is considerably to the north of the equator, and in coincidence of action when the sun gets to the south of the equator. (See Fig. 7.)

Let us take now the special case represented in Fig. 7. The same effects will occur that we have just considered; they will only be less in amount. But we may also discern effects produced by the primary ecliptic current. Before and after the equinox this current crosses the meridian under a large angle, early in the morning, and by the rapid shifting of its position tends to deflect the needle more and more toward the west during the forenoon. During the afternoon the tendency is reversed. It is thus opposed to the intertropical radial current during the greater part of the day both before and after the equinox. At the earlier date, at which the sun is north of the equator, the primary current crosses the meridian early in the morning in a direction from N. of E. to S. of W.; and at the later date it runs at the same hour in a direction from S. of E. to N. of W. The needle ought therefore to have a greater west declination early in the morning after than before the equinox. That such is the fact will be seen on inspecting Fig. 7.

It is to be observed that the easterly movement of the needle from 6 to 7 A. M., when the sun is north of the equator, and westerly movement at the same hour when the sun is south of the equator, is attributed to the action of the radial currents. When the sun has a south declination the westerly tendency resulting from the action of the radial currents, from 6 to 7 or 8 A. M., should be relatively greater at St. Helena than at Toronto. cordingly the radial may prevail over the ecliptic currents at that hour, at St. Helena, and the ecliptic over the radial at Toronto; (compare Figs. 6 and 7.)

It may be stated in general terms that the peculiar phenomena of the diurnal variation of the declination which have place at St. Helena and the Cape of Good Hope, and probably at intertropical stations generally, receive their explanation in the alternate preponderance of the extratropical currents of the Northern or Southern Hemisphere, according as the sun is north or south of the equator; together with the modifying action of the ecliptic and radial currents that are developed between the tropics.

(To be continued.)

ART. XX.-Observations on the fructification of the Arachis hypogaa; by HUGH M. NEISLER, Columbus, Geo.

IN studying our Stylosanthes a few years ago, my attention was attracted by a note in Torrey and Gray's Flora of North America, vol. i, p. 354. viz., "Mr. Bentham in a paper on the affinities of Arachis, read before the Linnæan Society in 1838, gives an account of the two kinds of flowers in Stylosanthes, and shows its affinity to Arachis, which he considers a genuine Hedysarea." I presumed that he supposed the Arachis to have two kinds of flowers, but, wishing to inform myself accurately as to his views, I mentioned the subject to Dr. Torrey in the course of our correspondence, who remarked in reply; "Mr. Bentham says, that Arachis has two kinds of flowers. Those that have all the parts, do not perfect their fruit, the ovary never ripens. The fructiferous flowers, have neither calyx, corolla, nor stamens, but consist at first of a minute ovary on a rigid stipe that arises from between two bracteoles. After fecundation, the minute ovary swells, at the same time burrows in the ground where it ripens.

On examination, I found in some specimens that had been in flower some days, in the axils of two or three of the lower leaves, minute, sessile (sometimes two or three in a kind of one-sided raceme) conical germs situated between two bracteoles; these gradually elongated themselves until reaching the earth, they penetrated beyond the reach of light, where their extremities becoming etiolated they grew succulent, enlarged and ripened their fruit. The stipe of the fruit varies much in length, in the prostrate forms of the plant from 1 to 3 or 4 inches-but in an upright variety which I cultivate, they grow 6, 12, and sometimes, even 18 inches before reaching the earth and in their growth hang around the stem like aerial rootlets. In the axils next above these fertile germs, in my specimens, I found petal-bearing flowers, which I at first (supposing Mr. Bentham's views of course to be correct) regarded as barren. But after close and repeated examinations, to my surprise, I found them in all respects perfect, and what at first sight I had thought a long peduncle which withered with the flower, proved to be a slender, tubular calyx, through which there was no difficulty in tracing the style to a minute conical germ, situated between two bracteoles-and in all respects identical with those in the axils below. And after examining a few plants, I succeeded in finding germs elongated to two or three inches with the marcescent calyx and corolla still adhering to their points, and stimulated into growth beyond a doubt by the perfect and fertilized ova. Younger plants just getting into bloom showed petal-bearing flowers in the lowest axils

and doubtless those that I first examined and which I thought achlamydeous, would have been found so, if seen a little earlier; for generally, the flower falls away entirely and is seldom found attached to the germ after withering. The flowers of the Arachis hypogaa are all petal-bearing and all fertile. The plant is in some respects a singular one-and I am not surprised that Mr. Bentham or any one else who had not watched it in all stages of its growth, should have fallen into error as regards its fructification.

ART. XXI.-On certain Physical Properties of the Light of the Electric Spark, within certain Gases as seen through a Prism; by D. ALTER, M.D., of Freeport, Pa.

In a former communication, I noticed the peculiar character of the light produced by interrupting the galvanic circuit, between different kinds of metal. I also mentioned that several bright bands were observed in the common electric spark. I have since employed various metallic conductors, but without producing the bands corresponding with those resulting from the interrupted galvanic circuit between the same metals-the bands always remaining the same, whatever metal was employed. It then occured to me, than the bands were characteristic, not of the metals, but of the atmosphere, through which the sparks passed. To test this view, I passed the spark through various gases, in succession, and found that they were characterised by their appropriate bands, with as much distinctness, as the metals are, by the galvanic spark; e. g., I discovered in atmospheric air, one red, one orange, two green, one blue, and one indigo: while in hydrogen, I discovered one very bright red, two faint green bands, and one very faint blue. In nitrogen, one red, one orange, and two green. In chlorine, two distinct green bands, and a knot of light in the yellow and also in the blue, which I think are composed of several bands close together in each. In carbonic acid, there are, one red, three orange and two green bands. In sulphuretted hydrogen, there are red, orange and green bands. In oxygen, no bands were discovered, but the light was strong throughout the spectrum. In the other gases it is feeble, except in the bands before mentioned. The quantity of light in the red band of the spark, in hydrogen, is quite remarkable, being so great, that the spark, seen without the prism, has a very red appearance, as also in the gases compounded of hydrogen. From this, we perceive the cause of the difference in color, in the flashes of lightning-for when the electricity has a watery conductor, in much of its course, it will emit red light, but when it passes through air, the light will be white: as in the spark through that medium,

the bands are well distributed among the colors of the spectrum. The colors also, observed in the aurora borealis, probably indicate the elements involved in that phenomenon. The prism may also detect the elements in shooting stars, or luminous meteors. Since from the preceding observations, it is evident that the electric light, whether from the interrupted galvanic current, or from the common electric machine, is principally resolved into several bright bands by the prism; and that the light, thus produced by one elementary body, differs in the number, brilliancy and situation of its bands, from every other element, so that it is immediately recognised, by mere inspection-we are led to the following inquiries. Is their such a fluid as electricity? or, are the phenomena, commonly reputed electrical, the result (as suggested by Prof. Graham) of chemical affinity? If so, are there only two poles, a chlorous and a zincous, to each molecule-or, are there as many poles or combining surfaces as are indicated by the number of bright bands of its refracted light? And (if the undulatory theory of light may be depended on) would not these bands give an indication of the size of those surfaces or poles?

On Daguerreotyping the dark lines in the Solar Spectrum. -Being desirous to know whether corresponding lines exist in the actinic rays, I adopted the following method. The sun's rays were admitted iuto a dark chamber, between the edges of two pieces of sheet brass about eight inches in length and separated about the thirtieth of an inch, at one end, but in contact at the other.

Near the outside of the aperture thus formed, was placed a large lens, five feet in focus. Near the focus of the lens in the chamber, the rays pass through a prism and through a second lens of about 20 inch focus, which shows the dark lines very distinctly on white paper, at its focus, for rays coming from the slit. The prepared Daguerreotype plate, placed in the focus and exposed for one or two seconds, produces the effect.

In the Daguerreotype, which I send you there are two spectra caused by filing the brass slips so as to cause an aperture on side of the point of contact. I have placed the letters on the lines as given in Brewster's Optics, 1837, page 79. They would correspond with Prof. Draper's (see this Journal, March, 1848) if the H occupied the place of I.

I could not see the spectrum farther than the breadth of the second broad line at I in the direction beyond that line, when looking through the prism and slit at the sun. But by receiving the spectrum on paper stained with alcoholic tincture of turmeric, several dark lines can be seen beyond these and the blue appears to be changed to violet down to the line F.

ART. XXII-Synopsis of the Ichthyological Fauna of the Pacific slope of North America, chiefly from the collections made by the U. S. Expl. Exped. under the command of Capt. C. Wilkes, with recent Additions and Comparisons with Eastern types; by L. AGASSIZ.

(Continued from p. 99.)

EXOGLOSSUM, Raf.

THUS far a single species of this remarkable genus is known, which was first described by Lesueur, under the name of Cyprinus maxillingua in the first volume of the Journal of the Academy of Natural Science of Philadelphia, p. 185. Lesueur however already suspected that this species would constitute a separate genus, but until the discovery of another similar species he would content himself with referring it to the genus Cyprinus. His expectation of such a discovery has however not been realized, since the three species soon afterwards referred to this type by Rafinesque, who first introduced for it the name of Exoglossum, and those described at a later period by Kirtland and Valenciennes do not in reality belong to it. This is another among the many instances which show that the importance of generic peculiarities does not necessarily depend upon the number of species in which they occur. Rafinesque states that he had thought of calling this genus Glossognathus, but that this name appearing to him rather harsh, he has proposed that of Exoglossum, for the sake of euphony. Valenciennes remarks that he would have preferred that of Glossognathus, which he had himself introduced for this genus, before he read Rafinesque's paper. As matters now stand, we can have no choice, the name of Exoglossum standing by the right of priority and general acceptance. DeKay is certainly wrong in referring this fish to the genus Catostomus with which it has no generic affinity, as I have already shown. In calling the typical species Exoglossum Lesueurianum, Rafinesque has paid a deserved tribute to the able French naturalist who discovered this fish; but in so doing, he has acted contrary to the universally acknowledged law of priority, which requires that specific names once established, should never be changed, unless they are absolutely objectionable, which is by no means the case in this instance. I do therefore not hesitate in restoring the specific name of maxillingua, first given to this fish by Lesueur, who discovered it in Pipe Creek, Maryland.

Valenciennes describes specimens from Pennsylvania. I have myself obtained numerous specimens from different localities; * I have introduced this and similar other remarks in my paper, not merely with reference to the subjects under consideration, but chiefly as hints to American Zoologists, who in their writings seem not always to take sufficiently into consideration the traditional rules which have guided Naturalists since the days of Linnæus, in matters of nomenclature.