe Arabique, ainsi que dans quelques eponges de la mer Adriatique. D'une pointe a 1'autre, le tinopore que nous venons de decrire a deux lignes de diametre." It is from the very distinct statement of DE MONTFOET (borne out by the vertical section rudely represented in his figure) of the cellulation of the interior of this organism on different planes, so as to give to its vertical section somewhat the appearance of that of a Num- mulite, that I am induced to believe that he had really distinguished Tinoporus from the type to which we now give the designation Calcarina. For, as we have seen, no such appearance is presented by vertical sections of Calcarina ; whilst, as will presently appear, Tinoporus is made up of several layers of cells superimposed one upon another ; and although its relation to Nummulite is really remote, yet the resemblance in aspect presented by vertical sections of the two may easily seem, to such as are unacquainted with the real meaning of their appearances, sufficiently close to justify the parallel. 210. I am not aware that any subsequent writer has adopted DE MONTFOET'S generic definition of Tinoporus, which seems to have been treated as one of his many valueless differentiations which systematists have agreed to disregard. In the ' Dictionnaire Universel d'Histoire Naturelle ' it is noticed as a synonym of Calcarina. 211. The plan of structure presented by Tinoporus differs so remarkably from any that has been yet described, as well to deserve being fully detailed. But it presents an additional feature of great interest, in the light which it throws upon the structure and character of the remarkable fossil genus Orbitoides, which, though first named by M. D'OEBIGNY, was first described by me*, and is scarcely less important in its geological relations than Nummulites itself. 212. Before proceeding, however, to the description of Tinoporus baculatus, I shall give an account of the structure of a simpler form of the same type, by which that of T. baculatus will be better understood. The form to which I allude has a much wider geographical range ; for though the largest and best developed specimens I have seen are those which I have obtained from Mr. JUKES'S Australian dredgings, yet I have met with it also in collections from the Fiji Islands, Mazatlan, and the Canaries ; and Messrs. PAKKEE and RTJPEET JONES (by whom this type has been recently noticed under what I cannot but consider the inappropriate designation Orbitolina-f) state that it occurs also in the East and West Indies, in the Mediterranean, and on the British coast as far north as Arran. To this type, which is destitute of the projections so remarkably characteristic of T. baculatus, the name of T. Icevis may be appropriately given. * On the Microscopic Structure of Nummulina, OrUtolites, and Orbitoides, in the Quarterly Journal of the Geological Society, February, 1850. t Annals of Natural History, Series 3. vol. vi, pp. 32, 33. GENUS TINOPOBUS: INTERNAL STEUCTUEE. 559 213. External Characters. The largest examples I have seen of Tinoporus Icems present the form of a short truncated cone, much resembling the lower half of a sugar- loaf (Plate XXI. fig. 1, #), having its margins rounded off, and attaining at the base a diameter of about one-tenth of an inch. The base commonly exhibits a slight central depression (fig. 1, V). Sometimes the cone is more depressed, and spreads out more widely at the base; and in this case the basal concavity is usually wider and deeper. The examples I have seen from other localities have for the most part a spherical or spheroidal shape ; but a careful examination will generally make it apparent that this shape is derived (so to speak) from that last mentioned, by the folding-inwards of its peripheral portion towards the centre of its lower surface, so as to leave a deep cavity at that part of the sphere, the relation of the two forms being very much like that which exists between the expanded pileus of an Agaricus, and the same pileus whilst still included within its volva. I cannot regard these diversities of form as possessing any specific value ; since they depend entirely upon mode of growth, and are not connected with any differences of internal structure. 214. In whatever form the Tinoporus Icevis may present itself, it is recognized by the absence of projection or angularity, and by a regular areolation over its whole surface, which a good deal resembles that of the cuticles of many leaves, the areolse preserving a tolerably constant average of size, but being very indefinite as regards form. The divisions are marked by very definite boundary lines; and in the interspaces between these, under a sufficient magnifying power, minute punctations may be seen. 215. Internal Structure. When the structure of this organism is examined by means of sections taken in different directions, it is found to be composed of an aggregation of minute chambers of nearly uniform size, which are piled one upon another in pretty uniform layers, each of these presenting an approach to a concentric disposition (figs. 2, 3). Although it is difficult to make out' with certainty the arrangement of the first-formed chambers, yet it is clear that as in other Foraminifera the point of departure is a sphe- roidal cell (fig. 3, a), which soon comes to be surrounded by a cluster of secondary cells (5, c) derived from it by gemmation. In what manner these are given off from the first, whether by a spiral or by a cyclical extension of the sarcode-body, I have not been able to satisfy myself, on account of the difficulty of precisely carrying the plane of section through this group of chambers. In T. baculatus I have been fortunate enough to do this in several instances, and have found that the early growth is unquestionably spiral (fig. 11), soon, however, giving place to the cyclical, as in those varieties of OrUtolites whose growth commences after this fashion (^[ 54) ; and whether this be or be not the case in T. Icevis, it is indubitable that before long the extension of the organism in diameter is effected by a budding-forth of new chambers from all parts of the circum- ference, not with such regularity, however, as to form distinct annuli as in OrMtolites. Whilst this extension is taking place peripherally, however, additional layers of chambers are formed, as in Orbiculina, above and below the central nucleus, meeting each other 560 DE. CAEPENTEE'S EESEAECHES ON THE FOKAMESTFEBA. on the equatorial plane ; and in this manner the increase of the organism in thickness is effected. The growth on the two sides of the equatorial plane, however, is seldom or never symmetrical ; and that of the more convex portion seems continually tending to overpower that of the opposite surface, so that the equatorial plane becomes more or less deeply concavo-convex. I have reason to believe that this inequality is due to the attachment of the flat or subconcave base to the surface of sea-weeds or zoophytes ; in virtue of which it will naturally happen that the free side will grow faster than the other. It is by an excess of this predominating growth that the spheroidal form is acquired, with its deep residual cavity, as just now described. 216. On more minutely examining the structure of the walls of the chambers and the mode of communication between their cavities, we find ourselves carried back to the lower type of Orbitolit.es and Orbiculina in this particular, that the septa are not double but single, and are formed of a simple homogeneous substance, presenting no vestige whatever of that fine tubulation which characterizes the dense almost ivory-like sub- stance forming the walls of the chambers in the more elevated forms of this group. It will be convenient to speak of the partitions between the chambers as horizontal, when their general direction is parallel to that of the equatorial plane, and as vertical when that direction is perpendicular to it ; their actual directions will of course vary with the curvature of the equatorial plane. The horizontal partitions or floors of the chambers are perforated by rounded apertures (Plate XXI. figs. 2, 3, 4) which closely resemble those of the shell of a Eotalia or a Planorbulina in their size and arrangement ; and these will allow of free communication, by pseudopodial threads of sarcode, between the segments that are lodged in the chambers piled one over the other in a vertical direction. The vertical partitions are much thicker, and are not thus minutely and regularly per- forated ; but they exhibit a small and variable number of large apertures (fig. 4, a, a), that lead into the adjacent cells which lie A in or near the same horizontal plane. I say in or near, because it is seldom if ever the case that the horizontal partitions or floors of two adjacent vertical piles of cells are on the same level ; and, in fact, the typical arrangement (though frequently departed from) &y_ seems to be, that there is an alternation in the levels of the floors of adjacent piles (as shown in the accompanying diagram, based on some parts of fig. 3), and that every chamber in any pile B normally communicates with two chambers in each of its adjacent piles A and c, by one passage above and the other below the floor that divides them. 217. The relation of this interesting type of structure to that of Planorbulina appears to me so clear that it can scarcely be questioned. For, as in that genus, the first-formed portion of Tinoporus Icevis will evidently consist of a flattened disk, consisting of nume- rous segments which are arranged in one plane, spirally in the centre of the disk, but clustered irregularly towards its circumference, and perforated on both sides with B GENUS TINOPOKUS: AFFINITIES. 561 numerous large pseudopodian foramina*. This relationship is admitted by Messrs. PARKER and RUPERT JONES; who state, however, that Planorbulince are differentiated by having " two or three tubular and margined apertures to each chamber, coarser pseudopodial pores, and no umbilical cells." Now I have shown that every chamber in the adult is connected with its adjacent chambers of the same or of alternate planes by two apertures ; so that at the free growing margin of the young disk there would pro- bably be at least two such pairs of apertures in the wall of every chamber. The difference in the size and number of pseudopodian pores is a very trivial character. And the superposed umbilical cells will probably be absent in the young Tinoporus, when as yet only a small number of rows of chambers have been formed around the central cell. I have, in fact, specimens in my possession which would be unhesitatingly characterized as Planorbulince by such as are unacquainted with the structure of the type we are considering ; yet which I cannot help regarding as in all probability young forms of Tinoporus, having been found in the same dredging, and presenting just the characters which I should expect from analysis of the structure of the adult to find in them. Moreover, I have Planorbulince whose early growth is so distinctly spiral as to correspond in every essential particular with the young of Rotalia. 218. It is not a little remarkable, however, that this organism should also be very closely related to a body of which the true nature has hitherto been doubtful, viz. the Millepora rubra of LAMARCK-^, the Polytrema miniacea of BLAINVILLE^. This grows parasitically upon shells, sometimes spreading over their surface in a laminated form, sometimes rising into a sort of stem and sending off branches. I have ascertained by examination of thin sections, that it is composed of minute chambers piled together very much in the manner of those of Tinoporus, and having the same kind of communica- tions ; and as Mr. PARKER has in his possession a specimen of a nearly globular form, attached to a projection of a bivalve shell, it may be questioned whether the difference between the two organisms is even of specific value. For the mode of growth which ordinarily characterizes each, shows a tendency to pass into that of the other ; Tinoporus Icecis occasionally flattening itself out and extending marginally, whilst Polytrema miniacea occasionally restricts itself within a compact spheroidal form. The probable relationship of Polytrema miniacea to the Foraminiferous type has been already suggested by the sagacity of Dr. J. E. GRAY ; but as he was not acquainted with the internal structure either of Tinoporus or of Polytrema, he could not make a more particular approximation. I should add that the bodies described and figured by Professor MAX. SCHULTZE|| under the generic name ofAcervulina appear to me to belong to the same type. I have in my possession a specimen growing round the stem of a * See Professor WILLIAMSON'S Monograph of the Eecent Foraminifera of Great "Britain, p. 57. t Hist. Nat. des Animaux sans Vertebras, troisieme edit., torn. ii. p. 309. J Manuel d'Actinologie, p. 410, pi. 69. fig. 16. Proceedings of the Zoological Society, April 27, 1858. || fiber den Organismus der Polythalamien, p. 67, plate 6. figs. 12-15. MDCCCLX. 4 E 562 DE. CAEPEXTEE'S EESEAECHES ON THE FOEAMINIFEEA. zoophyte, like the Acervulina acinosa of SCHULTZE, but with more of the compactness of Polytrema miniacea. 219. I now return to the form of Tinoporus on which the genus was originally esta- blished, the T. baculatus of MONTFORT; which agrees closely with T. Iccvis in the fundamental characters of its organization, but differs in being furnished with a variable number of radiating appendages that give it a strong resemblance to Calcarina. Of the specimens in my possession, the greater part present the aspect represented in Plate XVIII. figs. 2, 3, 4, 5, and on a larger scale in fig. 6 ; and these were collected from coral reefs in various parts of the Polynesian Archipelago, my earliest acquisition of them, however, having been from the contents of the stomach of an Echinus taken on the coast of Borneo, which were kindly put into my hands by Dr. J. E. GRAY. I am informed by Mr. DENIS MACDONALD that on certain coral islands which he has par- ticularly examined, these organisms are so extraordinarily abundant, that they accumulate in the lagoons in regular strata, commonly alternating with strata of Orbitolites. The more massive and ruder forms represented in Plate XVIII. figs. 7, 8, 9, 10, occur in Mr. CUMING'S Philippine collection. 220. External Characters. The typical form of the central portion of T. baculatus (Plate XVIII. figs. 2-6, and in section in Plate XXI. fig. 7) may be considered as an oblate spheroid; sometimes, however, it is nearly spherical, and sometimes it is much flattened out, especially when the body extends itself into the radial prolongations, as in Plate XVIII. fig. 4. Its surface is divided into areolse (fig. 6) very much as in . T. Icems; but the angles of junction of the partitions between the areolse are very commonly occupied by rounded projecting tubercles, strongly resembling those of Calcarina, The number and size of these tubercles vary greatly among different individuals, as will be seen on comparing figs. 2 and 4, Plate XVIII*. From the marginal portion of the central disk there spring a variable number of conical pro- longations having the furrowed surface of those of Calcarina ; and these appear seated (so to speak) upon extensions of the central disk itself, which is sometimes so deeply subdivided at its margin as to resemble the body of a Star-fish (Plate XVIII. figs. 4 and 7), the areolar division being continued nearly to the extremity of each ray, and its point only being formed by the furrowed prolongation. These appendages are usually from 4 to 6 in number ; I have occasionally seen only 2, and in no case have I met with more than 8. They usually diverge in or near the equatorial plane ; but they some- times come off in very different directions (Plate XVIII. figs. 8, 10). 221. Internal Structure. The general organization of T. baculatus, brought into view by sections taken in different directions, does not differ in any essential respect from that of T. Icevis ; the origin of the whole aggregation of chambers in a central cell, their subsequent multiplication both horizontally and vertically, and their methods of com- munication in both directions, being all the same. As already mentioned, I have very * They seem to be altogether wanting in the Philippine specimens, being apparently replaced by a multitude of small spines, which give to their surface a hispid aspect. GENUS TIIVOPOEUS: INTEENAL STETJCTUEE. 563 distinct evidence, in sections of this species, of a spiral commencement, soon giving place to an irregularly-cyclical growth ; and sometimes the first-formed portion (Plate XXI. fig. 11, ft) bears so close a resemblance to the innermost part of the spire of Calcarina, that in this earliest stage of their growth the two types could not be distinguished from each other. Thus Tinoporm baculatus seems to bear the same relation to Calcarina, that T. Icevis does (through Planorbulina) to the Eotaline type (^[ 217). The essential difference between T. baculatus and T. Icevis consists in the possession by the former of a " supplemental skeleton," which presents itself under two principal aspects. The piles of chambers extending vertically from the equatorial plane towards the two surfaces of the spheroid (Plate XXI. fig. 7) are partially separated by the interposition of pillars of solid shell (fig. 8, a) ; and it is by the projection of the summits of these pillars (as in Calcarina) that the tubercles of the surface are formed. The spines also, which form the extremities of the radiating prolongations, belong to the same system ; and they are shown, by sections of the Philippine type that pass in a favourable direction (fig. 6), to be extended from a solid framework which begins to be formed even with the first convo- lution, and which adds greatly to the thickness of the partitions between the chambers, giving off a multitude of minute spines from their external surface (fig. 10, c, c). This framework is penetrated by a canal-system, which not only forms passages through the solid axis that is prolonged into the spines (fig. 10, a, a), but also extends itself into the partitions between the chambers (fig. 9, b, b). The canal-system of the solid axis, moreover, communicates freely with the cavities of the chambers that are adjacent to it, as shown at fig. 10, b, b. These chambers are arranged around it with considerable regularity, as is shown in fig. 5, which is a transverse section of the base of one of the radiating prolongations, showing the solid axis with its radiating canals, surrounded by three rows of chambers. It would seem as if, in the Polynesian variety of T. baculatus, the material of the supplemental skeleton were appropriated rather to the formation of the solid pillars than to that of a solid axis for the radiating prolongations ; the latter being much less conspicuous than it is in the Philippine specimens, and sometimes appearing to be deficient altogether except at their extremities. On account of the variability of these differences, however, I cannot regard them as of any essential value. 222. If any further evidence had been required as to the essential relation between the " canal-system " and the " supplemental skeleton," I think that it must be satisfac- torily furnished by the comparison of the two species of Tinoporus now described. For in T. Icevis it is obvious that the system of communications which exists between its chambers is adequate for all the ordinary wants of an organism of this type, the structure of which is uniform throughout. But when, as in T. baculatus, an additional framework of solid walls is interposed in the midst of the building, for the support of the extensions into which it is prolonged, a special system of passages, originating from the cavities of the adjacent chambers, and extending throughout the solid framework, is provided for its nutrition. 4E2 564 DE. CARPENTER'S RESEARCHES OX THE FORAMINIFERA. 223. With reference to the relationship between Tinoporus and the fossil genus Orbitoides, to which allusion has already been made, I shall here content myself with stating briefly that if the chambers of the equatorial plane of T. Icevis were distinctly differentiated from those of the layers springing from it on either side, it would come to resemble in general conformation the simpler type of Orbitoides known as the 0. Mantelli ; and further that if, with this modification, there were also introduced the solid pillar-system of T. baculatus, we should have a corresponding resemblance to 0. dispansus*. The metamorphic condition of the shell of the fossil Orbitoides has hitherto prevented me from determining with certainty whether its elementary structure bears most resemblance to the inferior type presented by Tinoporus, or to the more elaborate structure of that of Cycloclypeus, to which type also it seems to be related^. On these points, however, I shall enlarge more fully elsewhere. Genus CARPENTERIA. 224. Of all the Foraminifera collected by Mr. COMING in the Eastern Seas, the last which I have to describe is perhaps the most interesting ; since the type of structure which it presents is not only altogether new, but seem to furnish the connecting link (which had been previously rather suggested than supplied by Thalassicolla and its allies) between SPONGES and FORAMINIFERA, two groups which accord most remarkably in their grade of organization, whilst they differ no less remarkably in plan of structure. 225. The larger number of the specimens of this type in the collection of Mr. CUMING are attached to the surface of a piece of Forties (coral) ; other specimens, however, are adherent to the shells of Pecten and Cardita ; and the attention of Mr. W. K. PARKER having been directed to these curious organisms, he has met with them on the surface of other bivalves, especially Chama gigas. It is not a little remarkable that the strong external resemblance which they present to the shells of certain sessile Cirripedes should have led not only Mr. CUMING, but other experienced conchologists, to regard them as belonging to that group. Their true nature was first suspected by Dr. J. E. GRAY, who was led by his study of them to consider them as the testaceous envelopes of a Rhizopod intermediate between Sponges and Foraminifera ; the grounds on which he came to this conclusion being, that he found the shell to be multilocular and minutely foraminated * I should take this opportunity of stating that in my former description of Orbitoides (Journal of the Geological Society, Feb. 1850) I fell into the same mistake in regard to these pillars, that I did in regard to the analogous structure in Nummulites ; regarding them as having been passages filled with solid calca- reous matter in the process of fossilization, an error which was pointed out to me by Professor WILLIAMSON at the time, and of which I have since come to be fully satisfied by the examination of the recent analogues. t The figure given by Professor EHBENBERG, in his remarkable memoir already referred to, " tlber den Griinsand und seine Erlauterung des organischen Lebens," plate iv. fig. 8, and by him designated as the internal cast of OMtoidesjavanicus, will be seen on comparison to present a most remarkable correspondence with figs. 10, 11, 12 of Plate XXIX. (Phil. Trans. 1856) illustrating my description of Cycloclypeus. GENUS CAEPEXTEEIA : EXTEENAL CHAEACTEES. 565 like that of certain Foraminifera, whilst the fleshy substance occupying its chambers is strengthened with spicules like those of Sponges. Hence he considered this organism in the light of a Sponge enveloped in a shelly case with a single terminal oscule. My opinion as to its character having been asked by Dr. GRAY, I soon found reason to accord with him in his general idea of its affinities ; the structure of the shell being most characteristically foraminiferous, whilst the substance occupying its chambers is no less characteristically spongcous. In communicating this result, however, to Dr. GRAY, I thought it right to suggest the possibility that this spongeous substance might be para- sitic ; the tendency of certain Sponges to find their way into even very minute fissures and passages having been observed by me in my researches on the structure of the Shells of Mollusks. Dr. GRAY agreed with me in thinking this improbable, for reasons which will be presently stated ; and he communicated a general account of this new type (to which he did me the honour of giving my own name) to the Zoological Society* ; at the same time expressing the desire that I should include a fuller account of its forma- tion and structure in my communications on the Foraminifera to the Royal Society. With this desire I have now the satisfaction of complying. 226. External Characters. The ordinary external aspect of Carpenteria, as represented in Plate XXII. figs. 1, 10 (taken from a group on the surface of Porites, of which the individuals are in close proximity to each other), at once suggests a resemblance to the Balaniform type ; the shell being conical, attached by its broad base, furnished with a single definite aperture at its apex, and presenting an appearance of irregular divisions into triangular segments, which might easily be supposed to be "valves" bounding a single undivided cavity. On breaking into the interior of the shell, however, it imme- diately becomes apparent that the foregoing resemblance is superficial only ; the entire cavity of the shell being divided into numerous chambers, which are completely separated from each other by septa, whose lines of junction with the external wall (indicative of the successional additions which the shell has received) give rise to the appearance of valvular divisions. And a closer examination reveals that these chambers are disposed upon a spiral type, each whorl completely investing (save on the adherent base) that which preceded it, so that only the external wall of the last whorl is anywhere seen on the surface. In the specimen represented in fig. 5, which is one of the isolated individuals occurring on the valve of a Pecten, the shell has a much less regular form, owing to the more or less complete divergence of the basal portions of the chambers of the last whorl, and the partial subdivision of some of those chambers into lobes which exhibit the like divergence. The shelly surface of the wall of each chamber presents a somewhat areolated aspect, which depends -upon its being raised into a multitude of low rounded elevations (fig. 8) ; and under a sufficient magnifying power these areolse are seen to be pretty uniformly marked with minute punctations (fig. 6). The form of the aperture at the summit of the cone, of which two examples are shown in figs. 13 & 14, presents a striking resemblance to that of the aperture of the Milioloid Foraminifera. * Proceedings of the Zoological Society, April 27, 1858. 566 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. 227. Internal Structure. On breaking-away a portion of the external wall of the last-formed chamber, so as to lay open its interior (as shown in fig. 8), we find that its cavity is closed-in on every side by its shelly walls, except where it has communications (b, c) with the apical aperture ; and each principal chamber is partially subdivided by a system of shelly septa, of which some are more and others less complete. The more complete of these secondary septa (fig. 8, e, e\ e*) resemble the principal septa (d, d l , (?) which separate the cavities of the chambers, in running from the base towards the apex of the cone ; they divide the lower portion of each chamber into three, four, or more digitations, which are sometimes marked by an external lobulation, as shown in fig. 5 ; they stop short, however, about half or two-thirds of the way towards the apex, leaving the upper third or half of the chamber undivided. Some of these septa do not reach the opposite surface of the chamber; and the least complete (f,f l ,f) form a sort of network of ridges slightly projecting from the inner surface of the outer wall into its cavity (as shown in vertical section at b, b, b, fig. 15), and there marking out an areola- tion which corresponds to that of the external surface. The areola3 of the internal surface, however, are concave instead of convex ; and the punctations, which are wanting on the ridges, are set more closely on the depressions between them. The reason of this peculiarity in their distribution will be presently seen (^[229). 228. The cavity of the last chamber communicates with the external orifice by a passage of considerable size ; and the wall of this passage is distinctly continued as an irregular ring around the apical aperture, so that this aperture may be considered in one sense (as described by Dr. GRAY) to belong to the last chamber alone. But it would be more correct to say that each cell as it is formed conceals, than (with Dr. GRAY) that it closes, the aperture of the preceding cell ; for a careful examination shows that the external aperture or vent is the termination of an irregular vertical canal, formed by the superposition of the oral rings of successive cells; and that through this canal the previously formed cells retain their original connexion with the exterior. In some of Mr. PARKER'S specimens, the oral ring is extended upwards into a tube or siphon at least equal in length to the radius of the cone. The general disposition of the chambers around the central canal is well shown by sections of the cone taken parallel to its base (fig. 7) ; such sections, however, may only bring into view the last or superficial whorl ; and they will generally show only one or two chambers in communication with the vertical canal, the communicating passage of each chamber being on a different plane. 229. The foramina which pierce the outer wall of each chamber are of considerable size, as compared with the minute tubuli of Cycloclypeus and Operculina (see - f[^[ 103, 154), and they are not nearly so closely approximated; -in both respects they correspond closely with the foramina of the ordinary Rotalioe and Globigerince. In fig. 12 they are shown as they appear in a section traversing the wall somewhat obliquely to its surface, whilst in fig. 15 they are shown as they appear in vertical section ; and in each case they are seen to present an annulated appearance, which is due to constrictions of the tubes at tolerably regular intervals. These tubes generally pass direct from one surface to the GENUS CAEPEXTEKIA : STEUCTITEE AND AFFINITIES. 567 other ; but at , a, fig. 15, it is seen that in the neighbourhood of the ridges which project from the inner wall into the cavity of the chamber, the tubes either bend or incline themselves in such a manner, that, whilst their external orifices are pretty uniformly distributed (fig. 6, ), their internal orifices do not show themselves upon the ridges, but are crowded together along their bases (fig. 6, b, and fig. 9). The septa, whether primary (separating the chambers from each other) or secondary (partially subdividing the chambers), are obviously formed by a doubling-in of the outer wall, so as to make each septum consist of two laminae (fig. 12, a, a) ; this is seen also in sections of the incomplete septa (fig. 12, b), as well as of the ridges which may be considered as rudimentary septa (fig. 15, b, b, I}. The two layers sometimes separate from each other, as shown in these figures, so as to leave intraseptal spaces ; and these form a tolerably regular canal-system, which may be traced throughout the network of ridges that covers the inner wall of each principal chamber, and, through the primary septa, into the ring that surrounds the vertical canal (fig. 7, g, g'). 230. Whilst, therefore, the general plan of conformation of Carpenteria seems to differentiate it strongly from that of the ordinary Foraminifera, so close an alliance to them is indicated by the minute structure of its shell, that it becomes of special import- ance to determine whether its peculiarities are original, or whether they are acquired during the progress of development. I have fortunately been enabled to determine this point by the comparison of several specimens in different stages of evolution, and by the removal from the older specimens of one whorl after another until the original nucleus was arrived at (an operation which has been very dexterously performed for me by my draughtsman Mr. GEORGE WEST); and 1 can state without hesitation that the early condition of this apparently anomalous organism accords with that of the Helicostegue Foraminifera generally, its approximation being the closest to Eotalia in general form, but its tendency being rather towards Globigerina in this particular, that its chambers do not seem to communicate directly with each other, but that each has a separate external orifice directed towards the umbilicus. Various aspects of this first-formed portion of the shell, two of them showing the animal substance contained in the chambers, are seen in Plate XXII. figs. 2, 3, 4. Now supposing that a Globigerina were to grow in such a manner, attached by one of its surfaces, that the walls of its successively- formed chambers came into mutual contact, and that these chambers were so shaped and so piled one on the other as to give to the entire shell a conical form, each chamber opening by its own separate orifice into an umbilical funnel, we should have the essential type (so far as its shell is concerned) of Carpenteria ; and this is really the mode in which the latter type is superinduced upon the former, as the development of the organism advances. It is further interesting to observe that the great size of the chambers which form the superficial whorl of Carpenteria, has every appearance of being due to the deficiency of that complete segmentation, in the later stages of growth, which characterizes the earlier ; for every one of the loculi marked out by the ridges projecting into the interior corresponds so closely both in size and general aspect 568 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. with an entire chamber of the earlier whorl, that the areolation of the outer wall may be regarded as a sort of attempt at that complete subdivision of the cavity, which we have seen to be fully carried out in Heterostegina (^f 114). 231. Thus, then, it appears that not only in the minute structure of its shell, but also in its general plan of conformation, Carpentaria essentially approximates to the ordinary Foraminiferous type, its affinity to the Spongiada being rather apparent than real ; for although each cone does at first view seem to resemble the papilla of a Sponge enclosed in a shelly case with a terminal oscule, yet the internal structure of that cone does not bear out that resemblance. The link of affinity, however, seems to be supplied by the spongeous character of the animal substance which occupies the chambers (fig. 9) ; this (according to the evidence aiforded by the dry specimens which alone I have had the opportunity of examining) not only possessing far more consistence than the sarcode- body of the Foraminifera, but being supported, in the large chambers at least (fig. 6), by sponge-like spicules (fig. 16), whose form resembles that of the simplest spicules of Halichondria, and whose composition is siliceous. 232. The idea which had occurred to myself of the possibly parasitic nature of this sponge-like substance, has been very strenuously advocated by Mr. BOWERBANK, on the ground of the frequency with which the surface of dead coral and the valves of living as well as dead shells are covered with Sponges, and the consequent probability that any multilocular organism growing on such surfaces would be so penetrated by the sponge that all its chambers would be filled by the parasite. The following consider- ations, however, seem to me strongly to militate against such a view : 1st. That neither on the surface nor in the substance of the specimen of Porites covered with the cones of Carpenteria, nor on that of the valves of the Pecten and Cardita on which isolated specimens of Carpenteria occur, is there the least trace of spongy structure : 2nd. That notwithstanding this marked difference in their habitats, all the specimens of Carpenteria yet examined have their cavities occupied by the same spongy substance : 3rd. That a firm brownish yellow substance of far greater consistence than the sarcode of Forami- nifera, is found to occupy even the smallest and earliest chambers of Carpenteria (figs. 2, 4, a), filling them so completely that it can scarcely be supposed to be anything but the animal body properly belonging to them ; and that although the substance in question is there destitute of spicules (the chambers being too small to accommodate them, as will be seen by the comparison of figs. 4 and 16, allowing for the difference of magnifying power), yet it is obviously the same with that in which spicules are copiously imbedded in the larger and later chambers : and 4th. That notwithstanding the multitudes of sections of various Foraminiferous shells which 1 have made during the la.st ten or twelve years, I have never found their chambers to be occupied by a parasitic sponge of any description. I may add to these considerations the fact men- tioned to me by Mr. DENIS MACDONALD, that he has met with various forms of branching Sponges*, possessing a peculiarly solid calcareous skeleton, and in many respects * These specimens were collected during the voyage of H.M.S. ' Herald ' in the Australian Seas, and CONCLTJDESTG SUMMAEY. 569 appearing to present the same kind of transition from Sponges towards Foraminifera, that, if my view be correct, is afforded by Carpenteria from Foraminifera towards Sponges. 233. The above reasons appear to me so strongly in favour of that idea of the essen- tially spongeous nature of the animal of Carpenteina, which had been from the first entertained by Dr. J. E. GRAY, as to leave me little room for hesitation in the belief that such is its real character; so that, until the contrary shall be proved, we seem justified in regarding this curious organism as a Rhizopod which in virtue of its shell is foraminiferous, whilst in virtue of the animal body which that shell contains it is spongeous. CONCLUDING SUMMAEY. [Added during the printing of the Paper.] 234. In bringing to a close the present series of Researches on the Organization of the Foraminifera, I think it desirable to combine a summary of the most important results which I have obtained, with some remarks on their bearing, not merely on the method to be followed in the attempt to frame a natural classification of this group, but upon some of those higher questions relating to the origin and value of differential characters generally, which have recently been brought prominently under the consider- ation of the scientific world. In so doing it is my desire to confine myself strictly to the scientific and practical aspect of these questions ; seeking in the first place to determine, on the legitimate basis of induction, what general principles may be either certainly or probably educed from the comparison of the large body of facts which have been brought together by myself and others as regards the mutual relations of Foraminifera ; and then briefly inquiring how far the results of similar comparisons made upon other types of organized structure justify the extension of the same principles to the Animal and Vegetable kingdoms at large. 235. It may be well for me to advert in limine to certain features in this inquiry, that render the group to which it relates singularly adapted for a comparison at once minute and comprehensive amongst a wide range of individual forms. The size of the greater part of these organisms is so small, that many hundreds, thousands, or even tens of thousands of them may be contained in a pill-box ; and yet it is usually not too minute to prevent the practised observer from distinguishing the most important peculiarities of each individual by a hand-magnifier alone, or from dealing with it separately by a very simple kind of manipulation. Hence the systematist can easily select and arrange in series such of his specimens as display sufficient mutual conformity, whilst he sets apart such as are transitional or osculant ; and an extensive range of varieties may thus be displayed within so small a compass, that the most divergent and the connecting forms are all recognizable nearly in the same glance. I am not acquainted with any other were transmitted by Mr. MACDOXALD to the Admiralty. I have not yet succeeded, however, iii obtaining a sight of them. MDCCCLX. 4 P 570 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. group of natural objects, in which such ready comparison of great numbers of individuals can be made ; and I am much mistaken if there be a single specimen of Plant or Animal, of which the range of variation has been studied by the collocation and comparison under one survey of so large a number of specimens as have passed under the review of Professor WILLIAMSON, Messrs. PAEKEE and KUPEBT JONES, and myself, in our studies of the types to which we have respectively given our principal attention*. 236. The general fact which I desire to bring prominently forward as the result of recent investigations into this group, is, that in all the types possessing a wide geogra- phical distribution, which have been specially studied by myself or by others, the range of variation has also been very wide ; so that not only what would commonly be considered as specific, but such as have been regarded as generic, and in some cases even as ordinal differences, present themselves among organisms, which, from the intimacy of the relationship that is evinced by the gradational character of those differences as well as by the variations presented by the several parts of one and the same organism, must in all probability have had a common origin. And it appears to me to be a justifiable inference from this fact, that the wide range of forms which this group contains is more likely to have come into existence as a result of modifications successively occurring in the course of descent from a small number of original types, than by the vast number of originally distinct creations which on the ordinary hypothesis would be required to account for itf . 237. The greater part of my First Memoir was devoted to the investigation of the single type Orbitolites ; and I there showed that not only as regards the size, shape, and other external characters of the organism as a whole, but even as regards the size and form of its elementary parts (in which greater constancy might be expected), is there so great a variation, the most marked diversities being apparent even in different parts of the same individual, that all attempts to found specific distinctions upon such variations are utterly futile. But further, I showed that a distinction on which almost any Natu- ralist would feel justified in relying as of at least specific if not of generic value, that between the simple type in which the chambers are arranged on only one plane, and the complex type in which there are two superficial planes more or less strongly differentiated from the median, is no less invalid. For although these types are usually distinguish- * I have the authority of M. DESHATES for the belief, that the excessive multiplication of generic and specific distinctions which so greatly impairs the value of the late M. D'ORBIGKT'S labours upon this group, was due to his having based these distinctions upon specimens selected for him as typical, and to his having disregarded the transitional forms which any large collection of these organisms is sure to exhibit in abundance, thus, to use the admirably discriminating phrase of the late Prince of CANINO, " describing specimens rather than species." t In order to avoid misapprehension, I would here remark that the production of any organism seems to me just as much to require the exertion of Divine Power when it takes place in the ordinary course of generation, as it would do if that organism were to be called into existence de novo ; the question being in reality, whether such exertion takes place in the way of continuous exercise according to a settled and com- prehensive plan, or by a series of disconnected efforts. CONCLUDING SUMMARY: EXTENT OF EANGE OF VARIATION. 571 able, the one from the other, without the least difficulty, yet they are often combined in the same individuals, and this in such a variety of modes, that the transition from the simple to the complex may be clearly seen, from the comparison of a sufficient number of specimens, to be by no means attributable to a mere advance of age. Further, having been furnished (by the kindness of Mr. H. J. CARTER) with specimens of the Scindian fossil which presents the characters ascribed by M. D'ORBIGNY to his genus Cydolina, I am now able most fully to confirm the suggestion I threw out on a former occasion (*f[^[ 49, 70), that this genus is founded on a mere variety of Orbitolites, in which the character of the surface-marking is more than ordinarily cyclical. 238. Not merely, however, does the range of variation of this type confound the ordinary distinctions of systematists in regard to species and genera; it extends also to that difference in plan of growth, which has been assumed by M. D'ORBIGNY to be of such fundamental importance as justly to constitute the essential difference between his two orders Cyclostegues and Helicostegucs. For, as I have shown, although Orlitolites is typically cyclical from its commencement, yet specimens frequently present themselves in which its early development has taken place so completely on the helical plan, that if such had been collected before their assumption of the cyclical mode of growth, their essentially Cyclostcc/ue character would not have been suspected. 239. Again, I have shown that a parallel variation is displayed by the genus Orbiculina, whose ordinarily helical plan of growth has caused M. D'ORBIGNY to range it among his Helicostegues, notwithstanding that in fully-developed specimens its mode of growth is not unfrequently cyclical. The occasional exchange, in this type, of one plan of increase for the other, at an advanced period of life*, of which exchange I think I have given adequate evidence (f[^f 85-87), is a fact which seems to me of very high interest, being much more decided in its nature than the corresponding change in Orlitolites. For whilst in the latter the tendency of the spiral form (whenever it presents itself) to pass into the cyclical, is apparent almost from the beginning, and the change is never long postponed, the helical plan is that on which the growth of the former not only com- mences, but continues to be carried on, often through the entire period of its increase. 240. It is important to remark that in each case the metamorphosis is simply due to the very rapid opening-out of the mouth of the spire, its two lateral extremities extending themselves round the shell on the one side and on the other, until they meet and completely enclose the portion previously formed (just as the lobes of the mantle in the adult Cyprcea spread themselves round the shell until they meet on its dorsum) ; and also that the mutual relations of the chambers of the shell and of the segments of the animal body which they contain, remain essentially unchanged. Again, it is a point of no mean significance, that when an Orbiculina has undergone this change, the outer * It has been remarked by Messrs. PAEKEB and BTJPEBT JONES, that whilst the assumption of the cyclical form in Orbiculina may often be the result of the continued growth of individuals under favourable circumstances, small starved forms also frequently take on the cyclical condition, leaving the young sub- lenticular stage without passing through the aduncal. See Annals of Natural History, March 1860, p. 181. 4F2 572 DE. CABPENTEB'S EESEAECHES ON THE FOBAMINIFEEA. or cyclical portion of its disk can in no way be distinguished from that of OrMtolites ; and that the only difference between these two types which has any semblance of validity, is the absence in OrMtolites of those successive encasings of the central nucleus, the presence of which seems to be a constant feature in Orbiculina. It is to be observed, however, that these successive encasings are due entirely to the extension of the later whorls of the spire over the earlier ; and they are no longer formed in Orbiculina when the helical mode of growth gives place to the cyclical. Hence it seems not unfair to surmise that if the helical growth of an aberrant OrMtolites were to continue until its spire had made several turns, instead of stopping before the completion of one, its nucleus would receive successive investments from successive whorls, just as in the typical Orbiculina ; and the only difference between these two types would thus vanish. On the other hand, if the helical growth of an Orbiculina were to give place to the cyclical at an unusually early period, the central nucleus would receive no investment, and would present the flatness by which that of Orbitolites is characterized when com- pared with that of the typical Orbiculina. 241. I cannot but believe that such as may have followed me through the details of my previous descriptions, will be disposed to agree with me in thinking it justifiable to assume that such a range of variation as to the period of the change in plan of growth, would be only analogous to that which both these types present in so many other par- ticulars ; and hence that the idea of the derivation of Orbitolites and Orbiculina from the same original is scarcely less probable than that of the derivation of the helical and cyclical types of Orbiculina, or of the simple and complex types of Orbitolites, from a common parentage ; particularly since, as was formerly pointed out (^[ 90), both types present analogous modifications in geological time. 242. Let us now apply the same mode of inquiry to Aheolina. It has been shown (^[ 93, 94) that this organism is closely allied in every other respect than its geome- trical plan of growth to the types we have just been considering; the structure of the shell and its relations to the contained sarcode-body, and the relations of the segments of that body to each other and to the external world, being essentially the same in them all*. Now, however improbable it might seem at first sight, that an OrMtolites which extends itself as a flat or biconcave disk by successive concentric growths, and an Aheolina acquiring a fusiform shape by successive turns round a progressively elongating axis, should have had a common original, yet when the intermediate links are duly studied, a continuous gradation is found to be established. For, as has just been shown, a longer continuance of the helical mode of growth in which OrMtolites often commences, would really produce an Orbiculina with its centre so invested by successive * I may take this opportunity of stating that the description which I based on the' examination of sections of the shell has been fully confirmed by the internal casts obtained by Messrs. PAEKEB and EUPEIIX JOSES from specimens whose chambers had been fdled by an infiltration of silicate of iron (^J 178) ; which casts most accurately represent the form of the sarcode-body and of its individual segments with their connecting stolons. CONCLUDING STJMMAKY: EXTENT OF EANGE OF VAEIATION. 573 whorls as to form a vertical linear axis ; and we find this axis in Orbiculina sometimes equalling in length the diameter of the spire, so that this organism at an early stage of its growth may be nearly spheroidal *. Now among the various types of fossil Alveolince, there are some whose shape, instead of being fusiform like that of the recent type I have described, is almost identical with that of a spheroidal Orbiculina ; and the general structure of two such organisms will be so nearly identical, that I cannot see any difficulty in referring them to a common original. And when we examine a series of such fossil types, we see that they present a wider and wider divarication from the Orbiculine type in this one particular alone, that whilst the later growth of Orbiculina tends .to liken it to the discoidal Orbitolites, that of Alveolina tends to the continual elongation of its vertical axis, a difference which the analogies of the Foraminifera generally would indicate to be one of far too small account to be fairly adopted as a ground of original distinction f. 243. In the assemblage of forms which I have thought myself justified in re-assem- bling under the designation Peneroplis, we encounter another remarkable series of variations, the principal of which have given occasion to the formation of the two additional genera Dendritina and Spirolina. With an exceedingly close conformity in the texture and in the superficial markings of their shells, as well as in their general plan of growth, we observe a marked diversity in the form and proportions of the spire, especially in the later stages of its growth, and a still greater divergence in the form and disposition of the septal apertures. For in the type to which M. D'OEBIGNY restricts the generic designation Peneroplis, we usually find the spire rapidly widening and becoming proportionally compressed in each succeeding convolution ; whilst in that which he distinguished as Dendritina, the spire widens but slowly, whilst increasing rapidly in turgidity. Further, in the one type as in the other, the later extension is often in a straight line, instead of continuing to follow the spiral course ; and on this variation alone, which is of no account whatever among Foraminifera (as will presently appear, ^[ 255), has been erected the genus Spirolina. Now in the typical Peneroplis, the septal plane presents a linear series of minute rounded pores ; whilst in the typical Dendritina we find in their place a single large orifice with radiating extensions ; the difference between these two modes of communication being as great as we find between almost any two types of Foraminifera whatever. Yet I believe that no one who will go through the details of the evidence I have collected from the study of transitional forms, will have any doubt that Peneroplis and Dendritina have had a common progenitor, and that the peculiarity in the mode of septal communication that characterizes each is intimately related to the compressed or turgid form of the spire in either case ; whilst the different forms of the Spirolina type, among which we find the most remarkable * See Philosophical Transactions, 1856, Plate XXVIII. fig. 8. t In this view of the relation of Alveolina to Orliculina I am supported by Messrs. PAKKEU and EUPEET Jo^E8, who remark that Alveolina " may be said to represent a small thick Orliculina drawn out trans- versely at its umbilici." Ann. of Nat. Hist., March 1860, p. 182. 574 DE. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. transitional conditions of aperture, are so obviously related to both the foregoing, that no reasonable doubt can exist as to their derivation from them. Now the geographical distribution of the two fundamental types is so far different, that where one prevails the other is either absent altogether, or presents itself under a modified form ; and thus we seem justified in the belief that, whether either of them has been derived from the other, or both have been derived from some intermediate form (such as that which seems common alike to the young of both), the modifications which have given rise to the marked differences they now exhibit are mainly due to diversities in the external con- ditions under which they have been respectively propagated. 244. But to what other type does Pencroplis itself present the closest approximation 1 By systematists in general the intimate relationship which I have shown it to possess to the helical type of Orbicullna (^f 130) has been so slightly regarded, that it has been considered as at least equally related to the Operculine type ; and yet, as I shall pre- sently show, these two types are removed from each other in all the most essential features of their structure, as far as any two polythalamous Foraminifera can be. And the idea of the derivation of Peneroplis from the same stock with OrMculina, seems justified by the fact that the young forms of the two are frequently so alike as not to be distinguishable by external characters alone, whilst their internal difference consists only in the presence or absence of the secondary or transverse septa, a character which I have shown reason to regard as variable in this group* (^[ 130). 245. Notwithstanding, therefore, the apparently wide divergence of the cyclical Orbi- tolitcs, the helical Orbiculina, the fusiform Alveolina, and the simply-chambered Pene- roplis and Dendritina, these several types must be regarded as most intimately related to one another; and that relationship seems to me much more likely to have arisen from a common ancestral descent, than from the original creation of independent types capable of graduating into each other so continuously as almost to assume each other's characters. It is very important to remark that they all possess that peculiar texture of shell, which is designated by Professor WILLIAMSON as porcellanous ; presenting an opake white hue when seen by reflected light, but a rich brown or amber colour when seen by light transmitted through thin natural lamella? or artificial sections. This substance is entirely structureless, and possesses no great density or tenacity. Moreover in all the foregoing types, each of the septa intervening between the chambers consists of only a single layer ; the passages of communication between them are for the most part so large and free, that the segments of the sarcode-body are but very imperfectly isolated from each other ; and, as might be anticipated from this incompleteness of separation, it is here that variations in the mode of communication between the chambers seem to be of * My statement on this point is fully confirmed by Messrs. PAEKEB and RTTPEET JONES ; who state that " not unfrequently, feebly- developed peneropliform varieties, as well as good-sized adunciform specimens, occur, in which the long narrow chambers are at times simple and undivided, being occupied by transversely elongate lobes of Barcode, instead of numerous minute subcubical blocks." See Ann. of Nat. Hist., March 1860, p. 180. CONCLUDING SUMMABY: EXTENT OF BANGE OF VABIATION. 575 least account. It is in this type that we recognize the nearest approximation towards such forms as Thalassicolla, which seem to connect Orbitolites with Sponges (^[ 67); while the relationship which Orbiculina and Peneroplis have been supposed to bear to the ordinary Helicostegues, being dependent only on plan of growth, and being utterly at variance with the essential characters of the two groups, must be regarded as one of analogy, not of affinity. Looking to the evidence already adduced in regard to the prevalence of particular modifications of Orbitolites in particular localities (^[ 62), and to the influence of the geographical distribution of the Peneroplis type upon the modi- fications it presents (\ 138), we seem justified in extending the same view to those larger (though not more essential) differentiations which these types must have undergone on the hypothesis of their derivation from the same original. The following may be suggested as the mode in which the existing forms might thus have diverged from each other and from their primary type. Orbiculine Type diverging into A^ . Peneroplis Orbiculina . Dendritiua, Peneroplis Orbiculina, Alveolina Dendritina, Spirolina, Pcneroplis. Orbiculina, Orbitolites, Alveolina. 246. Passing on, now, to an essentially different group, that which includes Nummu- lites and its allied forms, we find that the relation of the discoidal Cycloclypeus and the helicine Hetcrostegina is of essentially the same nature with that of Orbitolites and Orbiculina (^[^[113, 116); the minute structure of the shell and the physiological condition of the sarcode-body being essentially the same in the two organisms, and the only important divergence between them being in their plan of growth. From the rarity of Cycloclypeus, all the known specimens of which have been brought from one locality, I have not yet had the opportunity of ascertaining whether it ever presents in an early stage any approximation to the helical mode of growth ; but such a deficiency of affirmative evidence is obviously not equivalent to a disproof of what has strong analogy in its favour. The variations which I have described among the different forms of Oper- culina, although limited to the form of the spire and the character of the surface- markings, would be amply sufficient to justify the erection of numerous species, were it not for the connexion established between the most divergent forms by intermediate links, and the necessity for an almost indefinite multiplication of hypothetical originals which the adoption of such a method would involve. The existence of such a large extent of variation among the specimens collected in the same locality must be admitted as valid evidence of the facility with which differential characters develope themselves in this type ; and a strong probability is thus afforded in favour of the varietal character of larger differences among individuals whose conditions of existence are very diverse. 576 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. Hence the analogy of Operculina affords good ground to surmise that many of the reputed species in the nearly-allied genus Nummulites have no real title to that rank ; the differences among many of them being not nearly so great as those we have met with among the varieties of Operculina ; whilst those presented by many others do not exceed what might be reasonably expected to occur under a greater variety of modifying agencies. But I have shown (^[ 162) that it may be fairly questioned whether there is adequate ground for upholding the generic distinctness of Operculina and Nummulites ; the characteristic by which the latter has been asserted to be specially distinguished being not unfrequently observable as a varietal difference in the former. The form which I have described under the designation of Amphistegina Cuminyii* bears a striking resemblance to the ordinary Nummuline type in the early part of its growth, and to the ordinary Operculine in the later ; and may be regarded as in many respects a connecting link between the two. 247. There appears, then, strong reason for considering Cycloclypeus, Heterostegina, Operculina, Nummulites, and Ampliistegina as related to each other in the same manner and degree as the leading forms already enumerated under the Orbiculine group. And it is very curious to observe the perfect analogy which prevails in regard to the forms under which these two great types of structure essentially different as they are tend to develope themselves. As I have just pointed out, the relation of Cycloclypeus to Heterostegina is exactly that of Orbitolitcs to OrMculina. So if the transverse or secondary septa of Heterostegina were undeveloped, we should have an Operculina, Nummulina, or Amphistegina (these three types being in my view essentially one and the same), just as the like deficiency actually occurring in Orliculina gives to it all the essential characters of Peneroplis. And the parallelism seems to be completed by the existence in Fusulina^ of the same elongated condition of this type, that Alveolina is of the Orbiculine. 248. The mutual accordance of all these in the highly elaborated texture of the shell, in the relation which this bears to the segments of the sarcode-body, and in the presence of an intermediate skeleton with its canal-system, is extremely close. The substance of the shell is very dense, and of almost vitreous transparence where it is not perforated by the minute closely-set tubuli which usually pass direct from the interior of the chambers towards the external surface. Each segment of the body has its own proper envelope, so that the septa between the chambers are composed of two distinct laminae, which diverge from each other where they give passage to the canal-system, and which are often further separated by the intervention of a portion of the " intermediate skeleton." The passages of communication between the chambers are so narrow, that the segments * It is questioned by Messrs. PAEKER and RUPEET JONES whether this is a true Ampliistegina, chiefly on account of its bilateral symmetry (Ann. of Nat. Hist. Feb. 1860, p. 111). But 1 have met with perfect bilateral symmetry in specimens warranted as Amphistegince by those excellent judges of that type. t I have not yet been able to satisfy myself as to the precise affinities of Fusulina, the metamorphic con- dition of its shell interfering with the minute study of its structure ; but my view of its nature essentially corresponds with that of Messrs. PAEKEE and RUFEET JONES. (Sec Quart. Journ. of Geol. Soc , Nov. 1860, p. 458.) CONCLUDING- SUMMAEY: EXTENT OP EANGE OF VAEIATION. 577 of the body are much more isolated from each other than they are in the type already described ; and the proper walls of the chambers seem, as it were, to be moulded upon the segments, instead of merely filling up the interspaces between them. This filling-up, in fact, is the office of the " intermediate skeleton," which gives a solidity to the whole aggregation that would otherwise be wanting ; and special provision, as we have seen, is made in the canal-system for its nutrition. Altogether this type is the one in which the Foraminiferous structure attains its highest development, and which is most com- pletely differentiated from every other. And the morphological variations it is known to undergo seem to me fully to justify the inference, that such further variations as have been shown to occur in the Orbiculine type might be regarded as the probabl source of the divergence, from some common ancestral stock, of the several forms whose intimate relationship I have demonstrated. The analogy of that type would suggest Heterostegina as presenting the nearest existing approximation to such a common original ; and the stages of differentiation may be thus expressed : Heterostegine type diverging into Operculina Heterostegina A. ~ Amphistegina, Nummulites, Operculina. Heterostegina, Cycloclypeus. From my imperfect acquaintance with Fusulina, I do not feel justified in expressing its exact relationship to either of the forms included in this scheme ; and for the same reason I abstain from connecting OrMtoides with Cycloclypeus, to which it has some features of close relationship. 249. After this detailed examination of the general relations of the principal modifi- cations of two of the most strongly marked types to be found in the whole group of Foraminifera, it seems needless for me to do more with respect to the other forms whose structure I have investigated, than to inquire how far the peculiar characters by which they are respectively distinguished show evidence of a like variability. Thus we have seen that Calcarina is essentially distinguished from Rotalia by the extraordinary development of the " supplemental skeleton," and by the extension of this into radiating prolongations. But it has been shown (^[ 197) that the number, form, and proportions of these prolongations are subject to very considerable variations ; so that whilst they are sometimes so greatly multiplied and prolonged as to constitute the principal feature of the organism, they are so little developed in other instances that the contour of the disk is scarcely interrupted by them. Further, it has been shown (^[207) that the develop- ment of this supplemental skeleton is in great degree independent of that of the spire ; hence if this last be the essential component of the organism (as all analogy indicates), the supplemental skeleton must be regarded as a feature of minor importance. On the other hand, the development of radiating outgrowths is an occurrence not unfrequent among other helicine Foraminifera, even in species whose typical form is altogether MDCCCLX. 4 G 578 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINTFEEA. destitute of them (as Professor WILLIAMSON has pointed out in Potystomella crispa); and such forms differ much less widely, as regards this character, from the simpler forms of Calcarina, than these last do from the very complex forms which have been shown (^[ 197) to be connected with them by a continuously gradational series. Hence I cannot regard the remarkable development of the supplemental skeleton in Calcarina as affording any disproof of the idea of its genetic relationship to Rotalia, to which its affinity in every other particular is most intimate. 250. If, again, we inquire into the import of that remarkable development of the canal-system, which seems to be the distinctive feature of Polystomella, we find that if we base our judgment upon a sufficiently wide foundation of facts, its non-essential character becomes apparent. For although the large P. craticulata of the tropical and Australian seas presents the most symmetrical and extensive distribution of the canal- system that I have anywhere met with (^[ 187-191), the little P. crispa of our own seas exhibits but feeble traces of it (^f 192) ; yet of the intimacy of their relationship no doubt can be fairly entertained. We have traced a parallel difference between the gigantic Amphistegina Cumingii and the comparatively diminutive A. gibbosa (*jf 171). And the like difference has been shown to exist between the two forms of Tinoporus (^[ 222), where its presence or absence is obviously associated with the presence or absence of the radiating prolongations and of the supplemental skeleton from which these proceed. 251. In considering the import of the canal-system as a character for the systematist, the mode of its formation must not be left out of view. I have shown that the passages which altogether go to make up this system are not true vessels, but are mere sinuses, left in some cases by the incomplete adhesion of the two contiguous walls which sepa- rate the adjacent chambers, and in other cases apparently originating in the incomplete calcification of the sarcode which forms the basis of the solid skeleton ; certain portions of that substance remaining in its original condition, so as to maintain a communication between the contents of the chambers and the parts of the calcareous skeleton most removed from them, analogous to that which the Haversian canals afford in the case of laminae of bone not in the immediate vicinity of a vascular surface. As, therefore, the development of the Haversian system is related to the thickness of the bone-substance to be nourished, so does that of the canal-system in Foraminifera seem to be related to that of the consolidating substance which constitutes the supplemental skeleton. And it is to be specially observed that nearly all the forms in which (so far as we know at present) it attains any considerable development, are denizens either of tropical or of sub- tropical regions, in which the influence of external conditions appears specially to favour the largest growth and the most specialized evolution of the Foraminiferous type. 252. I think it better, in the present limited state of our knowledge of two of the types to the elucidation of whose structure the present memoir has been devoted, viz. Tinoporus and Carpenteria, to forbear to speculate further than I have already done upon their relationship to the forms already familiar to systematists (^[ 217, 230). CONCLUDING STJMMAET : EXTENT OF EANGE OF VAEIATION. 579 And I have now to show that the results of my inquiries in regard to such typical forms of Foraminifera as have passed under my review, are in complete accordance with those obtained by other naturalists who have proceeded upon the like method, that of the comparison, not of selected specimens, but of entire gatherings over extended areas and through various geological epochs. 253. Long ago did those excellent observers MM. FICHTEL and MOLL (1803) manifest a clear perception of the wide range of variation to which certain types of Foraminifera are liable, and give admirable descriptions and figures of such varieties. But their views were too philosophical for the species-making systematists of their time : thus we find MONTAGU remarking of the forms which they ranked under the designation of Nau- tilus (now Cristellaria) calcar, " If these can be admitted as the same species, we may bid adieu to specific definition;" whilst out of the same series DENTS DE MONTFORT con- structed no fewer than nine genera. It has been most unfortunate for the advance of this inquiry, that M. D'ORBIGNY should have prosecuted it under the influence of those ideas in regard to the differentiation of specific types, which he brought to the study of Foraminifera from that of the testaceous Mollusks to which he originally regarded them as allied; and that the influence of his comprehensive labours and high reputation should have given to his views, alike on the detailed arrangement and on the general classification of this group, a currency to which their entire inconsistency with its natural affinities entirely negatives their claim. It was in studying one of the simplest types of this series, that Professor W. C. WILLIAMSON was led (1847) to perceive " the extraordinary capacity for variation which Lagence exhibit in different states and ages ; extreme forms which appear to be very distinct from one another being connected together by specimens of an intermediate aspect to an extent only to be believed by those who examine a large series of specimens side by side*." And he adduced strong evidence, not only that most if not all of its reputed species are only varieties, but that the difference which separates the genus Lagena from Entosolenia, viz. that the neck of the flask-shaped body is prolonged inwards in the latter, instead of outwards as in the former, may be merely varietal. A like conclusion, strengthened by the results of his and my researches in regard to other types, Professor WILLIAMSON has extended to Fora- minifera generally, in his beautiful Monograph of the Recent Foraminifera of Great Britain (1857). "It may now be regarded," he says, "as an established truth, that most of the external characters on which both earlier and later writers relied for distin- guishing their species, possess but little value. The direction of growth in these shells, and the sculpturing of their exteriors, are alike influenced by age and local circum- stances; hence a dissimilarity between the different stages in the development of the same individual, such as finds few parallels amongst the Mollusca with which Concho- logists have so long identified them." Again, he remarks, " Nothing is easier than to throw the Foraminifera obtained by dredging over some limited area into defined groups, each of which has apparently a specific value. But as we extend our researches to more distant localities, new and intermediate forms perplex our minds as to what are the same * Annals of Natural History, Jan. 1848, p. 10. 4o2 580 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. and what different species. Long before our dredging-net has swept round the British coasts, we find that what was already difficult trenches on the impossible ; and when we test our results by applying them to collections made in remote parts of the globe, we become convinced that the limited amount of our information makes that impossibility absolute. The more extensive our experience, the weaker become our convictions respecting the limits of variation in any species*." 254. The relations of the forms belonging to the family Miliolitidce have recently been investigated by Mr. W. K. PARKER f on the same method of extensive and minute comparison; and his results are not only in perfect harmony with those obtained by Professor WILLIAMSON and myself, but even go beyond them in generality. Thus in each of the genera Comuspira, ffauerina, and Vertebralina, Mr. PARKER reduces all the reputed species to one, while he shows that even their generic differences are really but of small account. And he not only in like manner reduces all the reputed species of the genus Miliola to the level of varieties, but brings down to the same rank the reputed genera Spiroloculina, Biloculina, Triloculina, and Quinqueloculina ; the differences be- tween which, arising from asymmetrical growth, and from variations in the form and number of the chambers, cannot be regarded as even of specific value, the Milioline plan of construction being preserved throughout. " If," he remarks, " the forms kept themselves as distinct as those represented in the diagrams, a naturalist might be excused for regarding them as distinct types ; but between any two of these there may readily be found innumerable gradations, in large and small specimens, in the smooth and ornamented, in the shelly or the sanded, in attenuated and in distended individuals, and in specimens with symmetrical or non-symmetrical, or with two- or three-sided shells." I may add that I am fully prepared to endorse these conclusions; since they are entirely borne out by my own experience as to such forms of the Milioline type as have fallen under my notice. 255. In the large group of Nodosarince which has been carefully studied by Messrs. T. EUPERT JONES and W. K. PARKER, those gentlemen have felt themselves justified on the like grounds in reducing a multitude of reputed genera and species to a single type. Between the nautiloid Cristellarm and the straight moniliform or rod-like Nodo- sarice, which agree in essential characters of structure and mode of growth, they find such a continuous series of connecting links, that no line of demarcation can be any- where drawn, the straight, the curved, and the spiral forms passing gradationally one towards another. And the extreme forms being thus brought together, the various intermediate grades which have been distinguished by systematists under the generic names of Glandulina, Lingulina, Dentalina, Rimulina, Vaginulina t Planularia, Margi- nulina, Dimorphina, Flabellina, and Frondicularia, necessarily fall into the same category J. * Introd. pp. k, x. t Transactions of the Microscopical Society, 1858 (New Scries, vol. vi.), p. 53. J Annals of Natural History, Nov. 1859, p. 477 ; and Quarterly Journal of the Geological Society, Aug. 1860, p. 302, and Nov. 1860, p. 454. CONCLUDING STJMMAEY : DEEIVATION OF EXISTING FEOM ANCIENT FOEMS. 581 256. The same general doctrine having thus been shown to hold good in regard to all the chief natural subdivisions of the Foraminiferous group, it is not my purpose at present to prolong the inquiry in this direction. The systematic study of this tribe needs to be prosecuted far more extensively than my own time and opportunities have admitted, to enable even an outline-scheme to be framed, which should represent an approximation to the true relations of its principal families. But I think I have made it clear that such a scheme will be most likely to approach the truth, when the basis of it is laid in a thorough knowledge of the nature and extent of those variations which every chief modification of this type shows itself so peculiarly disposed to exhibit, and when, in building it up, the idea of natural affinity is accepted as expressing not only degree of mutual conformity, but actual relationship arising from community of descent more or less remote. For the occurrence of endless gradational departures from any types which we may assume as fixed, and of links of connexion between such as present the best-marked differentiations, seem to me to point unmistakeably to this as the only means of escape from that difficulty of indefinite multiplication which attends the doc- trine of distinct specific creations when applied to a group in which scarcely any two individuals are alike. The case, in fact, is very analogous to that of the relationship between the various members of the family of Mankind; for whilst the historical evidence of actual change in them is so incomplete as well as so limited in its range, as to be quite inadequate of itself to establish their community of descent, yet when that evidence is considered in its relations to analogous facts drawn from the far greater variations of domesticated animals, and to the manifold gradations by which the extreme types are connected, physiologists of the highest eminence have felt themselves justified in accepting that community as probable. Now the modifications which any single type of Foraminifera must have undergone, to give origin to the whole series of diversified forms presented by that group, are not greater in comparison with those of which we have direct evidence, than are those which the advocate for the Specific Unity of the Human Races has no hesitation in assuming as the probable account of their present divergence. 257. This view of the case derives great force from the fact, that there is strong reason to regard a large proportion of the existing Foraminifera as the direct lineal descend- ants of those of very ancient geological periods ; a doctrine first advanced by Professor EHRENBERG in regard to a considerable number of Cretaceous forms ; since fully con- firmed and extended as regards the Tertiary fauna by the admirable researches of Messrs. RUPERT JONES and PARKER on the Rhizopodal Fauna of the Mediterranean, as well as by my own comparison of the recent and fossil types of Orlitolites, Orliculina, Alveolina, Operculina, and Calcarina; and shown to be applicable also to the Secondary fauna, as far back as the upper part of the Triassic system, by the remarkable results of the investigations of the same gentlemen in regard to a well-preserved sample of it. Following out, by laborious and extended comparison, the method of inquiry I have so much insisted on, they have found ample evidence that the fact of a wide range of varia- 582 DE. CARPENTER'S RESEARCHES ON THE PORAMINIFERA. tion in this group is not confined to the present epoch, but that it is true also of the Foraminiferous fauna of all the geological periods to which their researches have extended. " Our own experience of the wide limits within which any specific group of the Foraminifera multiply their varietal forms, related by some peculiar conditions of growth and ornamentation, has led us to concur fully with those who regard nearly every species of Foraminifera as capable of adapting itself, with endless modifications of form and structure, to very different habitats in brackish and in salt water, in the several zones of shallow and abyssal seas, and under every climate, from the Poles to the Equator. In arranging our synoptical tables of the Mediterranean Rhizopoda, recent and fossil, and in comparing their numerous specific and varietal forms one with another, we have not confined ourselves to our collections from this region, but have necessarily made comparisons of forms from almost every part of the globe, from the Arctic and the Tropic Seas, from the temperate zones of both hemispheres, and from shallow as well as deep-sea beds. Geologically, also, we have reviewed the Foraminifera in their manifold aspects, as presented by the ancient Faunas of the Tertiary, Cretaceous, Oolitic, Liassic, Triassic, Permian, and Carboniferous times ; finding, to our astonishment, that scarcely any of the species of the Foraminifera met with in the Secondary rocks have become extinct ; all, indeed, that we have yet seen have their counterparts in the recent Mediterranean deposits. This is still more clearly found to be the case with regard to the Chalk of Maestricht and the Tertiaries*." 258. The same excellent observers, in summing up their description of the Forami- nifera of the blue clay met with in the alabaster pits at Chellaston near Derby, belonging to the Upper Triassic series, thus express themselves : " Having thus pointed out that, judging from these specimens obtained at Chellaston, the minute Nodosarince and other Foraminifera of the Triassic period have continued to exist through the intermediate ages to the present day without losing any of their essentially specific features, we will observe that the Nodosarias are present in rocks of still greater age than the Trias, namely, the Permian and Carboniferous, and probably even the Lower Silurian. Nodosarice and Dentalince abound in some of the Permian limestones of Durham and the Wetterau in company with Textularice. Nodosaria occurs also in the Carboniferous Limestone of Ireland, according to M'Coy ; and the green sand of the Lower Silurian series near St. Petersburg has granted to EHRENBEBG casts of chambers something like those of Dentalina, together with unmistakeable casts of Textularian and Rotalian shells. We may remark, too, that the Fusulina of the Russian, North American, and Arctic Mountain limestone carries back the pedigree of the Nonionina group to the paleozoic periods, and that it is accompanied with other Foraminifera of known types, among which Nummulina is not absent. This last-named type has rare representatives in the Lias and Oolite ; it acquired great potency in the Tertiary seas, and is not extinct now. Altogether we have here some remarkable instances of the persistency of life- types among the lower animals. Though the specific relations of the Paleozoic Forami- * Quarterly Journal of the Geological Society, August I860, p. 294. CONCLUDING SUMMARY: EXTENT OF BANGE OF VARIATION. 583 nifera require further elucidation, we feel certain that the six genera represented in the Upper Triassic clay of Chellaston by about thirty varieties, stand really in the place of ancestral representatives of certain existing Foraminifera, that they put on their several subspecific features in accordance with the conditions of their place of growth, just as their posterity now do, and that although we have in this instance met with only the minute forms of a 700-fathoms mud-bottom, yet elsewhere the contemporaneous fuller development of these specific types may be found by careful search in other and shallower-water deposits of the Trias period*." 259. It can scarcely, I think, be questioned that such a continuity of the leading types of Foraminifera maintained through so long a series of geological periods, and the recurrence of similar varietal departures from those types, are results of the facility with which creatures of such low and indefinite organization adapt themselves to a great variety of external conditions; so that, on the one hand, they pass unharmed through changes in those conditions which are fatal to beings of higher structure and more specialized constitution; whilst, on the other, they undergo such modifications under the influence of those changes, as may produce a very wide departure from the original type. Thus we have found strong reason for regarding temperature as exerting a most important influence in favouring not merely increase of size but specialization of development : all the most complicated and specialized forms at present known being denizens either of tropical or of sub-tropical seas ; and many of these being represented in the seas of colder regions by comparatively insignificant examples, which there seems adequate reason for regarding as of the same specific types with the tropical forms, even though deficient in some of their apparently most important features. The depth of the sea-bottom seems also to affect the prevalence of particular types, and to modify the forms under which these present themselves ; so that Messrs. RUPERT JONES and PARKER feel themselves able to pronounce approximatively as to the depth of water at which a deposit of fossil Foraminifera may have been formed, by a comparison of its specific and varietal types with those characterizing various depths at the present time. And it is specially worthy of note, that in the greatest depths of the ocean from which Foraminifera have been brought by deep-sea soundings, these belong almost exclusively to one type, Globigerina. 260. In applying the results of the foregoing inquiry to the Animal Kingdom generally, it may be at once conceded that no other group affords anything like the same evidence, on the one hand of the derivation of a multitude of distinguishable forms from a few primitive types, and on the other of the continuity of those types through a vast succession of geological epochs. But a nearly parallel case, as regards the first of these points, is presented by certain of the humbler groups of the Vegetable Kingdom ; in which it is becoming more and more apparent, from the careful study of their life-history, not only that their range of variation is extremely wide, but that a large number of reputed genera and species have been erected on no better foundation * Quarterly Journal of the Geological Society, November 1860, p. 458. 584 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFERA. than that afforded by the transitory phases of types hitherto known only in their states of more advanced development*. It would be very unreasonable to put aside these cases as so far exceptional that no inferences founded upon them can have any applica- tion to the higher forms of Animal and Vegetable life. For it is only in the degree of their range of variation, that Foraminifera and ProtopJiyta differ from Vertcbrata and Phanerogarnia ; and the main principle which must be taken as the basis of the system- atic arrangement of the former groups, that of ascertaining the range of variation by an extensive comparison of individual forms, is one which finds its application in every department of Natural History, and is now recognized and acted on by all the most eminent Botanists and Zoologists. It will be sufficient for me here to refer to the views recently advanced by Dr. J. D. HOOKER in his " Introduction to the Flora of Australia ; " the results of his extensive experience in the comparison of the Floras of different portions of the globe having led him to conclusions regarding the probable origin of the diversities they present, with which my own deductions from the study of the Foraminifera are in complete accordance. And I am authorized by Mr. THOMAS DAVIDSON, whose extensive knowledge of the Srachiopoda enables him to speak as the highest authority upon all that relates to that most interesting group (which, like the Foraminifera, is traceable through the entire series of fossiliferous rocks), to state that in proportion to the increase of his knowledge of its modifications of type, does he find reason to regard many of them as having had so wide a range of variation, as fully to justify him in making a large reduction in the number of specific types hitherto accounted distinct ; whilst in the same proportion he finds himself able to trace with considerable probability the same specific types through a succession of geological periods, certain Oolitic and Cretaceous Terebratulidce, for example, being the probable ancestors of existing forms ; and even the Lingula of the Wenlock Silurians not being distinguish- able by any characters which he can re.cognize as constituting a valid specific difference from the Lingula anatina of our present seas. 261. The following are the general propositions which it appears to me justifiable to base on the researches of which I have now given a resume : I. The range of variation is so great among Foraminifera, as to include not merely the differential characters which systematists proceeding upon the ordinary methods have accounted specific, but also those upon which the greater part of the genera of this group have been founded, and even in some instances those of its orders. II. The ordinary notion of species as assemblages of individuals marked out from each other by definite characters that have been genetically transmitted from original proto- types similarly distinguished, is quite inapplicable to this group ; since even if the limits * It is among the lower Fungi that the researches of TULASNE and others have shown the greatest variability to prevail ; whilst the recent inquiries of Dr. J. BBAXTON HICKB have hrought to light a most unexpected relationship between the supposed Unicellular Algae and the Gonidia of Lichens. See his Memoirs in the Quarterly Journal of Microscopical Science, October 1860 and January 1861. GENEEAL CONCLUSIONS. 585 of such assemblages were extended so as to include what would elsewhere be accounted genera, they would still be found so intimately connected by gradational links, that definite lines of demarcation could not be drawn between them. III. The only natural classification of the vast aggregate of diversified forms which this group contains, will be one which ranges them according to their direction and degree of divergence from a small number of principal family types; and any subordinate groupings of genera and species which may be adopted for the convenience of description and nomenclature, must be regarded merely as assemblages of forms characterized by the nature and degree of the modifications of the original type, which they may have respectively acquired in the course of genetic descent from a common ancestry. IV. Even in regard to these family types, it may fairly be questioned whether analo- gical evidence does not rather favour the idea of their derivation from a common original, than that of their primitive distinctness. V. The evidence in regard to the genetic continuity of the Foraminifera of successive geological periods, and of those of the later of these periods and the existing inhabitants of our seas, is as complete as the nature of the case admits. VI. There is no evidence of any fundamental modification or advance in the Forami- niferous type from the Palaeozoic period to the present time. The most marked trans- ition appears to have taken place between the Cretaceous period, whose Foraminiferous fauna seems to have been chiefly composed of the smaller and simpler types, and the commencement of the Tertiary series, of which one of the earliest members was the Nummulitic Limestone, which forms a stratum of enormous thickness that ranges over wide areas in Europe, Asia, and America, and is chiefly composed of the largest and most specialized forms of the entire group. But these were not unrepresented in pre- vious epochs ; and their extraordinary development may have been simply due to the prevalence of conditions that specially favoured it. The Foraminiferous fauna of our own seas probably presents a greater range of variety than existed at any preceding period ; but there is no indication of any tendency to elevation towards a higher type. VII. The general principles thus educed from the study of the Foraminifera, should be followed in the investigation of the systematic affinities of each of those great types of Animal and Vegetable form, which is marked out by its physiological distinctness from the rest. In every one of these there is ample evidence of variability ; and the limits of that variability have to be determined by a far more extended comparison than has been usually thought necessary, before the real relations of their different forms can be even approximatively determined. VIII. As it is the aim of the Physical Philosopher to determine " what are the fewest and simplest assumptions, which being granted, the whole existing order of nature would result*," so the aim of the philosophic Naturalist should be to determine how small a number of primitive types may be reasonably supposed to have given origin by the ordi- nary course of " descent with modification " to the vast multitude of diversified forms * MILL'S Logic, 3rd edition, vol. i. p. 327. HDCCCLX. 4 II 586 DE. CAEPENTEE'S EESEAECHES ON THE FOEAMINIFEEA. that have peopled the globe in the long succession of geological ages, and constitute its present Fauna and Flora. [Note. I cannot bring to a conclusion this series of Researches, without on the one hand gratefully acknowledging the liberality of the Council of the Royal Society, whose assistance (from the Grant placed at their disposal by Government) has greatly aided my investigations, by enabling me to have every important feature of form and struc- ture accurately delineated under a sufficient magnifying power; and, on the other, expressing my great obligation to the skill, intelligence, and patient assiduity of my draughtsman, Mr. GEORGE WEST, by whose careful study of these organisms under my direction I have been enabled to attain a much more thorough knowledge of their nature than my own more limited time would have permitted me to acquire.] EXPLANATION OF THE PLATES. PLATE XVII. "With the exception of figs. 9 and 10, all the figures in this Plate refer to Polystomella craticulata. Fig. 1. Polystomella craticulata, as viewed at a in its peripheral aspect, showing the septal ridges and the intermediate rows of ' fossettes,' and, on the left of the figure, the septal plane with the row of minute apertures along its inner mar- gin ; at b is shown one of the lateral surfaces of the same shell, over the central part of which the septal ridges are concealed by an exogenous deposit that is irregularly dotted with punctations resembling the ' fossettes.' Mag- nified 10 diameters. Fig. 2. Vertical section passing nearly through the umbilical axis, and laying open the outer seven whorls, but passing by those of earlier formation, so as to bring into view at e the external surface of the eighth (counting inwards). This figure shows the gradual increase of the distance between the two lateral surfaces of the successive whorls, of which those last formed do not invest the earlier, so that the shell would be biconcave but for the large amount of solid exogenous deposit, which not only occupies the umbilical region, but extends even to the last whorl ; it also shows the continuity of this deposit with that outer portion of the spiral lamina of each whorl which presents in section a plicated aspect; and along the inner margin of each of the septa displayed in the section, ia seen a row of minute septal pores, forming the only direct communication between the chambers. The relations of the various parts of the canal-system are well displayed in this section : Byv<'> jfel / Jw/$*y^-22.Mll!l]fy /' ///' ' I / f ' ' / ' ' V "" f ^v wlr^' ~^^JA.'^\ \\\ \ : ^^:ff\ (. *.v/ ,/,./' LrtsMilous. titk ':.. ,.,.'!!> ' PhilJrvis.MDCCCLZ Plate XXI. ';.:.>*, %.. J '?" .. -;*' '" l \ ">'-4. ' " :? "'V / "^"; 10^ ' V '",-;;V.. i; 'v" '''- ?' ; V- : 'f : ; V;' ''> "..S ' ! -' : '"" . .-//, Phtl Trans MDCCCLX PfaeXXll. , i v-V"'' ; -T* , u i""* ' * ..'.v * - - L 547 ] XXVI. Researches on the Foraminifera. By WILLIAM B. CARPENTER, M.D., F.R.S., F.G.S. $c. Received June 19, Read June 19, 1856. PART II. ON THE GENERA ORBICULINA, ALVEOLINA, CYCLOCLYPEUS AND HETEROSTEGINA. Genus ORBICULINA. 78. History. X HE interesting group of organisms belonging to this type seems to have early attracted notice, probably on account of the great abundance in which it presents itself on the sands of many of the West Indian shores. Three species are described and figured by FICHTEL and MOLL*, under the names of Nautilus orbiculus, N. angulatus, and N. aduncus. LAMARCK, however, in his first systematic treatise-^-, separated them from Nautilus, and raised them to the rank of an independent genus, to which he gave the name of Orbiculina ; and he also changed two of the specific names, the three standing respectively as O. numismalis, O.angulata, and O.uncinata. By DENYS DE MONTFORT^:, these species were raised to the rank of independent genera, under the names of Helenis, Archaias, and Jlotes ; but these genera have not been adopted by any other systematise M. D'ORBIGNY^, in his first classification of the Foraminifera, not merely adopted LAMARCK'S generic designation, but affirmed that the three reputed species were really nothing else than one and the same organism in different phases of growth, O. angulata being the youngest, O. numismalis the next in age, and O. uncinata the .adult. He arrived at this result, of the truth of which I am myself well assured, by the comparison of a great number of specimens, a process which it would have been well for science if he had more constantly adopted. The name of the adult form should of course stand as that of the species ; but the organism in question is more commonly known under the designation Orbi- culina adunca, which seems to have been conferred upon it by M. DESLONGSCHAMPS||. 79. A considerable number of figures of this species are given by M. D'ORBIGNY in his treatise on the Foraminifera of Cuba, which forms part of the great work of M. RAMON DE LA SAGRA on the Natural History, &c. of that island^. These figures, * Testacea Microscopica, Vindob. 1798. f Systeme des Animaux sans Vertebres. Paris, 1801. J Conchyliologie Systematique. Paris, 1808. Tableau Methodique de la Classe des Cephalopodes. Paris, 1825. || Encyclopedic Methodique, Zoophytes. Paris, 1824. ^f Histoire Physique, Politique, et Naturelle de 1'Ile de Cuba. Paris, 1840. MDCCCLVI. 4 C 548 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. however, serve only to give a general idea of the diversities of external conformation which had presented themselves to him ; and notwithstanding their number and variety, they do not include some of the most important among the protean shapes of these bodies, nor do they throw any light upon their internal structure. 80. The memoir of Professor EHRENBERG*, in which his group of Bryozoa was originally constituted, contains the first recognition of the relationship between Orbitolites (^[ 4) and Orbiculina ; of which the first had previously been ranked among the Zoophytes; while the second (until the Rhizopodous nature of the whole group of Foraminifera was made known by M. DUJARDIN in 1 835) had been associated with the Cephalopods. Professor EHRENBERG'S description and figures of Orbiculina, however, being just as inaccurate as I have shown those of Orbitolites to be (partly, it seems likely, through his unacquaintance with the mode of making thin sections), there is no occasion for me to make further reference to them. 81. The excellent memoir of Professor WILLIAMSON-|~ " On the minute structure of the Calcareous Shells of some recent species of Foraminifera," contains the first approach to a correct description of the internal conformation of Orbiculina; and the fact was fully recognized by him, that in advanced age, when the spiral type of growth has given place to the cyclical, there is no other difference between the structure of Orbitolites and that of Orbiculina, than that, which arises from the dis- similarity of their earlier mode of development. Although I am satisfied that, as to one or two points of minute structure, Professor WILLIAMSON has fallen into error, I am disposed to attribute this to the want of a sufficient number of well-preserved specimens for examination ; and for having it in my power to correct and extend his description, I am chiefly indebted to Mr. HUGH CUMING, the specimens of this type included in his Philippine collection being remarkable both for their high develop- ment, and for their very beautiful state of preservation. 82. The investigations of Professor WILLIAMSON are entirely unnoticed by M. D'ORBIGNY in his latest classification of the Foraminifera;}:; Orbiculina being ranked in the order Helicostegues, and being defined as follows, " Coquille nautiloide, com- primee, forme'e de loges divise"es inteVieurement en compartiments regnliers, percees de nornbreuses ouvertures en lignes longitudinales a 1'enroulement special ;" whilst Orbitolites (his definition of which has been already cited, ^| 5) is placed in the order Cyclostegues. The separation of the two genera by so wide an interval, is grounded, therefore, on the assumption that the type of growth in Orbiculina is spiral, whilst that of Orbitolites is cyclical. I have already shown that this assumption is incorrect as regards Orbitolites, the early plan of whose growth is frequently spiral (^[ 54); and I shall presently show that it is equally incorrect as regards Orbiculina, whose later plan of growth is typically cyclical. And after having closely compared these and * Transactions of the Royal Academy of Berlin, 1839. f Transactions of the Microscopical Society, 1st series, vol. iii. p. 120. J Cours Elementaire de Paleontologie, torn. ii. Paris, 1852. GENUS ORBICULINA: ORGANIZATION. 549 other types, we shall be in a position to inquire what value is really to be attached to such a character as a basis for classification. 83. Organization. Among- the diversified forms that are presented by the species before us, it is of course necessary to select some particular type as that with which others may be compared; and this I consider to be the form delineated in Plate XXVIII. fig. 5 ; since by far the larger proportion of the very numerous specimens I have examined show such a degree of approximation to it, that their differences may fairly be set-down to the account of incomplete development. The general aspect of a typical Orbiculina, then, differs but little from that of Orbitolites, except in the pro- minence of its nucleus, and in the peculiar spiral disposition of the bands by which the surface of the nucleus is marked. The species never attains, so far as I am aware, the dimensions of Orbitolites ; the diameter of the largest recent disk in rriy possession (one of Mr. CUMING'S Philippine specimens) being -20 of an inch; whilst of the specimens which I have seen from the West Indian and ^gean seas, none exceed -12 of an inch. Mr. CARTER*, however, describes under the name of Orbito- lites Malabaricus a fossil Orbiculina (^[ 90), whose specific identity with the type before us I see no reason to doubt, as attaining a diameter of from 7 to 8 lines. The thickness of the disk is usually less in proportion to its diameter, than it is in Orbi- tolites; but this is by no means a constant difference, depending as it does merely on the relative increase of the columns of sarcode in the vertical direction, as compared with the extension of the surface by the addition of new annular bands. 84. On more closely comparing the marginal portion of such a disk with the cor- responding portion of one of those forms of Orbitolites, in which the concentric annu- lation is strongly marked externally whilst the transverse divisions of each annulus are scarcely indicated (^[49), no essential difference is perceptible between them; and although these transverse divisions are usually but very faintly indicated in Orbiculina (Plate XXVIII. figs. 13, 14), yet they are sometimes very obvious, as is seen in figs. 15, 16. In the fossil Orbiculina just adverted to, the surface-markings (figs. 21, 22) are entirely conformable to the ordinary type of Orbitolites. The margin itself exhibits one or more ranges of pores (figs. 6, 7, 18, 19) arranged after precisely the same fashion as those of Orbitolites: these ranges are more numerous in the fossil than in the recent forms of this type. 85. The internal structure of Orbiculina presents such a general conformity to that of Orbitolites, that it will not be requisite here to do more than specify the points of agreement and of difference. The general aspect of the horizontal section of such a typical specimen as is shown in Plate XXVIII. fig. 1 1, does not differ from that of a corresponding section of Orbitolites in any other essential particular than the dispo- sition of the central portion, in which the successive additions to the first-formed chambers are so made as to produce a spiral, instead of a concentric disk. When we examine this central portion more closely, we see that its plan of growth is exactly the * Annals of Natural History, New Series, vol. xi. p. 425. 4 c 2 550 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. same as that which is presented by those aberrant forms of Orbltolites, in which the central cell buds-out only on one side, instead of all around (see ^[ 54, and Plate IX. fig. 4). The spiral mode of increase is usually carried-on much further in Orbiculina, than it ever is in Orbitolites ; several turns being made before it gives place to the cyclical plan. But as this is a mere question of degree, such a difference would scarcely alone afford a valid distinction between these two types. There is, however, this very definite positive distinction, that in Orbiculina each turn of the spire not only surrounds the preceding, but completely invests it above and beneath, every band of new chambers being continued to the very centre ; so that, whilst the spiral mode of growth continues, the thickness of the shell increases with each turn ; and after this has given place to the cyclical, the central nucleus, thus augmented in thickness, projects above the plane of the disk. This peculiarity in Orbiculina, which I have never found to be wanting, and to which there is never the least approach in Orbitolites, is best seen in vertical sections. Thus in Plate XXIX. fig. 3 is shown that part of a vertical section of a large disk, which has passed through its nucleus; the innermost and therefore first-formed portion of which is seen to be invested above and below by three layer.*, formed by three turns of the spiral. The same peculiarity is shown in Plate XXVIII. fig. 17, which represents a portion of a fossil disk near the nucleus. 86. The transition from the spiral to the cyclical mode of growth is effected in Orbiculina, exactly as in the aberrant forms of Orbitolites just referred-to, by the opening-out (so to speak) of the mouth of the spire; the successive rows extending themselves more and more on either side, until they meet around the previously- formed portion ; after which each new row forms a complete zone or annulus. The commencement of this change is seen in Plate XXVIII. fig. 1 ; and its subsequent stages are shown in figs. 2-5, a comparison of which will show that the specimens which they represent are in different phases of this transitional state. 87. But this transition by no means constantly occurs; for the original spiral plan of growth is not (infrequently maintained, apparently through the whole of life; so that specimens are often met-with, which are not inferior in size or in number of rows to the larger disks, but which retain the aduncal form. Such a series of speci- mens is shown in figs. 810; the first of which represents a very young Orbiculina, in that stage which is common to both types of growth ; whilst it is obvious from a com- parison of this with the two following, that their increase has continued to take place upon the same plan, each row that is put-forth from its predecessor terminating abruptly like it at its free extremity, without any such disposition to extend itself as would carry it round the nucleus so as to form a complete annulus. A horizontal section of such a specimen as is represented in fig. 10, is shown in fig. 12. It may, of course, be urged that such a difference ought to be accounted sufficient to sepa- rate the spiral and the discoidal types of Orbiculina, as two distinct species ; but the following reasons appear to me quite sufficient to negative such a mode of viewing GENUS ORBICULINA: ORGANIZATION. 551 them : First, they so closely resemble one another, as to be undistinguishable, in their early condition. Second, they correspond in every particular, so far as regards the structure of their minute parts. Third, the assumption of the cyclical plan of growth does not take-place at any one fixed epoch of development, but may occur at various periods. Fourth, the persistence of the original plan of growth throughout life, can- not be fairly regarded as anything else than an arrest of development, such as we shall presently see to be a common occurrence in Orbiculina, as in Orbitolites, in regard to other particulars (^[ 90). 88. Turning now from the general plan of growth to the minute structure of the individual parts of Orbiculina, we continue to find a very close conformity to the type of Orbitolites. The texture of the shell is precisely the same ; and it exhibits no other peculiarity than a minute punctation of the superficial layers (Plate XXVIII. fig. 13), which at first suggests the idea of apertures*, but which is found on careful examination of transparent sections (Plate XXIX.fig. 2) to be due to a mere thinning of the shell at certain points, so as to give an appearance of cellular areolation closely resembling what is seen in Orbitolites (Plate VI. fig. 5). 89. Not having had the good fortune to obtain specimens in which the animal body has been preserved, I cannot speak as confidently on the subject of its conforma- tion, as I could in regard to that of Orbitolites; but a comparison of the features which present themselves in the structure of the calcareous skeleton of the two types, can leave no reasonable doubt that the general arrangement of its segments of sarcode must have been precisely the same. For the conformation of the chambers and passages, as displayed by a horizontal section (Plate XXIX. fig. 1), shows that the soft body must consist of a succession of bands of sarcode, each band swelling at intervals into larger segments ; that the segments of each band usually alternate with those of the bands internal and external to it, so as to be opposite to the intervals between them ; and that the stolons which connect one band with another pass-forth from pores in the intervals between the segments of one, into the segments of the next, those of the outermost band emerging from the margin as pseudopodia. By the coalescence of these a new band would originate, which would become thickened into segments opposite the pores of the preceding, and would give off its own pseudo- podia from the intervals between the segments. 90. Again, by an examination of the natural margins of Orbiculinas (Plate XXVIII. figs. 6, 7, 18, 19), and by a comparison of these with vertical sections (Plate XXIX. fig. 3), it becomes evident that the same variety exists as in Orbitolites, in regard to the increase of the disk in thickness by the vertical elongation of the segments of sarcode. For some specimens are altogether conformable to the 'simple type' of Orbitolites, in having but a single floor of chambers (so to speak), with a single row of marginal pores; whilst others correspond with the complex type of Orbitolites, in * Professor WILLIAMSON (loc. cit.) has described these punctations as perforations for the passage of pseu- dopodia ; but I am quite certain that such is not the case. 552 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. having many such floorSj with numerous rows of marginal pores. In the former, the segments of sarcode, with their single annular stolon, would resemble rounded beads strung at short distances on a cord. In the latter the segments would be columnar, with constrictions at intervals, and would communicate with each other by two or more annular stolons. It is true that we seldom find any such complete differentia- tion of the superficial cells, as the fully-developed type of Orbitolites presents; but we have seen that such differentiation is by no means a constant character in that genus (^[58); and the structure of the most developed specimens I have examined among the recent Orbiculince closely corresponds in this respect with that of the fossil Orbitolites of the Paris basin, as will be seen on comparing the left-hand portion of Plate XXIX. fig. 3 with Plate VI. fig. 1 1 of my former memoir. But it is not a little remarkable, that while the fossil Orbitolites of the Paris tertiaries are less deve- loped in this respect than their existing representatives in the South Seas, the fossil Orbiculince of the Malabar tertiaries should be more developed than existing speci- mens of the same type; for in them we find the superficial cells differentiated from the intermediate layers (Plate XXVIII. fig. 20), precisely as in Orbitolites*. Looking to the far larger dimensions, as well as to the higher development, which these fossils present (^[83), as compared with the existing specimens of Orbiculina, I am disposed to believe that this type attained its highest evolution in a period long since passed, and that we now have, so to speak, only the degenerate descendants of an ancestry of higher rank; whilst in the case of Orbitolites, I am inclined to think that the type is most fully evolved at the present time. 91. General conclusion. From the foregoing details it is obvious that the relation- ship between Orbitolites and Orbiculina is extremely close ; the only essential point of difference between them being that which is furnished by the structure of the nucleus. Whether or not they ought to rank as types of distinct genera, or whether they ought (as Professor WILLIAMSON maintains) to rank as cognate species of the same genus, is a point as to which it is impossible to arrive at a satisfactory conclusion, until the characters which should serve for the distinction of genera and species in this class shall have been determined on a physiological basis. This much I think myself entitled to assert with confidence, that even if they are to be regarded as distinct genera, they must be ranked in the same family, and in immediate proximity to each other ; and that no classification can have any claim to be considered as natural, in which they shall be widely separated. Genus ALVEOLINA. 92. History. Although the form and aspect of the Foraminifera which are refer- able to this genus, would seem to remove them altogether from proximity to the pre- * Although Mr. CARTER has described this fossil as a species of Orbitolites, yet it is really an Orbiculina, as is shown by the spiral conformation of its central portion (Plate XXVIII. fig. 17), and by the investment afforded by each turn of the spire to its predecessor, as shown in fig. 18. GENUS ALVEOLINA: HISTORY AND ORGANIZATION. 553 ceding, yet the two really bear a very close relationship in all essential points of minute structure, as will appear from the particulars I shall presently have to detail. The following outline of what has been previously ascertained respecting it (for which I am indebted to M. D'ORBIGNY*), will show how little the nature of its organization has hitherto been understood. Most of the species at present known are fossils, occurring in association with Nummulites, Orbitolites, &c. in the Num- mulitic limestone, or in other formations which represent it; and the examples first described (by FORTIS) were confounded with Nummulites and Orbitolites under the term Discolites. By FICHTEL and MOLL, they were ranked as a sub-type of their comprehensive genus Nautilus. The designation Alveolites was first given to this type by Bosc-f-; but it was not generally adopted ; and MONTFORT, according to his wont, raised three of Bosc's species to the rank of genera, under the names of Borelia, Clausalia, and Meliolites. LAMARCK did not adopt either Bosc's or MONTFORT'S generic designations, but substituted a new one, Melonia ; and this was adopted by CUVIER and FERUSSAC. DEFRANCE proposed yet another name, Orizaria. And finally M. D'ORBIGNY, in 1825', adopted Bosc's name, with a slight alteration in its termination, which served at the same time to mark the continued existence of the type, and to distinguish it from a genus of Corals which also had received the name of Alveolites. The name Alveolina was soon afterwards adopted by M. DESHAYES ; and it may now be considered as the established designation of the genus. The following is the latest definition given of this type by M. D'ORBIGNY : "Coquille allongee dans le sens de 1'axe d'enroulement, formee de loges divise"es par des canaux capillaires, ronds, perches de nombreuses ouvertures placees en lignes transversales a 1'enroulement spiral." Not the least idea seems to me to be conveyed by this defini- tion of the real structure of these bodies, such as is brought into view by thin sections; and no one, so far as I am aware, has previously attempted thus to eluci- date it. 93. Organization. My investigations have been made upon specimens which were tolerably abundant both in Mr. JUKES'S Australian dredgings, and in Mr. CUMING'S Philippine Collection. These were obviously identical specifically, but the latter con- siderably exceeded the former in average size. The length of the longest complete specimen in my possession is '35 of an inch; but I have a specimen whose shape is somewhat abnormal indicating that it has increased with unusual rapidity in length, as proportioned to its diameter, which, though incomplete at one end, measures '50 of an inch. The ordinary form, from which any considerable departure is very rare, is that which is exhibited in Plate XXVIII. fig. 23 ; and it is obviously produced, as correctly stated by M. D'ORBIGNY, by the involution of a spiral around an elongated axis. The surface is marked-out by longitudinal furrows into a succession of bands of tolerably uniform breadth ; and each of these is crossed by secondary furrows, * Foram. Foss. de Vienne, p. 140. t Bulletin des Seances de la Societe Philomathique, No. 61. 554 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. which lie so closely together as to mark-out each band into a series of very elongated cells, which remind us of the oblong superficial cells of the complex type of Orbi- tolites, but are much narrower in proportion to their length. The mouth of the spire is closed by a solid wall, the surface of which is nearly flat, and which is perforated by three, four, or five rows of rounded pores, bearing a very close resemblance to those at the margin of Orbitolites and Orbiculina. 94. The resemblance suggested by external configuration is fully borne-out by the examination of the internal structure of this genus, as brought into view by longitu- dinal and transverse sections. For it is shown in Plate XXVIII. fig. 24, that the whole organism has originated from a single large nearly-globular cell, around which are successive layers of chambers freely communicating with each other, each layer completely enveloping the preceding, and adding much more to its length than to its diameter. A portion of such a section, more highly magnified, is shown in Plate XXIX. fig. 9. The formation of the successive layers by spiral involution, is well displayed by a series of transverse sections taken at different points, such as are delineated in Plate XXIX. figs. 4 to 7 ; the first (fig. 4) having crossed near the narrow termination of the fusiform shell, and having consequently traversed only the two last-formed turns; the last having crossed near the middle of its length, and having traversed all the six turns by which it was formed. Each whorl seems to be, as to all essential particulars, a repetition of the rest; and hence it will be sufficient to make a detailed examination of only a small part of a section, such as is shown on a larger scale in fig. 8. It is there observed that the spaces occupied by the Barcode-body and bounded by the shell, though far from regular, conform to a certain general plan. Each lamina is made-up (as the surface-furrowing and the marginal pores indicate) of a succession of bands of superimposed cells, of elongated form ; each band answer- ing to one of the annuli of QrbitoUtet. At each point corresponding with the external furrow, the superficial layer of shell sends a prolongation inwards (a, a, a), which thus marks-off the superficial cells more completely from each other than is the case with the subjacent cells; just below this projection is a rounded perforation (b, b, b), which marks the passage of a longitudinal channel ; and at a little distance beneath this is another (c, c, c), which is of very considerable dimensions. Some- times there are more than two such channels; but this is comparatively rare. These channels must have given passage to longitudinal bands of sarcode, running from one end of the lamina to the other, and freely connecting together all the segments occupying the piles of elongated cells of which it is composed. The marginal pores are the orifices of the chambers of the last-formed band ; and it will be observed in Plate XXVIII. fig. 23, that one row of them is contiguous with the preceding whorl of the spire, along which, as is shown in Plate XXIX. fig. 8, there is always a free passage. Thus we may consider the sarcode- body, as in the case of Orbitolite, to be composed of a mass of closely-connected segments, between which a calcareous skeleton grows-up according to a certain tolerably-regular type. It seems to me GENUS CYCLOCLYPEUS: ORGANIZATION. 555 probable from the conformation of the shell, that the segments formingsuccessive bands do not communicate directly with each other so much as with the great longitudinal stolons ; just as the successive annuli of superficial cells in Orbitolites communicate with each other, not immediately, but through the annular stolons (^[ 28). The pseudopodial prolongations issuing from the marginal orifices, probably first coalesce into a longitudinal cord of sarcode, when a new band of cells is to be formed ; and from this another series of segments is then budded-off. In the texture of the shell, in the relations of the different chambers, in their mode of communication with the exterior, in all (to speak concisely) which marks the physiological condition of this organism, its conformity to the types previously described is so close, that, notwith- standing the marked difference in their mode of increase (on which depends their form), they must rank, in any natural classification, in very close proximity with these*. Genus CYCLOCLYPEUS. 95. The organisms which 1 have now to describe, and to which I shall give the generic designation Cycloclypeus (suggested to me by Dr. J. E. GRAY), are amongst the most interesting of all the Foraminifera at present existing; on account both of the large dimensions which they sometimes attain, and also of the complexity of their structure. The only specimens of them yet known, were dredged by Sir EDWARD BELCHER from a considerable depth of water off the Coast of Borneo. Two of these, which are now in the British Museum-f-, are complete disks measuring no less than 2^ inches in diameter ; and by the kindness of Dr. J. E. GRAY, I have had the oppor- tunity of making microscopic sections of a fragment of a disk, which, when entire, must have nearly equalled these in size. Smaller disks of various dimensions pre- sented themselves in the same dredgings. 96. Organization. The external aspect of these disks is sufficiently like that of Orbitolites, to prevent the two genera from being readily distinguished by a super- ficial examination, especially when young specimens of Cycloclypeus are compared with Orbitolites of the complex type ; since, on the two snrfaces of the former (Plate XXX. fig. 1), there can be distinguished concentric rings of oblong cham- bers, which are not at all unlike the similarly-disposed superficial cells of the latter. The peculiarly-compact texture of the shell of Cycloclypeus, however, gives to its sur- face a smooth and glistening appearance, which is very different from that of Orbi- tolites. And further, the forms of the two disks ordinarily differ in this, that whilst the centre of Orbitulite is usually rather depressed than elevated, and the thickness * I have confined myself to an account of the existing species, as I have not had the opportunity of making a similar examination of any large number of fossil forms of this type. It is well, however, for me to mention, that the existing species seems to me to be certainly identical with the A. Boscii of the Paris tertiaries. f These are the disks referred-to by Professor WILLIAMSON in his Memoir on Orbitolites, &c., Trans, of Microsc. Soc. ser. 1. vol. Hi. p. 127. He appears to have considered them as gigantic Orbitolites, not being acquainted with their peculiarities of internal structure. MDCCCLVI. 4 D 556 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. of the disk generally increases towards the periphery, the central portion of Cydo- clypeus always presents a knobby elevation, on the surface of which the oblong boundaries of the chambers are superseded by rounded 'punctations,' whilst the thick- ness of its disk gradually diminishes towards its margin, where it is so reduced as to come to a sharp edge. In older specimens of Cydoclypeus, the boundary-markings of the chambers are scarcely distinguishable, save near the margin; their concentric annuli are marked-out, however, by rows of ' punctations,' similar in appearance to those of the central eminence. In either case, it is usually observable that the breadth of the annuli is far from constant; and that the annuli are not unfrequently incomplete, extending round only a portion of the disk. This irregularity has been noticed in Orlltolites (^[ 20) as of rare occurrence ; in Cydoclypeus it is so common that I have not yet met with specimens which are entirely free from it. 97. Whilst agreeing with Orbitolites in those external features which result from the cyclical mode of growth that is common to both forms, Cydoclypeus presents as wide a contrast to it in every other feature of its organization, as is anywhere known to exist within the limits of the Foraminiferous group. For whilst, as we have seen, the general plan of structure of Orbitolites removes it but little from the grade of Sponges the several segments of its aggregate body being but very imperfectly separated one from the other, and the shell which grows-up in the midst of them having no discoverable organization, that of Cydoclypeus closely approximates to the Nummulitic type, in which the successive segments are as completely isolated as they can be without entire disconnection, and in which, by the peculiarly-elaborate construction of the shelly covering, a special provision is made for their independent nutrition. 98. On making horizontal and vertical sections of the Cydodypeus-disk, its central plane is found to be occupied by chambers, disposed (ordinarily in a single layer) in concentric annuli ; these being covered-in above and beneath by compact plates of shell, which are thicker towards the centre, thinner towards the circumference (Plate XXX. fig. 1). The typical form of these chambers seems to be a parallelo- gram with its angles rounded off, whose sides are to each other as 1^ to 1, or as 2 or even 3 to 1, the longest side lying in the direction of the radius of the disk; but owing to the variation in the length of the chambers which results from the before- mentioned irregularity in the breadth of the annuli (^[ 96), the breadth of the cham- bers remaining more constant, their proportions vary greatly in different parts of the same annulus, or in adjacent parts of different annuli, as shown in Plate XXIX. fig. 12. I have occasionally met with chambers whose length was to their breadth as 4 to 1 (Plate XXXI. fig. 3). The vertical thickness or depth of the chambers, seems usually to be pretty constant in different parts of the disk, except near its centre; the thinning-away towards its margin being due, not so much to a diminution in the vertical height of the chambers, as to the reduction of the thickness of the shelly plates that enclose them above and below. GENUS CYCLOCLYPEUS: ORGANIZATION. 557 99. But although the existence of only a single layer of chambers is obviously the rule in this species, yet exceptions to it are not unfrequent ; a subdivision of the entire stratum into two or three presenting itself when its thickness is above the average, as is shown in Plate XXXI. fig. 8, a. Occasionally one or two chambers only are thus subdivided, as shown in figs. 4, 5. The cavity of each chamber is surrounded by a proper wall of its own, quite distinct from that of the chambers which it adjoins ; and hence the septum by which each chamber is divided from the adjacent one on either side, is formed of at least two lamellae (Plate XXIX. fig. 12). These come into close contact with each other at the junction of the vertical septum with the horizontal roof and floor of the chamber, as shown in Plate XXXI. figs. 2, 4, 5 ; but elsewhere they diverge from one another, leaving an interseptal space, which is partly filled-up by an interposed lamina of shell-substance, but is partly occupied by the interseptal canals to be presently described. A thicker space of the same kind is in like man- ner left between the proper walls of the chambers forming one annulus, and those of the chambers forming the annul! internal and external to it : this space is almost entirely filled-up by a shelly deposit, the interseptal canals which pass between the successive annuli being less numerous than those which run between the chambers of the same annulus (^[ 105). 100. As in Orbitolites, the chambers of each annulus usually alternate in position with those of the annuli internal and external to it. But this is by no means con- stantly the case ; since additional chambers are ' interpolated' here and there, so as to increase the number according to the augmented diameter of the annulus ; and such an interpolation disturbs the regular arrangement of the neighbouring chambers. 101. The adjacent chambers of the same annulus have not, so far as I have been able to ascertain, any direct communication with each other; an indirect communi- cation, however, is perhaps established through the system of interseptal canals. But each chamber normally communicates with two chambers of the annulus within it, and also with two of that which surrounds it, by large passages (shown in horizon- tal section in Plate XXIX. fig. 12, and in vertical section in Plate XXXI. fig. 2, c, c, and represented in perspective view in Plate XXX. fig. 4,ffandgg), which traverse the annular septa ; of these passages there seems to be normally but a single one for each pair of chambers thus to be brought into communication ; but I have frequently met with two, and occasionally three, one placed directly or obliquely above the other. Thus at each extremity of the oblong chamber, there are normally two pas- sages leading to two chambers of the annulus next internal or external to it ; but since to each of these chambers there may be two or even three passages, the total number at each end may be three, four, five, or six. The variations as to this point of structure which are presented in adjacent, chambers, are shown in Plate XXXI. fig. 2. But since, on the other hand, each of the 'interpolated' chambers communicates with only one chamber in the annulus next internal to it, there may be but a single passage, in place of two or more. 4 D 2 558 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. 102. The shelly plates which bound the chambered plane above and below, are formed of a succession of superimposed lamellae (Plate XXXI. fig. 10). These lamellae, which are of tolerably uniform thickness, are most numerous in the older or more cen- tral portions of the disk, and diminish in number towards the marginal or last-formed portions ; so that it seems pretty certain that new lamellae must be added from time to time, as the disk is augmented by the formation of new annuli. I have often met with appearances, which might seem to indicate that the formation of a new lamella over the entire surface of the disk, and the addition of a new annulus at its margin, were parts of one and the same act of growth, the new lamella being continued into the annular septum ; but if this were constantly the case, the number of lamellae which form the ceiling or floor of any chamber, would always correspond with the number of annuli external to it, which I do not find to hold-good. 103. Each of these lamellae is perforated by an assemblage of parallel tubuli very closely set-together, which pass from its inner towards its outer surface (Plate XXXI. figs. 9, 10) ; and there is such a continuity between the tubuli of successive lamellae, that a communication is thus established between the cavity of the thickest-walled chamber, and the external surface of the disk. These tubuli, however, are very mi- nute, their diameter being not above i ,ooo tn of an inch. They are wanting in cer- tain parts of the shell, which then presents a transparence that contrasts strikingly with the semi-opacity produced by the tubular perforations. By the comparison of vertical with horizontal sections taken in different planes, it appears that these trans- parent portions of the shell have a conical form, the base of each being on the sur- face of the shell, and its apex pointing to one of the angles at the outer margin of a chamber (Plate XXX. fig. 4, cc, dd). Their gradual widening towards the surface causes the diameter of their bases to increase with every addition to the thickness of the shell ; and thus it is on the older portion of the shell, and especially on its cen- tral protuberance, that they become most conspicuous as rounded ' punctations' ( 96). In horizontal sections of the superficial lamellaa, they form a large proportion of the area (Plate XXXI. fig. 6); whilst in similar sections near the chambered plane (fig. 9), they become blended with angular projections of the annular partitions, that fill-up the spaces left between the proper walls of the chambers by the rounding-off of their angles. 104. The lamellated structure is seen in these conical pillars (Plate XXXI. fig. 10, b), the lamellae being continuous with those of the tubular part of the shell; so that at each increase in thickness, a tubular and a non-tubular portion must be super- imposed upon the corresponding parts of the preceding lamella. Both in the tubular structure of the shell, and in the presence of these non-tubular columns, there is an exact conformity to the structure of Nummulite and its congeners*. * I avail myself of this opportunity of correcting a mistake into which I fell in my original description of the structure of Nummulite (Quart. Journ. of Geol. Soc., 1850, p. 26), in regarding the non-tubular columns of the shell as having been passages which had become filled-up by the infiltration of carbonate of lime GENUS CYCLOCLYPEUS: ORGANIZATION. 559 105. I have now to speak of another feature in the structure of this organism, which most strikingly differentiates it from Orbitolites and its congeners, and at the same time furnishes an additional proof of its close approximation to Nummulites, notwithstanding the difference in its plan of increase. I allude to the system of interseptal canals, which establish a direct communication between the external sur- face, and the parts of the interior most removed from it. Such radial canals are seen both in horizontal and vertical sections (Plate XXIX. fig. 10, Plate XXXI. fig. 4, c) excavated in that shelly substance, which occupies part of the space that intervenes in the radiating partitions between the proper walls of adjacent chambers of the same annulus. When the canal reaches the end of the radial septum, it usually sub- divides into two, which diverge at a considerable angle from each other, so as, by traversing the annular septum, to reach the two alternating radial partitions of the next annulus ; and as each branch, before entering the partition towards which it runs, unites with another branch that inclines towards it from the radial canal next adjacent, it follows that just as every chamber communicates (normally) with the two alternating chambers in the annuli internal and external to it, so do the inter- septal canals of every radiating partition communicate with those of the partitions alternating with it in the internal and external annuli. This arrangement, which cannot be described verbally without some complexity, will be readily comprehended by an inspection of Plate XXIX. fig. 11. In each radial partition there are at least two, and very commonly three tiers of such canals, as is best seen in vertical sections that cross the radial partitions transversely (Plate XXXI. figs. 4, 5). Short trans- verse branches, apparently communicating with the cavity of the chambers (Plate XXIX. fig. 11), are sometimes seen to proceed from the longitudinal canals ; in regard to these communications I would not speak with confidence from what I have seen in Cydoclijpeiis ; but that they exist in other organisms, hereafter to be described, is un- questionable. There can be no doubt, moreover, that the horizontal radiating canals communicate with vertical canals which pass directly towards the two surfaces of the disk, whereon they open (Plate XXXI. fig. 5, c) ; these canals are best seen in hori- zontal sections taken near the upper or under surfaces of the chambers (Plate XXXI. figs. 3, 9), in which they present themselves in regular rows, d, d, corresponding to the radial partitions ; whilst in similar sections taken nearer the surface, they are seen to be less regularly disposed, in consequence of their following a somewhat oblique in the process of fossilization. I was led to this by the very marked contrast which exists in Nummulite be- tween the tubular, and non-tubular portions of the shell, and the peculiarly inorganic semi- crystalline appear- ance of the latter, closely resembling that of the calcareous infiltration which usually occupies the interior of the chambers. Subsequent examination, however, of Nonionina and other recent forms most closely allied to Nummulite, has satisfied me that these columns were part of the original shell, as my friend Professor WILLIAM- SON maintained from the first. It is not a little curious, however, that in certain other species of Nummulite described by MM. D'ARCHIAC and HAIME, a system of passages should exist, very analogous to those which I thought I had discovered in N. lavigata. 560 DR. CARPENTER'S RESEARCHES ON THE FORAM1NIFERA. direction. Still their continuity is maintained through all the successive layers of which even the thickest part of the shelly disk may be composed. 106. Besides the radial and vertical systems of canals, there is an annular system, which traverses the thick band of shell-substance that usually intervenes between the successive annuli, and which is continually brought into view in horizontal sections (Plate XXIX. figs. 10, 12). It appears from vertical sections traversing the annular septa, that several tiers of these annular canals may exist. I have frequently traced them running continuously for a considerable distance, without appearing either to give off any branches, or to communicate with the radial canals; but I have occa- sionally seen appearances which indicate that such a communication is established by means of canals passing vertically downwards at the angles of the chambers, so as to unite the three sets of canals into one continuous system, furnished with a mul- titude of orifices upon the surface of the disk. A representation of the whole canal- system, as I believe it to exist in this organism, is given in Plate XXX. fig. 4. 107. The uses of this canal-system can only be a matter of speculation. Not having had the opportunity of examining specimens in which the soft animal substance had been preserved, I am unable to affirm whether the interseptal canals of Cyclodypeus are occupied in the living state by a portion of the sarcode-body, or whether they are empty ; but as I have unquestionable evidence that the former is the case in Poly- stomella, I should think there can be little doubt that it is also true of this genus. Now if we come to examine the purpose of this canal-system, we are at once struck with the fact, that it can scarcely be requisite for the nutrition of the segments of the sarcode-body enclosed within the chambers; since the mutual communication which these segments have with each other, seems fully as adequate for the purpose in Cyclodypeus, as it is in Orbitolites, Orbiculina, or Alveolina. If we examine wherein this organism so differs from the foregoing as to require such an additional system, we may find a not improbable answer in the possession of that additional skeleton which intervenes between the proper walls of the chambers ; for the canal-system, excavated in the very substance of this, would seem to furnish the appropriate chan- nel for its nutrition. And that such is its object, will be shown in a future memoir to be almost certainly proved, by the comparison of facts then to be adduced from the structure of other genera. 108. Monstrosities. Although the number of specimens of this type which I have had the opportunity of examining is but small, yet two among them exhibited the same kind of monstrosity as that which is common in Orbitolites ; namely, the super- position of a vertical plate upon the horizontal disk (Plate XXX. fig. 3). And in each it is sufficiently apparent that this plate has originated from the central cell, and that its increase has taken place pari passu with that of the horizontal disk. 109. General Summary. If, now, we review the principal facts relating 'to the structure of Cyclodypeus, and compare them with those furnished by Orbitolites on the one hand and by Nummulites on the other, we shall see that, notwithstanding GENUS CYCLOCLYPEUS : GENERAL SUMMARY. 561 its resemblance to the former in external aspect and plan of growth, it is far more closely allied to the latter in those features of its organization which indicate its physiological condition. It has been shown that in Orbitolites the communication between the different portions of the sarcode-body is so free, that the whole may be regarded as a continuous mass in which the segmental division is but imperfectly indicated (^[ 67); and this view is in complete harmony with the fact, that every addition made to the shelly disk forms (save in a few rare cases) an entire annulus. In Cycloclypeus, on the other hand, the chambers are so completely separated from each other laterally, that no other communication exists between them than such as may be established by the interseptal canals; and the communications between the chambers forming successive annuli, are only large enough to allow the passage of narrow bands of sarcode. Hence we see that there is here as much segmental inde- pendence as is consistent with the existence of these animals, which involves the maintenance of a communication between the innermost and outermost chambers, for the transmission of nutriment to the segments of the sarcode-body contained within the former. And this independence is strikingly manifested, by the frequency with which incomplete annuli are added to the previous margin of the disk, extend- ing (it may be) along not more than a third, a half, or two-thirds of the entire circum- ference. The want of constancy in the number and position of the communications between the chambers (^[ 101), even in this high type of Foraminiferous structure, is a point of fundamental importance in the determination of the value of the shape of the aperture, as a character of discrimination between genera and species in this group of organisms. I have elsewhere shown that a like want of constancy exists in Num- mulites (op. cit. p. 24), the different septa of one and the same specimen having aper- tures of very varied forms. We are not in a condition to assign a positive function to the minute tubuli that traverse the shelly layers intervening between the chambers and the two surfaces of the disk. But it is quite possible that these tubuli may give passage to pseudopodial prolongations of extreme minuteness, which may spread themselves forth from the whole surface of the disk, as we know that they do from the larger pores of Rotalia, and may coalesce so as to form upon it a continuous layer of sarcode, by whose instrumentality a lamina of shell is added from time to time to those previously existing. That either by giving passage to threads of sarcode, or by conveying organizable fluid, they furnish the means for progressive increase of the shell in thickness, would seem a very probable account of their use. Whether this hypothesis, however, be correct or not, there can be no reasonable doubt that the minute organization of these shelly layers in Cycloclypeus and Nummulites, an organization as high as that of dentine, is a feature of high elevation, as compared with the simple concretionary condition of the calcareous skeleton in the three genera previously examined. 110. Thus, then, in the almost complete isolation of the segments, in the enclosure of each of them in its own proper wall, in the interposition of an intermediate skele- 562 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. ton and of a canal-system between the contiguous walls of adjacent chambers, and in the minutely-tubular structure of the shell, all of them points of high physiological importance, Cycloclypeus differs entirely from Orbitolites, and agrees with Nummu- lites ; whilst it agrees with Orbitolites, and differs from Nummulites, in the single circumstance that its mode of increase is cyclical instead of helical, a difference which we have seen to present itself at two different periods of life of the very same specimens of Orbitolites and Orbiculina, and which must, therefore, be a character of quite subordinate importance. Genus HETEROSTEGINA. 111. The correctness of the views just advanced is fully borne-out by the occur- rence of a type, which bears precisely the same relation to Cycloclypeus, that Orbicu- lina bears to Orbitolites; one, namely, in which the form and connexions of the individual chambers, the minute structure of the shell, and the distribution of the canal-system, being essentially the same the plan of growth is helical, at least during the earlier period of life. This is the case with the genus Heterostegina, which was established by M. D'ORBIGNY in his memoir of 1825, but the essential structure of which he has altogether misapprehended. In his latest classification of the Foraminifera*, he ranks this genus in his order Entomostegues, which is com- posed of Foraminifera, whose segments are disposed in a spiral, but in two different planes alternating with each other, so as to render the entire shell inequilateral. Of the genus Heterostegina, which he ranks in close approximation to Amphistegina, he gives this definition : " Coquille a spire embrassante, dont les loges sont sepa- rees inte"rieurement par des cloisons trans versales." Now in the first place, I am quite satisfied that the chambers of Heterostegina do not alternate one with another, but are arranged in one plane about the same axis, as is shown in figs. 1, 7, Plate XXXI.; and secondly, it gives by no means a correct idea of its structure, to liken its chambers to those of Amphistegina save for their division by transverse partitions. 112. I have had the opportunity of examining, by the kindness of Mr. CUMING, a very extensive series of specimens of this genus (belonging, apparently, to the species H. costata, D'ORB.^), from the Philippine islands ; many of these are of large size, attaining as much as half an inch in diameter ; and the appearance of the adult speci- mens scarcely differs less from that of the young (which latter are alone figured by M. D'ORBIGNY), than it does in the case of Orbiculina. The dredgings of Mr. JUKES have furnished me with numerous specimens of Heterostegina from the Australian coast ; these closely correspond with the figures of M. D'ORBIGNY, being of compara- tively small size, and not exhibiting that peculiar mode of development which is characteristic of the adult. As the Australian forms correspond precisely with the young of the Philippine, there can be no doubt of their specific identity. I recognise the shells of the same species as almost the sole components of a fossilized deposit, * Cours 61ementaire de Paleontologie, torn. ii. p. 201. t Foram. Foss. de Vienne, p. 212. GENUS HETEROSTEGINA: ORGANIZATION. 563 which I understand to be very commonly met-with in Malta in fissures of the rocks, but of which the age is uncertain. 113. Organization. The older specimens of Heterostegina (Plate XXX. fig. 2) pre- sent a form which, when regular, may be characterized as discoidal. There is, how- ever, a knobby elevation or nucleus, which is usually somewhat excentric ; and from this the turns of a spire are seen to commence, the last of which usually becomes con- tinuous with one part of the margin of the disk (a b c), which there possesses a thick and defined border. As this spire opens-out, however, it becomes thinner and flatter; and this thinning is especially noticeable at that part of the margin of the disk (a dc) which corresponds with the opening of the spire. An examination of this portion of the disk shows that it precisely corresponds in structure with Cycloclypeus ; the form and disposition of the chambers, their mode of communication, the structure of their shelly walls, and the interposition of the intermediate skeleton and of its canal-system, being all points of such close resemblance, that, as there is positively no other point of difference than a somewhat inferior thickness of the intermediate skeleton between the successive rows of chambers in Heterostegina, a fragment of this marginal por- tion of the spirally-formed disk of Heterostegina might be taken for a fragment of the cyclical disk of Cycloclypeus, without the possibility of certainly distinguishing them, and vice versd. It is interesting to observe, moreover, how close is the con- formity of these two types, even as regards their irregularities ; for it will be seen, on an inspection of the figure, how little uniformity there is in the breadth of the successive rows of chambers, and how frequently it happens that a row is incomplete, just as in Cycloclypeus (^[ 96). 1 14. If, now, we examine the structure and arrangement of that spirally-coiled por- tion of the disk, which constitutes its nucleus, and which is best shown in younger specimens, we see that, as in the other cases, the first chamber (Plate XXXI. fig. 1, a) is globular, that the second (b) buds-forth from one side of this, and each successive chamber from the outer side of the preceding, just as in Nummulite or any other simple helical form. But before one turn of the spire is completed, each newly- formed chamber is seen to be double (cc) instead of single, a small portion being divided-off (as it were) near the marginal part of the whorl ; and just about the part where the second turn is completed, the gradual opening- out of the spire gives room for the interposition of a third chamber in each ro.v (d) ; and the number is soon further augmented, in accordance with the progressive increase in the breadth of the spire, the dimensions of the individual chambers retaining a pretty close conformity to a constant average. An examination of this figure will further show, that the increase in the number of chambers in successive rows always takes-place at the inner margin of the spire; some of those nearest the outer margin dying-out, as it were, without giving origin to new chambers in the next row. This may, I think, be connected with the fact, that there is always a large free opening (e, e) between one row of chambers and the next, at the inner margin of each spire (the situation of the open- MDCCCLVI. 4 E 564 DR. CARPENTER'S RESEARCHES ON THE FORAMIN1FEUA. ing in Nummulite}, and that the chamber of the row abutting on the preceding whorl is nearly always much larger than the rest, and gives origin to two or even three chambers in the next row. Further, it is shown by vertical sections (Plate XXXI. fig. 7), that the innermost chambers of the whorl are not only broader but thicker, their upper and under walls diverging from each other where they are to be con- tinued over the spire they invest. Hence it is pretty obvious, that this portion of the whorl is that wherein the most active nutrition takes place ; and it is here that the marked accession to the number of chambers occurs, which tends to carry the later rows around the whole circumference of the disk. 115. Each of the early turns of the spire not only surrounds, but completely invests its predecessor; as is best shown by a vertical section, such as that represented in fig. 7- The investing whorl does not, in the younger part of the spire, come into immediate contact with the two surfaces of that which it includes, but is separated from it by the prolongation of the chambers and of their septa, very much as in ordinary NumrnuUtes. But between the later whorls, there are no such interspaces. The successive layers come into absolute continuity with one another ; both the tubuli and the cones of non-tubular substance being continued from each into the one external to it. From the time that the rapid thinning-away and opening-out of the spire commences, the investment of the previously-formed whorls seems to discon- tinue. It is at the margin of each whorl, that we find the intermediate or additional skeleton most remarkably developed ; and the canal-system sometimes forms quite a network in its substance (fig. 11). 116. General Summary. It is obvious, from the foregoing details, that the physio- logical condition of each individual segment of the animal of Heterostegina must be essentially the same as that of each segment of Cycloclypeus ; and that the only difference in the condition of the two organisms arises out of the mode in which these segments are increased in number. In Cycloclypeus, in which each, row of seg- ments is (normally at least) a complete annulus, a new annulus is formed around its predecessor by gemmation from its several segments along the entire circumference ; and this mode of increase may be traced-back to the central cell, which buds-out equally on all sides. But in Heterostegina, each row is limited by the breadth of the spire; and while most of its chambers are formed by the like kind of gemination from their predecessors, there is a special provision for an augmentation in the number of chambers at the end of the row nearest the previous whorl. Tracing-back the spire to its origin, we find that it commences in the one-sided gemmation of the central cell, just as we found it to do in Orbiculina (* 85) and in those forms of Orbitolites which have a spiral commencement (^J 54). Hence there is nothing but the plan of increase, which separates Heterostegina from Cycloclypeus ; and their relation is exactly the same as that of Orbiculina and Orbitolites. For in Heterostegina, as in Orbiculina, the first-formed portion is a spire, of which each turn invests its predecessors ; but after three or four turns have been made, the spire spreads-out, CONCLUDING REMARKS. 565 tends to surround the whole disk with its mouth or growing margin, and thencefor- ward the growth is cyclical, as in Cycloclypeus and Orbitolites. Concluding Remarks. ll/. Looking now to the general organization of the five genera which I have described, and to the peculiarities by which I have shown that each is characterized, I think that we are in a position to inquire into the value of the system of classifica- tion which has been erected by M. D'ORBIGNY on the exclusive basis of plan of growth ; on which inquiry, the facts which I have now brought together have an obvious and direct bearing. I. The very close physiological relationship which has been shown to exist between Orbitolites and Orbiculina, requires that they should be associated in the same Family, if not in the same Genus. In the classification of M. D'ORBIGNY they are ranked under different Orders (Cyclostegues and Hellcostegues). II. In like manner, the close physiological relationship which has been shown to exist between Cycloclypeus and Heterostegina requires that they should be associated in the same Family, if not in the same Genus. In the classification of M. D'ORBIGNY they would be ranked under different Orders (Cyclostegues and Entomostegues). III. Again, the strongly-marked physiological difference which has been shown to exist between Orbitolites and Cycloclypeus, would seem to require that they should be separated by the widest possible interval ; yet the system of classification adopted by M. D'ORBIGNY would have forced him to associate them (if he had been acquainted with the last-named type) side by side in the same Order (Cyclostegues}. IV. In like manner, the corresponding difference which has been shown to exist between Orbiculina and Heterostegina, would seem to require that they should be separated by the widest possible interval ; yet the system of classification adopted by M. D'ORBIGNY would have forced him to associate them (if he had been acquainted with the real plan of structure of the last-named type) side by side in the same Order (Hellcostegues) . V. The doctrine which I base on the foregoing facts, that physiological con- formity in the condition of each individual segment, as indicated by the structure of its shelly investment, is a character of primary importance, whilst the plan of growth, that is, the mode of increase in the number of chambers, is a character of subordi- nate importance, is further borne-out by the following considerations : 1. In Orbitolites, the general plan being cyclical, the early plan of growth is fre- quently spiral. 2. In Orbiculina, while the early plan of growth is uniformly spiral, and this is sometimes continued throughout life, it is very commonly exchanged in adult age for the cyclical. 3. In Alveollna, whose physiological approximation to Orbitolites and Orbiculina is unquestionable, a plan of growth is followed, which differs more from 4 E 2 566 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. that of either of them, than the plans of the two latter differ from each other. 4. In Heterostegina, as in Orbiculina, the early plan of growth being uniformly spiral, there is a tendency in adult age to the assumption of the cyclical. 118. I think myself justified, therefore, by the foregoing comparisons, in asserting that the system of classification proposed by M. D'ORBIGNY is founded on an estima- tion of the value of characters, which is entirely erroneous ; and that any classifica- tion which shall be really natural, must be based on an order of facts relating to the economy of the animal, of which his imperfect methods of observation have left him in entire ignorance*. It is not my intention, in this stage of the inquiry, to propose the erection of any new system ; my sole aim, at present, being to establish the funda- mental principles upon which alone can a natural arrangement be securely built-up. EXPLANATION OF THE PLATES. PLATE XXVIII. Figs. 1-5. Successive stages of growth of the ordinary type of Orbiculina adunca, showing the change from the spiral to the cyclical plan of development : 16 diam. Fig. 6. Edge of a disk of Orbiculina adunca, showing but a single row of apertures, as in the simple type of Orbitolites: 50 diam. Fig. 7- Edge of a disk of Orbiculina adunca, showing three rows of apertures, as in the complex type of Orbitolites: 50 diarn. Figs. 8-10. Successive stages of growth of the less common type of Orbiculina adunca, in which the spiral plan of development is retained throughout life: 16 diam. Fig. 11. Horizontal section of a disk resembling fig. 5 : 16 diam. Fig. 12. Horizontal section of a spiral resembling fig. 10: 16 diam. Figs. 13-16. Portions of the superficies of Orbiculina adunca, showing varieties in the surface-markings : 50 diarn. Fig. 17. Central portion of a disk of a fossil Orbiculina (Orbitolites Malabaricus, CAR- TER), showing its spiral commencement on the plan of fig. 8: 16 diam. Fig. 18. Marginal portion of a similar disk, showing the investment of the early whorl by the later, with the characters of the surfaces and edge : 50 diam. Fig. 19. Marginal portion from a similar fossil disk, which had extended itself like fig. 5 : 50 diam. * I am constrained to make a similar remark respecting the classification proposed by Professor SCHULTZR (Ober den Organismus der Polythalamien) ; which, although in many respects an improvement upon that of M. D'OBBIGNY, is almost equally far from representing the natural affinities of these organisms, as revealed by minute investigation of the structure of their testa. EXPLANATION OF PLATES. 567 Fig. 20. Inner surface of one of the annul! of a similar fossil disk, showing the differen- tiation of the superficial from the intermediate layers of cells: 50 diam. Figs. 21, 22. Portions of the superficies from similar fossil disks, showing varieties in the surface-markings : 50 diam. Fig. 23. External aspect of Alveolina Boscii (recent) ; a, a, growing margin, showing multiple apertures resembling those at the margins of Orbitolites and Orbi- culina: 40 diam. Fig. 24. Longitudinal section of Alv eolina Boscii, showing its internal structure and the successive stages of its growth : 40 diam. * s PLATE XXIX. Fig. 1 . Section of disk of Orbiculina adunca parallel to the surface, showing the cells and their communications: 100 diam. Fig. 2. Surface-layer of disk of Orbiculina adunca, showing its punctuated appear- ance : 100 diam. Fig. 3. Vertical section of disk of Orbiculina adunca, passing through its central nucleus, and showing columnar arrangement of its cells, and the manner in which the earlier whorls of the spire are invested by the later: 100 diam. Figs. 4-7. Transverse sections of Alveolina Boscii, showing the increase in the num- ber of turns of the spire from its terminal to its central portion : 40 diam. Fig. 8. Portion of a similar transverse section enlarged; showing a, a, a, internal prolongations of the surface-layer ; b, b, b, outer longitudinal canals ; e, c, c, inner longitudinal canals : 80 diam. Fig. 9. Portion of a longitudinal section (Plate XXVIII. fig. 24) similarly enlarged ; showing 1, 2, 3, 4 successive layers formed by the involution of the spire; and in each the passages a a, b b, and c c, between one band and the next. (N.B. The variation in the appearances presented by the other layers, depends upon the difference of relative direction in which the sec- tion traverses each of them respectively) : 80 diam. Fig. 10. Thin section of Cycloclypeus, taken parallel to the surface, and close to the covering of the chambers ; showing part of the system of interseptal canals : 50 diam. Fig. 11. Diagram of a single chamber of Cycloclypeus, showing its relations to other chambers, and to the interseptal system of canals: a, cavity of chamber; b, b', adjacent chambers of the same annulus, each separated from a by a double septum; cc' nnddd', chambers of internal and external annuli, separated from a by the annular partitions ee, eV, but communicating with it by the passages f, f, f, f; in the septa between a and b, b' are seen the interseptal canals, each of which sends two oblique branches across the annular septa, to communicate with corresponding canals in the septa 568 DR. CARPENTER'S RESEARCHES ON THE FORAMINIFERA. dividing cc' and dd' ; these interseptal canals seem to communicate by short lateral twigs with the cavities of the adjacent chambers, whilst at g> g> S^ S tne y become connected with vertical branches, which unite them with those of other planes: at h h, h' h' are seen the canals proper to the annular septa. Fig. 12. Thin section of Cydoclypeus, passing through its central plane, showing por- tions of four annul!, 1 1, 22, 33, 44, with the general relations and con- nexions of their chambers, the double septa by which they are separated, and (in parts) the interseptal system of canals: 36 diam. PLATE XXX. Fig. 1. General view of a disk of Cydoclypeus, showing the aspect of its surface, and the appearances presented by horizontal and vertical sections: 12 diam. Fig. 2. Surface-view of a full-grown specimen of Heterostegina, showing its tendency to assume the discoidal form by the opening-out of the spire: a be, the thick- ened margin of the spire; cda, its growing edge or mouth : 10 diam. Fig. 3. Monstrous specimen of Cydoclypeus, having a vertical plate superimposed upon the horizontal disk : 5 diam. Fig. 4. Ideal figure of a portion of a Cycloclypeus-d'isk laid-open to show the details of its structure: a, a, a, upper stratum, consisting of superimposed tubular laminae; b, b, b, portion of lower stratum ; c, c, c, cones of non-tubular substance, sometimes perforated by larger canals ; d, d, d, their bases pro- jecting on the surface ; e, e, plates of non-tubular substance, continuous with the septa between the chambers ; f,f, passages of communication between the chambers, through the inter-annular partitions, as seen in section ; g,g, the same as seen from the interior of the chambers ; h, h, inter- septal canals cut across ; i, a chamber on the walls of which the system of interseptal canals is represented as fully displayed ; k, k, passage of the principal canals along the line of junction between the roof of the chambers and the vertical septa : 60 diam. PLATE XXXI. Fig. 1. Section of a young specimen of Heterostegina, taken parallel to the surface, partly through the chambered plane and partly (owing to a slight inequality of the specimen) through the shelly investment of the chambers, the boun- daries of which, however, are still distinguishable, owing to the difference of texture between the shell that covers the chambers and that which unites with the septa : a, first cell ; b, second cell ; c c, first subdivision ; f, second subdivision ; e, e, large passages connecting the chambers that abut on the pre-formed whorl : 60 diam. EXPLANATION OF PLATES. 569 Fig. 2. Vertical section of Cyclodypeus, taken in the direction of one of the annuli, showing- the upper and under layers a b, a 1 b', with the septa between the chambers, each of which is seen to be composed of two layers with an interposed lamella, pierced by two or more canals whose orifices are shown ; on the left side of the figure, the chambers are seen to be closed by the inter-annular septum, which is pierced by several irregularly-disposed passages, c, c, that establish communications between each chamber and those of the annuli internal and external to it : 60 diam. Fig. 3. Section of Cyclodypeus, taken horizontally (or parallel to the surface of the disk) through the shelly covering of the chambers, showing its minutely tubular structure, and its junction with the two interannular non-tubular septa a b, a' b', and with the intercameral septa c c, c' c', along the course of which last are seen the divided ends, d, of vertical interseptal canals : the chamber thus inclosed is remarkable for its great length in comparison with its breadth : 100 diam. Figs. 4 & 5. Vertical sections of Cyclodypeus, corresponding with fig. 2, and showing irregularities in the division of the chambers: at cc, fig. 4, are seen the orifices of the interseptal canals laid open ; and at c, fig. 5, is shown a part of the vertical canal system, passing upwards and downwards towards the two surfaces of the shell, through its transparent portions: 60 diam. Fig. 6. Section of Cyclodypeus, taken in the same direction as fig. 3, but nearer to the surface, showing the relation of the tubular and non-tubular portions of the shell, and the orifices of the canals : 100 diam. Fig. 7- Vertical section of young specimen of Heterostegina, showing the investment of the first-formed whorls by those which succeed them, and (on the right) the commencement of the thinning-out: a, a, a, passages of communica- tion between the successive bands of which each whorl is composed : 35 diam. Fig. 8. Vertical section of young disk of Cyclodypeus, showing at a an irregular sub- division into two or three layers of chambers: 22 diam. Fig. 9. Horizontal section of a portion of Cyclodypeus, corresponding with fig. 3, but more highly magnified : 150 diam. Fig. 10. Vertical section of the floor of one of the chambers, showing its lamellated arrangement, the minutely-tubular structure of one portion of it, a a, and one of the non-tubular cones, bb, which is continuous with the annular septum c : 150 diam. Fig. 11. Horizontal section of a portion of Heterostegina, taken near the margin, showing the disposition of the chambers, the communications a a between those of successive bands, the minutely-tubular structure of their roof, the interseptal canals, and the high development of the canal-system in the thickened non-tubular margin, bb: 150 diam. PJiiL.Trans.MDCQCLT: 'X'TIIL k?V ****>*'/;** ;*,*i '.''*!" *"J(H ^^^'. .***. *_'....* **^KJ ' Phil. TransMLCCCLVI Plate, Thl Trims. mcCCI,~rr. Pl&te-XZX . Trans. MDCCCW. Plate IXIf. '::;' '..' '- Wcet., M a, Heh. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED BIOLOGY LIBRARY TEL. NO. 642-2532 This book is due on the last date stamped below, or on the date to -which renewed. Renewed books are subject to immediate recall. 3 11968 M LU 21A-12m-5,'6S (J401slO General Library University of California Berkeley I