mines single vision by corresponding points; but whether is any such fusion of corresponding fibres as supposed Müller, or any such reflection of illusive images from eye as supposed by M. Pictet, can never by determined e by anatomical investigation; and even if so determined in affirmative, could not possibly show itself in any visual pl menon, since by supposition every such illusive image mus absolutely identical with, and absolutely inseparable fro real image seen by the other Oakland, California, June 9, 1871.

eye.

ART. LIV. The American Spongilla, a craspedote, flag th Infusorian; by H. JAMES-CLARK, A. B., B. S.. Prof Hist., Kentucky University, Lexington, Ky. (With a P

THE argument of Hæckel, and others, that the Sponges essentially compound Polypi, is virtually based upon the s sumption that the minor (afferent) and major (efferent) o of the former correspond to the mouths of the latter; and t the profusely branching afferent and efferent canals of the S ges are strictly comparable with similar canals in the polypica of Halcyonarians: and, by implication, that the cilia-bear cells of the interior, lining-wall of the Zoöphyte find their be? logues in the ciliated, cell-like bodies of the interior chart of the Porifera. If, now, it should turn out that these last re not altogether mere cell-components of a tissue, but are e severally, an independent body, although closely conne with others in a coinmon bond, then the attempted paralle between the two groups must utterly fail of confirmation. tendency of Carter's later investigations, and our own too, show that this is no vain supposition.

For ourselves, we hold that each ciliated body of the Sp is a cephalic member (a cephalid in this case) of a polycepe individual. We believe, as far as we can understand his decided, rather hesitating position, Carter's latest decis that the Sponge is a community of Amoebous individuals+d not a polycephalic unit. Yet, whichever view prevails tendency is the same, and the Polyp theory is negatived: unquestionably. The incompatibility of the interior orgats of the two groups, above mentioned, is so great that it w seem as idle to elaborate a proof of it, as to attempt the de stration of an axiom. The question is really circumscr

* See our article on "Polarity and Polycephalism," this Journal, January, + See Carter, On Fecundation in the two Volvoces; on Eudorina, Sponge Annals and Magazine Nat. Hist., January, 1859; also for July, 1871, Sponges, &c.

ording to the method of Hæckel, to arguing that, since a sysof branching canals in the Sponge reminds one very ongly of the intricate network of passage-ways in the basal ts of certain Polyps, therefore the two are homologous, and r an identical relation to the rest of the organism. Carter i new Sponges, &c., Ann. Mag. Nat. Hist., July, 1871) has anred this far-fetched homology with considerable detail in ecent paper; and we do not, therefore, feel called upon to 1 more to it.

The principal aim of this article is to furnish new material proof of the polycephalism of the Spongiæ, and particularly regard to their relation with the Protozoa Flagellata. We highly pleased to find that Carter has lately (ut sup., July, 1) confirmed our earliest observations* as to the organization che collar-bearing monads of Leucoselenia, by an investigation Grantia compressa. He has also accepted our interpretation the horn-like processes of the sponge-cell of Spongilla alba; It they are the outlines of a membraneous collar in profile. We have now to bring forward a fourth example of a craslote, flagellate monad-cephalid in a Sponge. It seems to be Spongilla, but specifically, at least in its monads, it differs m the English forms. For convenience sake, we will call it mgilla arachnoidea, from its resemblance to an irregular ler-web. It lives in fresh water streams and ponds, usually but the bottom of the stems of water-plants, or wherever re is considerable shade; apparently avoiding the light, as seldom, if ever, found it in open water. In size it varies m a few inches to half a line in diameter; of no definite pe; and has a uniform fuscous or yellowish-brown color; I is wrapped about by a filmy, transparent, colorless envee ("investing membrane" Carter). The brown color is inherent the interior mass, in which the groups of monads are imIded; in fact the latter are themselves as strongly colored by own granular contents. The "investing membrane " is also htly tinged with amber color by the large and small spicules ich are imbedded in it. Excepting in very small specimens, eign matter is often so thickly spread over the surface as to scure the view and seriously interfere with a correct intertation of the relation of parts. We have been most fortunate our endeavors with the minuter individuals, which occasion, we found, would allow a view through and through their enbulk, and of course left full opportunity for a satisfactory dy of the details of special parts, without our resorting to the secting needles. Anyone who knows, by experience, the inse contractility of the living sponge, can appreciate the adMemoirs Boston Soc. Nat. Hist., vol. i, 1867, "On the Spongia Ciliutæ as Inria Flagellata."

vantage of not being obliged to destroy and sever parts organism from their natural relations. Premising thest everything has been studied "in place," even to the detai the monads, we shall endeavor to describe this sponge as were to be the type for future comparison.

C

General plan. The whole individual sponge is endo with a double envelope (fig. 1 a,a', c, d,) the outer and i parts of which are directly continuous into each other at points. The outer division (a, a1) lies at a considerable dista from the monadigerous mass (g), and is, as it were, suspen on the points of the larger, far-projecting spicules (e); just si tent canvas is supported on the ends of poles. The inner vision (c) closely embraces the monadigerous mass like epidermis, and even plunges between the hollow group monads, forming to them a basis of support. The outer: inner divisions are continuous with each other at many pois as stated just now, but only where the larger spicules proje There the envelope (d) runs along the spicules, completely bracing them, as if in a sheath, from their tips to their bases where they rest on the brown mass of monads. In brief. e might say that the sponge is covered with a miniature col nade, whose ceiling is the outer division of the envelope, the pillars are the bundles of spicules, and the floor is tapestried by the inner division, which about the pillars hangs from the cel ing in lofty folds. The continuity of the outer division of the envelope is broken by numerous, round or oval openings, of various and frequently changing sizes, sometimes very large which allow a free ingress of the water to the space just he neath. These are the afferent ostioles (os), through and int which a constant current of floating particles may be seen mor ing with considerable vivacity. Here and there, scattered 21 wide distances, finger-like, hollow processes from the outer di vision arise singly, and at various angles. Each is terminate by a large aperture, the efferent ostiole, from which a curre of water and floating matter emerges with more or less spas modic irregularity. The smaller individuals, from half a re to half an inch in diameter, possess only one such ostiole; and those an inch in diameter seldom have more than two or thre like conduits; but they are very large, sometimes a quarter o an inch in length when fully extended, and of the proportion and taper of the human forefinger.

Plunging the focus of the objective to the floor of the colon nade, the inner division (c) there is found to be pierced by much more numerous openings (i), but far smaller in diameter and quite methodically arranged, each one corresponding t and overlying a hollow group of monads (h). The outer divis ion is further embellished with irregularly scattered minute

icules (e1), which lie imbedded in the cytoblastema, parallel th the surface of the envelope, and occasionally crossing ch other at various angles. To complete this general sketch, will state more definitely the relation of the constituents of e monadigerous mass. There are essentially but two eleents here; namely, the inner division (c) of the investing embrane, and the groups of monads (h) which are imbedded it, below its surface. In a fully expanded individual these oups seldom lie so closely as to touch each other. They ry considerably in size and are usually globular or spheroidal, d form a single stratum, with rather narrow interspaces (c1) etween them.

It seems proper here, at least for the sake of precision, that e cytoblastematous basis, in which the monad groups are imedded, should be considered apart from the epithelium-like, ner (c) investing membrane which overlies it, although the 70 are essentially one; the epithelioid membrane, by prolongg itself between (at c1) and beneath the groups, forming for em a continuous foundation. In this light, then, we shall eak of the monadigerous mass as consisting of three elements, mely, the inner investing membrane proper, the group of onads, and the cytoblastematous basis. This basis seems to onstitute a large part of the bulk of the body, since it occupies 1 of the interior space beneath the monad groups. In speciens which grow over flat surfaces in depressed patches, or ound stems of plants, it forms a relatively thin layer; but here the body stands out an irregularly rounded mass, somemes an inch in diameter, the cytoblastematous basis fills up he interior, in enormous proportion to the bulk of the monad yer.

ORGANOGRAPHY.

The Investing Membrane. The investing membrane (fig. 1, a, , c, d,) consists essentially of two histiological elements, namely, very diffuse cytoblastema (a1) and irregularly disposed cells , b1, b2) scattered through it. The intercellular cytoblastema orms a very thin layer (a1) between the cells (b); but where e latter are imbedded in it, its outer and inner faces are as wide part as the considerable depth of the cells demand; and thus - happens that the membrane (both the outer and the inner ivisions) presents in profile (a', c, d) such an irregular thickness. 'he cytoblastema (a1) is colorless, hyaline, and apparently homoeneous under a low power; but, when magnified to about four undred diameters, it displays a very finely granular aspect. t occupies wide intervals between the cells, certainly more than ne-half, and fully three-fifths of the whole area of the memrane. Its apparent extent, in a general view, is even more

than that, owing to the extreme transparency of the cells, s their consequent inconspicuousness. That the cytoblase notwithstanding its low undeveloped state, is the true contrat element in this membrane, there can scarcely be a doubt, w we consider both its wide spread preponderance, and its tive continuity, as contrasted with the scattered, disconne condition of the cells (62) which are imbedded in it. Sometics it is barely possible to discover even the trace of a cell on th border of an afferent ostiole (os), and in that case we must inte inevitably, that it is cytoblastema which opens and closes aperture. We find it, too, embracing the extreme tips of th larger spiculæ, where the cells utterly fail to appear.

The cell-element (b) of this membrane is also in a lowly ex dition; only partially developed. There is no cell-wall. Wh may appear to be a wall is really the thin stratum of cytobla tema (a) overlying the distal and proximal faces of the ce This is our conclusion after the most critical scrutiny, with: carefully-corrected objective. Were it not, indeed, for the usually constant presence of a distinct nucleus (n) in each ce we would be strongly inclined to look upon it as merely a dense collection of coarser granules than are generally diffused through the cytoblastemic layer. The irregular and jagged outline, and the caudate projections of the cells (b2) also tend to tempt one to the latter view. The cell element in this case, then, corres ponds only to what is usually considered the cell contents, and a nucleus. The contents are composed of coarse and fine grer granules, which at times are quite conspicuous, but most fre quently are so transparent and slightly refractive as to appear, collectively, unless specially focussed upon, as a faint blotch in the investing membrane. This renders it all the more difficul: to trace the outline of the cell, and particularly where it throws out irregular, caudate prolongations, to blend with those of other cells. We have been able to detect but one layer of cells in this membrane* when it is well stretched out. The depth of the cells, as may be seen in a sectional profile view (b), is about equal to their breadth, and their length is from one-half more than to twice their breadth; but frequently they are as broad as long. They stand in no particular relation to the ostioles; and, as stated above, sometimes scarcely touch their border. The nucleus (n) may be readily detected by its peculiar. strong refraction, and its considerable superiority in size over the granules. Its bright refractiveness in this connection reminded us of a contractile vesicle, but, although suspecting it of such a function, we could detect no change other than might be produced by the varying length and breadth of the cell, and the *Carter figures two or three cells overlying each other in Spongilla alba. Mag. Nat. Hist., July, 1857, Pl. 1, fig. 7.

Ann.