The author recommends the study of the form and development of the embryonic parts of the skeleton as affording the best criterion of the correctness or want of correctness of the alliances here suggested.

Fossil Sponges in the British Museum.*-This fine work from the pen of Dr. G. J. Hinde, a pupil of Prof. Zittel, the great leader in the modern development of the paleontology of sponges, is a fitting tribute to the excellence of the principles of classification which have been laid down by the distinguished professor. Its classification is based essentially on the principle of the employment of the minute structure of the skeleton for its fundamental distinctions. While the adoption (in the Introduction) of the older, Häckelian, grouping of the tissues of living sponges into syncytium (ectoderm and mesoderm of Schulze) and ciliated cells (entoderm of Schulze) is not a happy feature, yet, on his own ground, Dr. Hinde does good service in his careful account of the mineralogical characters of fossil sponges. He upholds the ready replacement of organic silica by calcite. Diagnoses are given of the British species and the new foreign ones; references are given in the case of the remaining described species from foreign horizons; the rich collections of William Smith, Toulmin Smith, Mantell, and Bowerbank render the British part of the collection particularly interesting, and the work may be regarded as a manual of British fossil spongology.

The new genera are 27 in number; of these, Climacospongia, Lasiocladia, and Acanthorhaphis are siliceous Monactinellids; the first strongly resembles Reniera in the recent series. Acanthorhaphis is perhaps related to the recent Metschnikowia of the Caspian Sea. The Lithistidæ, as might have been anticipated, contribute largely to the long list of new types, viz. 10 genera. Of these, the Megamorina are represented by Placonella, Holodictyon, Pachypoterion, Nematinion, chiefly distinguished by general form or by points in the arrangement of the internal "canals." To the Tetracladine group of Lithistids Dr. Hinde adds no less than six new types, viz. Bolospongia, Kalpinella, Thamnospongia, Pholidocladia, Phymaplectia, Rhopalospongia; in the last alone does any considerable divergence from the normal character of Tetracladinæ appear in the spiculation, viz. a part of it inclines towards the Rhizomorine type. It is noteworthy as illustrating the advance made by the new system of classification, that a close resemblance is to be traced between the external form of some of these genera and that of genera belonging to quite different groups. Considering its wider range in time and its greater comprehensiveness, the order Hexactinellida is not so richly represented by new types in this collection as the Lithistida. Among the forms with a continuous skeleton (Dictyonina), the Euretidæ contribute two such types, viz. Strephinia, which forms irregular or cup-shaped expansions-a habit unusual in the recent members of this family, and Sestrodictyon. Ventriculitidæ are represented by a new form, Sestro

*Catalogue of the Fossil Sponges, in the Geological Department, &c., with descriptions of new and little known species.' 4to, London, 1883, 248 pp. (38 pls.).

cladia, which is remarkable for its dendroid growth; the branches are hollow. Of the Staurodermide of the collection the new forms are Placotrema (allied to Porospongia), Cnididerma (distinguished by having the level dermal surface divided up regularly into squares by the arrangement of its spicules), and Plectoderma, differing slightly from Dictyophyton, but of which the form is unknown. The Callodictyonida have Porochonia, based on an old species of Ventriculites provided with a delicate surface-tissue besides the usual dermal layer, and Sclerokalia, a nest-shaped sponge, with vertical rows of apertures on the inner surface, and shallow canals leading from them. No new Lyssacine Hexactinellid genera are described.

The Calcarea, as defined by Dr. Hinde, are very numerous, owing to the inclusion by him of the Pharetrones (distinguished by the possession of a fibrous skeleton) in the group, in which course he follows Zittel and the more recent views of Steinmann and Dunikowski; this step is in partial opposition to Carter and Sollas, who regard some, at any rate, of its members as siliceous. Unlike Dunikowski, who places them under the Leuconidae, he regards them as constituting a distinct family; he relies largely on the character of the fibre and the methods of arrangement of the spicules in it, for his definition of genera. Few species are described as new. The new genus Tremacystia unites a number of already known species, distinguished by a metameric segmentation of the sponge. Inobolia is distinguished by its form and the absence of canals. Trachysinia has a cylindrical form and may be compound; it is based on three new Jurassic species. Diaplectia has the growth of Pharetrospongia, but contains tri- and quadriradiate spicules. Raphidonema has elongate triradiate spicules like those of Corynella. Among the Calcarea, but as incerta sedis, is introduced Bactronella n. g., from the Upper Jura; it resembles the recent Leuconidæ, but the spicules are spinous. No Horny sponges find mention.

Tables and lists are given showing the known distribution in time of all the species, from which it appears that the Cretaceous beds contribute to the collection by far the greatest number, viz. the large total of 250 species, of which 103 are Lithistidæ, 85 Hexactinellida, and 46 Calcarea (including Pharetrones); the total number of species enumerated is 399. A bibliography is given.

Vosmaer's Manual of the Sponges.*-G. C. J. Vosmaer completes the review of the literature of sponges commenced in the first instalment of this work. He devotes seven pages to an account (A) of the best methods of investigation, under the heads (1) Investigation of the soft parts.-Killing and preserving, staining, preparation and preservation of sections, decalcification and desilicification. (2) Investigation of the skeleton.-Skeleton of the Calcareous Sponges, of the true Horny Sponges, of the Siliceous Sponges. (B) Preservation for collections. (C) Rearing larvæ under the Microscope. For hardening, absolute alcohol, picro-chloric and osmic acids, and corrosive sublimate;

'Dr. H. G. Bronn's Klassen und Ordnungen des Thierreichs. Band ix. Porifera. Lief. 3-5, 1884, pp. 65-144 (10 pls.). See this Journal, ii. (1882) p. 797. † See this Journal, i. (1881) p. 611.

for staining, hæmatoxylin, picrocarmine, iodine, chloride of gold, and nitrate of silver are respectively recommended. Under the heading Morphology is given a general sketch of the range of shape, size, colour, consistency, and character of the surface in the group. Under Anatomy is given an account of the different parts of the canal system of sponges, and their chief modifications. The four types under which the leading modifications of this system are arranged by the author in a previous work are adopted here also.

Protozoa.

Nucleus and Nuclear Division in Protozoa.*-A. Gruber passes in review the different groups of the Protozoa; commencing with the Rhizopoda, he points out that, though their nuclei diffor considerably, they are all referable to the type of the so-called vesicular nucleus. There is a more or less distinct nuclear membrane, and a clear and apparently homogeneous nuclear substance in which are deposited one or more nuclear corpuscles. Such nuclei are to be found in the lowest myxomycetoid plasmodia, and Bütschli is probably right in regarding this form of nucleus as the primitive one; previous to this, however, there was, in all probability, a stage in which small granules of nuclear substance were scattered through the whole of the protoplasm, and these were only later collected into a proper nucleus. As a fact, there are organisms which exhibit such characters; as, for example, some of the forms described by Maupas, the very low Trichosphærium sieboldi (Pachymyxa hystrix), and, probably, the Pleurophrys gennensis discovered by the author. In all these we find small spheres which are strongly coloured by certain reagents scattered through the body. Moreover, as Brandt was the first to show, Amoeba proteus contains not only a definite nucleus, but also small granules of nuclear substance.

Amaba verrucosa is cited as an example of a form which, though it seems to have a very definite vesicular nucleus, is found on examination with higher magnifying powers (e.g. Hartnack Oc. 3, Obj. 12 Imm.) to have its nuclear corpuscles made up of smaller spherules. When stained, these bodies gradually become less distinctly visible; there appear in the substance of the nucleus excessively fine granules, so fine as to have the appearance of a red dust; these would seem to be true chromatin particles, which may become united into fine filaments; they form lines arranged radially around the nucleolus. They are best seen in specimens that have been treated with absolute alcohol or picrocarmine. Although, therefore, there is a nuclear network in Amoeba verrucosa it is very incomplete, and, as observation has shown, takes no part in the division of the nucleus. Multinucleolar nuclei are derived from the uninucleolar by repeated division of the nucleolus.

The division of the vesicular nucleus is effected by constriction or by cleavage. In the former case, the chromatic substance is first diffused through the whole nucleus; in the latter case the nucleolus

* Zeitschr. f. Wiss. Zool., xl. (1884) pp. 121–53 (2 pls.).

divides first, the halves separate from one another, and then the rest of the nucleus is cut through.

Among the Rhizopoda two other kinds of nuclei are also seen; in one of them we distinguish a nuclear membrane, and substance, within which are scattered, more or less irregularly, particles of chromatic substance. On division these become arranged into filaments, which, at first coiled, become later on arranged parallel to the long axis of the extending nucleus, and so are equally divided on its constriction. The other form is distinguished by the presence of a cortical zone, generally consisting of granules, which lies just beneath the nuclear membrane. Here there is but little nuclear substance and a large central nucleolus. In division, the nucleolus divides first, and the parts separate from one another; the cortical zone is then divided equatorially, and finally the whole nucleus is cut through.

Lastly, the nuclei of some Foraminifera are remarkable for being distinguishable into two halves, one of which is quite filled with chromatic substance, while the other has one or more nucleoli. The mode of their division is as yet unknown.

The Heliozoa are next taken up, and their nucleus found to consist of a nuclear membrane, clear nuclear substance, and a central nucleolus, or there are several nucleoli; or, finally, there is a membrane, a cortical layer, nuclear substance, and central nucleolus. In this last, division always begins; when there are several they unite into two plates, which separate from one another.

In the large nuclei of the intermediate Radiolaria, we find (a) vesicular forms, exactly like those of many Rhizopoda and Heliozoa, (B) nuclei with a cortical layer (as in Actinophrys); (v) nuclei with a very strong membrane, dark and often granular nuclear substance, in which radiating bands may sometimes be seen; and (8) nuclei with a plexiform arrangement of the chromatic substance, and nucleoli imbedded in the meshes; the fissive methods of none of these are satisfactorily known.

The small nuclei of the multinucleate Radiolaria are either amoeboid and divide by simple constriction, or are quite round or oval when division commences with the radiate arrangement of the chromatic substance.

The nuclei of the Gregarinida have a vesicular structure, and one or more nucleoli; their mode of division is not known. The nuclei of the spores are quite homogeneous and divide by constriction.

The different groups of the Infusoria are discussed separately; in the true Flagellata we find vesicular nuclei, which divide by the regular constriction of all the parts, and the formation of parallel longitudinal lines in the nucleolus. In Noctiluca the nucleus forms a granular mass in which nucleoliform corpuscles are distributed; but in Leptodiscus the nucleus is formed of a larger, darker, and granular portion together with a smaller and clearer part. Unlike Rotalia, the hyaline part here contains the chromatic substance. In Noctiluca, as observed by Robin, the nucleus elongates, and the central part becomes longitudinally striated. In the Cilio-flagellata

the nucleus is formed on the "massive" type; that is to say, the nuclear membrane incloses a thick mass of nuclear substance, which, in all probability, contains the chromatin in the form of small granules. So far, the nucleus of the Cilio-flagellata resembles that of the next group the Ciliata.

The description of the nuclei of the Ciliata offers considerable difficulties in consequence of the numerous variations which are to be seen in the structure of even closely allied species. The nuclear substance may be so dissolved in the cell-substance, and the granules may be so fine as to be only distinguishable with the highest powers; or the constituents of the nucleus may be larger, and formed (as in Oxytricha) of spherical corpuscles which, before division, unite into a mass. This substance may form bands and plexuses, and sometimes, as in Benedenia and Plagiotoma, break up into pieces; this leads to the rosette-like nuclei of Stentor, or the band-like nuclei of Vorticella. It is rare for an Infusorian to have more than one nucleus, but the number of the paranuclei is by no means so constant. The nucleus is generally "massive" and surrounded by a membrane; its substance is very rich in chromatin-granules, which are very variously arranged; the paranuclei are likewise massive, and apparently always granular. On division, the chromatin-granules form filaments which lie parallel to the long axis of the nucleus, and become constricted in the middle.

The nucleus of the Suctoria, the last group of all, is either branched or rounded; there is a thick massive nuclear substance, in which chromatin-granules are often very distinctly visible. On division, the nuclei break up into filaments which undergo constriction.

The author thus sums up the results of this important investigation:

There are Protozoa in which the nuclear substance may be distributed through the protoplasm of the cell in the form of numerous granules; and these are often so small that after staining they only appear, on examination under high powers, as a precipitate. In others there are nuclear particles of this kind, but they are not only more numerous, but are also larger, and, in fact, more regularly arranged, so that they may be better spoken of as small nuclei; these lead us to the truly multinucleate forms. He thinks it possible that in those Protista which appear to us to be non-nucleate, the nuclear substance is more or less completely dissolved in the cell-substance; and that in the history of race development there was not at first a definite and formed nucleus, but rather fine nuclear granules. In any case, the formation of a true nucleus is intimately associated with the process of reproduction, and, primarily, with regular division.

A most important piece of evidence is afforded by those Protozoa which, after conjugation and division, are for a time filled with small nuclear particles. It would appear that there is a regular distribution of the chromatin in the daughter-individuals.

The nuclei of the Protozoa belong, as a rule, to one of two types: they are either vesicular, as in most Rhizopods, Heliozoa, Sporozoa, and all true Flagellata, as well as in some Radiolaria and Ciliata, or