entertained any other opinion in regard to them, than that they are the only passages of communication that could have existed between the softer parts occupying the ambulacral furrows of the arms, and the interior of the body, Mr. Billings was the first author, so far as we are at this time aware, who called especial attention to them in this regard. We regret that we have not space to quote a portion, at least, of his remarks on this subject, and would advise the student to read attentively the whole of both of his articles alluded to.

The specimens at Mr. Billings's command enabled him to trace the courses of the ambulacral canals from the arms, through the walls of the body at the arm-bases, and to ascertain the additional fact that, after passing through the walls, they seemed to have turned upward; but beyond this he had not the means of tracing them farther.

A single specimen of Actinocrinus proboscidialis, however, in Mr. Wachsmuth's collection, is in a condition (thanks to the great skill of that gentlemen and the exceedingly fortunate state of preservation, by which its delicate internal parts remain almost entire, and without any surrounding matrix) to throw much additional light on this subject. By very dextrous manipulation, Mr. Wachsmuth succeeded in removing about half of its vault, so as to expose the internal parts, in place, and in an excellent state of preservation. The convoluted organ already described in other species is in this comparatively large, subcylindrical in the middle, apparently tapering at the lower end, and a little dilated at the upper extremity. It seems to be rather dense, and shows the usual rough appearance, but as we had no opportunity to examine any detached fragments of it by transmitted light, we did not determine whether or not it has pores passing through it, though it probably has, at least when entirely free from any inorganic incrustation. Its slightly dilated upper end seems to stand with its middle almost, but apparently not exactly, under the middle of the nearly central proboscis of the vault; while at the anterior side of its upper margin, and a little out from under the proboscis, it shows remains of a kind of thickened collar, which we found to be composed of minute calcareous pieces. From this there radiate five ambulacra, composed of the same kind of minute pieces as the collar itself, each ambulacrum consisting of two rows of these minute pieces alternately arranged. They are each also provided with a distinct furrow along their entire length above. As they radiate and descend from their connection with the top of the convoluted frame work of the digestive sack, they all bifurcate, so as to send a branch to each arm-opening, those passing to the posterior rays curving

a little at first above, so as not to pass directly under the proboscis. These ambulacra, although passing along obscure furrows in the under side of the vault, which are deepest near the arm-openings, are not in contact with the vault, or visibly connected with any other parts than the top of the convoluted digestive sack, and the outer walls at the arm-openings. Each of their subdivisions can be traced into an arm-opening, and it is very probable that they continued on out the ambulacral furrows of the arms and tentacula. At one point in one of these ambulacral canals, beneath the vault, some evidences of the remains of two rows of minute pieces were observed alternating with the upper edges of those composing the under side of these canals, and thus apparently covering them over. condition of the parts is such however, as scarcely to warrant the assertion that this was really the case, though we are much inclined to think it was. If so, these canals must have been, at least under the vault, hollow tubes, formed of two rows of pieces below, and two above, all alternately arranged.

The

We are not aware that any evidences of the existence of these delicate ambulacral canals, composed of minute calcareous pieces, and passing beneath the vault from the arm-openings to the summit of the convoluted digestive sack, have ever before been observed in any Crinoid, recent or extinct; and we can but think it probable, that the extremely rare combination of circumstances that brought them to light in this instance may not again occur for centuries to come, with regard to another specimen. That they correspond to the ambulacral canals seen extending from the arm-base to the mouth, on the outside the ventral disc in Comatula, is clearly evident.

The presence of furrows radiating from the central region of the under side of the vault to the arm-openings, in various types of paleozoic Crinoids, must have been frequently observed by all who have had an opportunity to examine the inner surface of this part. Messrs. de Koninck and Lehon figure a portion of the vault of Actinocrinus stellaris, in their valuable Recherches sur les Crinoides du Terr. Carb. de la Belgique, pl. iii, fig. 4 f., showing these furrows, which they seem to have regarded as the impressions left by the muscles of the viscera. The inner surface of the vault of most of our western Carboniferous Crinoids is known to have these furrows more or less defined, either from specimens showing this inner surface, or from natural casts of the same. In some instances they are very strongly defined from the central region outward to the arm-bases, to each of which they send a branch. In Actinocrinus ornatus Hall, for instance, they are generally so strongly defined as to raise the thin vault into strong radiating ridges,

separated by deep furrows on the outer side. In Strotocrinus, the vault of which is greatly expanded laterally, and often flat on top, these internal furrows, in radiating outward, soon become separated by partitions, and as they go on bifurcating, to send a branch to each arm, they actually assume the character of rounded tubular canals, some distance before they reach the arm-bases.

That these furrows or passages of the inner side of the vault were actually occupied during the life of the animal by the ambulacral canals as they radiate from the top of the convoluted digestive sack to the arm-openings, we think no one will for a moment question, after examining Mr. Wachsmuth's specimen of Actinocrinus proboscidialis, which we have described, showing all these parts in place. It is also worthy of note, that in all the specimens of various types in which these furrows of the under side of the vault are well known, whether from detached vaults, or from casts of the interior of the same, they never converge directly to the opening of the vault, but to a point on the anterior side of it, whether there is a simple opening or a produced proboscis. The point to which they converge, even in types with a decidedly lateral opening of the vault, is always central or very nearly so, and even when the ing is nearly or quite central, the furrows seem to go, as it were, out of their way to avoid it, those from the posterior rays passing around on each side of it to the point of convergence of the others, a little in advance of the opening. That the ambulacral canals here, under this point of convergence of the furrows in the under side of the vault, always came together and connected with the upper end of the convoluted frame-work of the digestive sack, we can scarcely entertain a doubt.

open

Now in looking at one of these specimens, especially an internal cast of the vault, showing the furrows (or casts of them) starting from a central, or nearly central point, and radiating and bifurcating so as to send a branch to each arm-base, while the opening or proboscis of the vault (or the protuberance representing it in the cast) is seen to occupy a position somewhere on a line between this central point from which the furrows radiate, and the posterior side, one can scarcely avoid being struck with the fact, that this point of convergence of the ambulacra, under the vault, bears the same relations in position to the opening of the vault, that the mouth of a Comatula does to its anal opening. And when we remember that eminent authorities, who have dissected specimens of the existing genus Comatula, maintain that these animals subsisted on microscopic organisms floating in the sea-water, such as the

AM JOUR. SCI-SECOND SERIES, VOL. XLVIII, No. 142.—JULY, 1869.

Diatomaceæ, minute Entomostraca, etc., which were conveyed to the mouth along the ambulacral canals, perhaps by means of cilia, we are led from analogy to think that paleozoic Crinoids subsisted upon similar food, conveyed in the same way to the entrance of the digestive sack. If so, where would there have been any absolute necessity for a mouth or rather opening directly through the vault, when, as we know, the ambulacral canals were so highly developed under it from the arm openings to the entrance into the top of the alimentary canal? Indeed it seems at least probable, that if the soft ventral disc of Comatula had possessed the power of secreting solid vault pieces, as in most types of paleozoic Crinoids, that these vault pieces would not only have covered over the ambulacral furrows, as in the paleozoic types, but that they would also have hermetically covered over the mouth, and converted the little flexible anal tube into a solid calcareous pipe, such as that we often call the proboscis in the extinct Crinoids.

From all the facts therefore now known on this point, we are led to make the inquiry whether or not, in all the paleozoic Crinoids in which there is but a single opening in the vaultwhether it is a simple aperture or prolonged into a proboscis, and placed posterally, subcentrally, or at some point on a line between the middle and the posterior side-this opening was not, instead of being the mouth, or both mouth and anus as supposed by some, really the anal aperture alone; and whether in these types the mouth was not generally, if not always, hermetically closed by immoveable vault pieces, so far as regards any direct opening through the vault?

We are aware of the fact, that at least one apparently strong objection may be urged against this suggestion, and in favor of the conclusion that the single opening seen in these older Crinoids was the mouth, or at least performed the double office of both anal and oral aperture. That is, the frequent occurrence of specimens of these paleozoic species, with the shell of a

* Bronn mentions the fact (Klassen des Thierreichs. Actinozoa, II, p. 211), that the remains of Diatomacea, of the genera Navicula, Actinocyclus, Coscinodiscus, and of Entomostraca, were found in the stomach of Comatula, and suggests that, when such objects, in floating in the sea-water, came in contact with the ambulacral furrows of the pinnulæ, they were conveyed along these furrows to those of the arms, and thence in the same way into the mouth. He ridicules the idea, sometimes suggested, that the food may have been handed by the pinnulæ or arms directly to the mouth.

Dujardin and Hupé also state (Hist. Nat. de Zoophytes Echind., p. 18), that the living Comatula was "nourished by microscopic Algae and floating corpuscles, which the vibratile cilia of the ambulacra brought to the mouth." That they may have sometimes swallowed a larger object, that accidentally floated into the mouth. however, is not improbable, and would not, if such was the case, by any means disprove the generally accepted opinion that these animals received their food almost entirely through the agency of their ambulacral canals.

1

Platyceras in close contact by its aperture, either with the side or the vault of the Crinoid, and not unfrequently actually covering the only opening in the vault of the latter, so as to have led to the opinion that the Crinoid was in the very act of devouring the Mollusk at the moment when it perished.

Amongst the numerous beautiful specimens of Crinoids. found in the Keokuk division of the Lower Carboniferous series at Crawfordsville, Indiana, there is one species of Platycrinus (P. hemisphæricus), that is so abundant that probably not less than two hundred, and possibly more, individual specimens of it have been found there by the different collectors who have visited that noted locality; and, judging from those we have seen, apparently about one-half of these were found with a moderate sized, nearly straight, or very slightly arched and conical Platyceras (P. infundibulum), attached to one side by its aperture, between the arms of the crinoid, and often so as to cover the single lateral opening in the vault of the same.* From the direction of the slight curve of the apex of the Platyceras, it is also evident that it it is always placed in such a manner, with relation to the Crinoid, that the anterior side of the mollusk was directed upward, when the vault of the Crinoid was turned in that direction.† A species of Goniasteroidocrinus (G. tuberosus Lyon & Casseday), found at the same locality, also has frequently a Platyceras attached to the top of its nearly flat vault, so as to cover the only opening in the same. It is worthy of note, however, that it is always another, subspiral, Platyceras (very similar to P. æquilaterum), that we find attached to this Crinoid, so that here at least, it would

* We at one time thought these shells attached to the side of this Platycrinus, to be out of reach of the opening, or supposed mouth, because we had not seen, specimens showing the position of the opening in this species, and had supposed from its similarity to Platycrinus granulatus Miller, and other species without a lateral opening, that such was also the case with this. We have since seen specimens, however, showing that it has a lateral opening, and therefore belongs to the group Pleurocrinus, so that it is probable these shells often cover this opening. + Prof. Richard Owen has noticed, in his Report on the Geological Survey of Indiana, p. 364 (1862), the frequent occurrence of a Platyceras attached to this same Platycrinus, at this locality, and proposed to name the Platyceras P. pabulocrinus, from the supposition that it formed the chief food of these Crinoids. It is probable that the Platyceras for which he proposed this name, is the same we named P. infundibulum, but as he gave no description of the species, and but an imperfect figure, we cannot speak positively as to its identity. Prof. Hall has also proposed the name P. subrectum for this Crawfordsville Platyceras, but he had previously used the same name for a very different, New York, Devonian species of this genus.

Prof. Yandell and Dr. Shumard have also figured in their paper entitled "Contributions to the Geology of Kentucky," a specimen of Acrocrinus, with a very similar Platyceras apparently attached to its vault.

Amongst all the Crinoids found at Burlington, Iowa, we are aware of but a single instance of one being found with a Platyceras attached, and that is a specimen of Actinocrinus ventricosus in Mr. Wachsmuth's collection, which has a crushed shell of a Platyceras connected with its vault.