From the square form of the prism Mr. Dana has suggested that Algerite may be a result of the alteration of scapolite from which species however it differs widely in composition; while the latter affords from 13 to 24 p. c. of lime, and from 4 to 7 of soda, Algerite contains neither lime nor soda, but 10 p. c. of potash, so that the change must have consisted in the removal of lime and soda, and the substitution of magnesia, water, and a large amount of potash, which reactions, especially the replacement of one alkali by another, seem very questionable. As scapolite has not yet been found with this mineral, it would have been better to regard algerite as derived from the alteration of the associated idocrase, a process which presents less difficulty to the chemist than the other. It may fairly be questioned whether the extreme views maintained by some, with regard to the alteration of mineral species, do not tend to discourage mineralogical chemistry, and to embarrass the science with groundless hypotheses.

In support of the view that Algerite is an altered scapolite, Mr. J. D. Whitney has given at page 296 of this volume, the results of his examination of the mineral. He observes of the material selected for analysis, that "after the ignition it was noticed that portions of the ignited mineral remained nearly unaltered in appearance, while the larger part acquired a brick-red color, and on examination with the microscope was seen to contain small silvery scales apparently of mica. As only a small quantity could be used for analysis, the results can be relied upon only as approximatively correct." p. 208. He gives the following as the result of his analysis: Silica 52-09, alumina 18-63, phosphate of lime 8-22, carbonate of lime 4:41, water 6 68, loss, potash and soda? 9.97100 00.

Had the analysis of such a compound agreed with my own, I might have had reason to doubt the results of Mr. Crossley and myself. Since reading Mr. Whitney's remarks I have examined several crystals of Algerite for phosphate of lime, by digesting their powder with heated hydrochloric acid, and testing the solution with molybdate of ammonia; this delicate reagent indicated the presence of traces only of phosphoric acid. The apparently pure calcite in which the crystals are imbedded, gives a more decided reaction, and traces of this acid are seldom wanting even in feldspars and tourmalines. While, therefore, we do not question that Mr. Whitney's parti-colored mixture may have contained apatite as well as mica, we find in his confessedly approximate results, no reason for doubting that Algerite is a homogeneous mineral, or for supposing it to be produced by the alteration of scapolite. Further examinations of the mineral may throw more light upon the variations in the proportions of water and magnesia.

Montreal, C. E., March 7th, 1854.

ART. XXXVIII.-Notice of a collection of Fishes from the southern bend of the Tennessee River, in the State of Alabama; by L. AGASSIZ.

(Concluded from page 308.)

CYPRINODONTS, Agass.—Only two species of this family have thus far been discovered in the waters of the Tennessee River, and both of them have already been described by Dr. Storer under the names of Poecilia catenata and olivacea, Synopsis, p. 178. Having made lately however, a thorough revision of the genera and species of this family found in the United States, I would remark that Poecilia catenata, St., ought to be referred to the genus Hydrargyra, and that Paecilia olivacea belongs to my newly established genus Zygonectes. These species ought therefore to stand in future in our systematic catalogues under the names of Hydrargyra catenata, and Zygonectes olivaceus.*

CYPRINOIDS, Cuv.-This is one of the most interesting families of our freshwater fishes, both on account of the number of genera and species inhabiting our lakes and rivers, and of the diversity of their forms and habits.

CARPIODES, Rafin.-In the great French Ichthyology, Valenciennes has established a new genus under the name of Sclerognathus, for Lesueur's Catostomus cyprinus, and this genus has deservedly been acknowledged by subsequent writers. In con

*The species of the genus Zygonectes may be arranged in two groups: 1, those in which there are several more or less distinctly dotted lines along the sides of the body, and in which a broad black band extends across the eye and cheek. To this group belong: Z. Nottii, Agass. The darker continuous longitudinal lines alternate with fainter interrupted ones. Males with distinct, transverse bands. Dark olive above, fading upon the sides, silvery below. Operculum, throat, and space in advance of the eye light orange color. Mobile, Alabaina. Collected there with Dr. Nott. Mississippi: Col. Deas.-Z. lineolatus, Agass. Longitudinal lines broader and undulated or serrated, the transverse bands of the male very distinct and broader than the longitudinal ones. Olive colored, darker along the back and fading upon the sides, lower parts silvery Discovered by Dr. W. I Burnett at Augusta, Ga.-Z. guttatus, Agass. A large dark spot upon the centre of each scale on the back and sides, forming longitudinal rows of disconnected dots. The transverse bars of the males are much narrower and nearer together than in Z. lineolatus. Dark olive above, fading upon the sides; abdomen silvery. Mobile, Alabama.-Z. dispar, Agass. Longitudinal lines of minute dots particularly distinct in the anterior part of the body, alternating backwards with continuous lines in the males, which are besides transversely barred, whilst the female has only continuous serrated lines upon the sides. Light olive above, silvery upon the sides and below. In small creeks near St. Louis, Mo., on the Illinois side of the Mississippi, and also in the Illinois River at Beardstown.-Z. hieroglyphicus, Agass. Anterior and upper part of the body irregularly sprinkled with dark spots, passing into longitudinal rows backwards. Light olive above, silvery upon the sides and below. Mobile, Alabama. 2. The second group includes species with one broad longitudinal black band extending from the tip of the lower jaw to the base of the tail, passing in a straight line through the eyes and along the sides of the body. To this group belongs the species mentioned above from the Tennessee River, and also Z. lateralis, Agass, which is a more elongated species from Mobile, Alabama; also dotted above the broad lateral band, and Z. zonatus, Agass., from St. Louis, Mo., which has no spots upon the sides of the back, and in which the outlines of the longitudinal band are serrated.

sidering this type of Cyprinoids as a distinct group among the Catostomi, Rafinesque has however the priority over the able professor of the Jardin des Plantes; for we find in his Ichthyologia Ohiensis that the third subgenus of Catostomus, which he calls Carpiodes, though not characterized with the precision with which Valenciennes has circumscribed his genus Sclerognathus, exactly corresponds to it. I do not hesitate therefore to adopt Rafinesque's name as the older; the more so, since this writer has at the same time wisely separated from the common Catostomi at that early day two other types of the same group, which are even now left among Catostomi by all ichthyologists. I allude to the subgenus Ictiobus, with Calostomus Bubalus as its type, and to the genus Cycleptus for the Missouri sucker; for though Rafinesque did not himself examine this latter fish, and ascribes to it two dorsals, it must be evident to any one who has had an opportunity of investigating this rare species that the few words with which it is mentioned apply to it, and that the indication of two dorsals is easily explained by the very form of that fin, the anterior part of which rises like a separate fin in advance of the following low part which extends uniformly far behind. I should add that Catostomus elongatus belongs also to this genus Cycleptus. As to Ictiobus, it resembles Carpiodes in external appearance, but is at once distinguished by its thin lips and more terminal mouth.* Nothing is to be more regretted for the progress of Natural History in this country, than that Rafinesque did not put up somewhere a collection of all the genera and species he has established, with well authenticated labels, or that his cotemporaries did not follow in his steps, or at least preserve the tradition of his doings, instead of decrying him and appealing to foreign authority against him. Tracing his course as a naturalist during his residence in this country, it is plain that he alarmed those with whom he had intercourse by his innovations and that they preferred to lean upon the authority of the great naturalists of the age then residing in Europe, who however knew little of the special Natural History of this country, than to trust the somewhat hasty man who was living among them, and who had collected a vast amount of information from all parts of the States, upon a variety of objects then entirely new to science. From what I can learn of Rafinesque, and from a careful study

* In connection with the genera mentioned above, I may remark here that Rafinesque has established another sub-genus under the name of Morostoma, which fully deserves to be recognized as a distinct genus, as far as I am able to judge from the three species belonging to it, with which I am especially acquainted, which are Catostomus anisurus of the West, C. gibbosus or tuberculatus of the East, and C. Suceti of the South. After acknowledging these alterations of the genus Catostomus, as it is now generally understood by ichthyologists, there would still remain a group of species to constitute the genus Catostomus proper of which C. hudsonius, for which the name Catostomus was first proposed, may be considered as the type. Thus freed of all unjustifiable additions engrafted upon it in course of time, the genus Catostomus would be restored to its primitive natural circumscription.

of his works, I am satisfied that he was a better man than he appeared. His misfortune was his prurient desire for novelties, and his rashness in publishing them, and yet both in Europe and in America he has anticipated most of his cotemporaries in the discovery of new genera and species in those departments of science which he has cultivated most perseveringly, and it is but justice to restore them to him, whenever it can be done. Personal considerations should no longer be allowed to interfere with this late act of redress. May the example of Rafinesque not be lost for those naturalists in this country who describe new species without taking the least care to preserve the original specimens of their descriptions, or to circulate authentic ones among other naturalists.

Besides the well known type of the genus Carpiodes, the C. Cyprinus, and the other species described by Valenciennes and Rafinesque, I have ascertained the existence of five undescribed species, of which I give below short comparative descriptions. These species bear to one another similar relations as the species of Cyprinus described by Heckel; indeed they truly represent upon the Continent of North America the genus Cyprinus of the old world to which they bear the greatest resemblance in outward appearance, though they differ strikingly in their generic characters. I have applied to the new species here mentioned names reminding us of the common name of Buffalo applied to all of them throughout the country. The large number of specimens including all sizes, which I have been able to collect of some species of this genus, has enabled me to ascertain the range of variation in their characters.

1. Carpiodes Urus, Agass.-From the Tennessee River. It grows very large, weighing occasionally from 30 to 40 pounds. The body in this species is not so high as in C. Cyprinus, nor is it so compressed above; the scales are also not so high, but more. angular behind, and the anterior portion of the dorsal is not so elongated. The gill cover is larger, and the distance from the hind border of the eye to the inferior angle of the subopercle, near the base of the pectorals, and the distance from the same point to the superior and posterior angle of the opercle, are nearly equal. In C. Cyprinus the distances differ by nearly one-third. The subopercle is not triangular, but its hind border is nearly regularly arched from the upper angle to the posterior angle of the interopercle. The anal has its posterior margin full, and not lunate; the caudal is not so deeply furcate as in C. Cyprinus. The ventrals do not reach the anal. All fins are of a dark color. I am indebted to Dr. Newman for this species.

2. Carpiodes Taurus, Agass.-From Mobile River, Alabama. The form of the body is intermediate between that of C. Cyprinus and C. Urus. The gill cover has the same form as in

C. Urus, but it is larger and more strongly arched behind. The hind margin of the scales is waving, owing to a somewhat prominent middle angle. The anterior rays of the dorsal equal in length two-thirds of that of the base of the fin. Anal not lunate behind. The ventrals do not reach to the anal opening. Caudal not so deeply furcate as in C. Cyprinus.

3. Carpiodes Bison, Agass.-From the Osage River, Missouri. This species is more elongated than C. Taurus. The head is smaller, the opercle also smaller, and the subopercle triangular. The dorsal has its anterior rays longer, hence its hinder border is more deeply emarginate. Anal more deeply lunate. Horizontal diameter of scales greater. I have received this species through the attention of Mr. George Stolley.

ana.

4. Carpiodes Vitulus, Agass.-From the Wabash River, IndiThis seems to be a smaller species than the preceding ones. The form of the body resembles that of C. Taurus; but the eyes are smaller; the opercle is more broadly rounded behind; the subopercle has its posterior and free border regularly arched above and below, and not emarginate as in C. Taurus. The direction of the numerous water tubes on the head and cheeks also differ. The upper and lower borders of the scales are nearly straight. The dorsal does not extend quite so far forwards. I am indebted to Col. Richard Owen of New Harmony for this species.

5. Carpiodes Vacca, Agass.-From the Susquehannah River. This species resembles more closely C. Cyprinus than any other; the anterior rays of the dorsal are also very elongated, yet they do not reach beyond the base of the fin itself when bent backwards; the caudal is not so deeply furcate, and the scales have a greater horizontal diameter. I owe this species to the kindness of Professor S. S. Haldeman.

CATOSTOMUS, Lesueur.-The following species of this genus have been collected by Dr. Newman in the vicinity of Huntsville: Calostomus communis, Lesueur.-Called Fine-scaled Sucker at Huntsville.

Catostomus nigricans, Lesueur.-Called Hog Sucker at Hunts

ville.

Catostomus Duquesnii, Lesueur.-Called May Sucker at Huntsville.

Catostomus melanops, Kirtl.-Also called May Sucker at Huntsville. This species agrees with Kirtland's description of C. melanops, except in having longer pectorals and in the reddish color along the sides. Rafinesque's description cannot apply to Having no specimens from the localities mentioned by Rafinesque and Kirtland, I do not venture to pursue further a comparison between these fishes.

RHINICHTHYS, Agass.-This genus was established by me in "Lake Superior," page 353. Several new species have been dis