band, which was reproduced by Kämtz in 1840, and by Schübler* in his four splendid engraved plates in 1849. In fine, the plates which were published by the Smithsonian Institution at Washington, and in 1859 by the chart department of the French minister of marine, ‡ to serve as instructions for seamen and observers, were reproductions of those of Kämtz. The French edition alone contained two plates. of which the first embraced the simple forms of clouds, cirrus, stratus, cumulus, and nimbus, and the second the compound forms, cirro-cumulus, cirro-stratus, and cumulo-stratus, with four variable cirri. Howard's Plate VI, on the contrary, published in 1803 for the first time, in Tilloch's Philosophical Magazine, represents the stratus as a mist spreading above a lake surrounded by hills. Thus far in relation to the erroneous interpretation given to Howard's stratus. We now proceed to note another error hitherto propagated, relative to the nimbus. In the first place, Kämtz's definition, which all the late meteorologists have adopted, is as follows: When the cumulus is piled up and becomes more dense it passes to the state cumulo-stratus, which often takes at the horizon a black or bluish tint and passes to the state of nimbus, or rain-cloud. This is distinguished by its uniform gray tint and its broken edges, the clouds which compose it being so much confounded that it is impossible to distinguish them. Thus the single primordial and distinctive character which results from this definition is that of a rain-cloud, and therefore we shall call every rain-cloud nimbus, as has been done to the present time. Its secondary characteristics are: 1st, a tint of uniform gray; 2d, its edges broken; 3d, confusion of all the clouds of which it is composed. All this gives no idea of the most essential element, namely, the real form of the rain-cloud. Though the definition of Kämtz, and other writers, is not that of Howard, this observer has not given one more intelligible and exact. We see that he unconsciously felt the formation of rain-cloud, but was not very settled in his description. I am speaking of the double superposed stratum, the inferior formed of cumulus, which it designates, and the second superior stratum of cirrus, whose existence was vaguely felt by Howard. The remainder of his definition is cumulus cirro-stratus vel nimbus def. nubes vel nubium congerico (superné cirrata) pluvian effundens. A cloud or system of clouds from which rain is falling is called a rain-cloud. It is a horizontal sheet, above which the cirrus spreads, while the cumulus enters in laterally and from beneath. In other isolated passages of his long description of the nimbus, the idea of this double stratum appears yet more plainly. "Clouds, in any one of the preceding modifications, may increase so as completely to obscure the sky. Before this effect takes place there exists, at a greater altitude, a thin, light veil, or at least a hazy turbidness.

* SCHÜBLER.-Grundsätze der Meteorologie, Leipzig, 1849.

+ SMITHSONIAN INSTITUTION, Washington, D. C.

‡ VANEECHOUT.—French translation of the "Explanations and Sailing Directions of Lieutenant Maury," Paris, 1859, 4to.

When this has considerably increased we see the lower clouds spread themselves till they unite in all points and form one uniform sheet. It will rain during this state of the two strata of clouds, one passing beneath the other and each continually tending to horizontal, uniform dif fusion, (the superior stratum is often seen in this case to partake of cirrus,) although they should be separated by an interval of many hundred feet in elevation. The intermediate space on these occasions is not supposed to be at any time free from a conducting medium of different watery particles, enabling the opposite electricities to neutralize each other." Hence we see that this description of Howard's nimbus has no connection with that given by Kämtz and the late writers, of a cloud intermediate in form, which, according to the latter, is supposed to have the property of producing rain. It is not the property of any cloud to produce rain, but it is produced by the reunion by the electric action and reaction upon the aqueous vapor of two superposed strata of clouds, the superior being cirrus and the inferior cumulus. It is this latter cir cumstance that Howard has neglected to properly establish. All the plates which have been published, not excepting that of Howard, give no idea of this double stratum which constitutes the rain-cloud, that is, the nimbus of Howard, or properly my pallium, (pallio-cirrus and pallio-cumulus.) I now proceed to point out the error equally inherent in the three orders of clouds denominated cumulus, cumulo-stratus, and strato-cumulus, and to show that they can all be reduced to the second type of cumulus. I remark, in the first place, a great confusion, or it may be a great similitude, between the cumulus and the cumulo-stratus; afterward I observe in the sky other cumuli, whose characters partake of more than these two types; for example, M. Kämtz's definitions, which are generally accepted. M. Kämtz says: "The cumulus is often seen in the form of a hemisphere, reposing upon a horizontal base. Sometimes these hemispheres are piled one upon the other, forming those great clouds accumulated at the horizon, which resemble at a distance mountains of snow."

Now for the cumulo-stratus: "When the cumulus is piled up and becomes more dense, it passes to the state of cumulo-stratus, which often takes at the horizon a black or bluish tint and passes to the state of nimbus or rain-cloud." Thus, it is sufficient that the cumulus be piled up and become more dense in order to be changed into cumulo-stratus, strikingly implying great development in the horizontal base, which is not mentioned, though it is one of the most essential characters of this type of clouds. While, on the other hand, this horizontal base is expressly pointed out in Kämtz's definition of cumulus. In a word, these fundamental points are found equally in the formation of cumulus and cumulo-stratus: 1st, a horizontal base; 2d, a superior hemispherical dome; 3d, a formation by increase from below upward. The points of difference as to the cumulo-stratus rest upon the irregularity of the base of the convex summits, and the aggregation from below upward;

upon the greater density, dark coloring, and semblance of their isolated masses to a mountain chain. These differences are so accidental and insignificant that Howard and Forster could not distinguish them in their definitions, and they are forced to aver that these characters were found here and there at such points that they were easily confounded. We see there is no reason to maintain a radical separation and a double denomination for two forms of clouds which are of the same kind and whose slight modifications are not even continued. It is therefore more exact to preserve the single denomination of cumulus for this second type, which embraces at once the two characteristic forms of cloud-on the one side the accumulation of hemispheres, and on the other the horizontal base, which are always inseparable.

Let us now pass to Kämtz's description of strato-cumulus: "It is composed of dense cloudy masses, round or extended, badly defined at the edges, which appear in the afternoon, increasing toward evening, until during the night the whole sky is completely covered, disappearing the next morning some hours before sunrise and finally replaced by the true cumulus. These strato-cumuli are composed of very dense versicular vapor, like the cumuli and cumulo-strati. They differ by their dependence on different hours of the day; they have also some analogy with the strati by reason of their extension, but are distinguished by greater altitude; yet they approach the cumuli. In winter the strato-cumulus often covers the sky during entire weeks; but as the sun approaches the zenith, his rays dissolve these clouds, the vapors rise, and the cumuli are formed." We find again in Kämtz's definition of stratocumulus the same confusion as in Howard's description of cumulo-stratus, which I pointed out above. The terms "cloudy masses rounded, extended, or yet badly defined at the edges," embrace three expressions which are mutually exclusive, so that it is impossible to know the true form of the cloud, for if the cloudy masses are rounded they are not extended in Kämtz's sense, and still less are their edges badly defined. As to the physical constitution of the strato-cumulus, it is the same, according to Kämtz, as that of cumulus and cumulo-stratus, that is to say, composed of very dense versicular vapor.

In fine, "the strato-cumulus approaches the stratus by its exterior, but is separated from it by greater altitude." I have already said that we can compare the stratus-mist with no other form of cloud. Hence nothing remains but the hour of the appearance and disappearance of the strato-cumulus, (which seems to be the fundamental distinction Kämtz wished to establish,) to separate them from the cumulus and cumulostratus, added to their continuance in winter for whole weeks at a time. In a word, the strato-cumuli, with Kämtz, are the clouds of the night and of winter, predominating during the absence of the solar rays, and dissolved on the appearance of the sun. On this latter circumstance, we remark that the distinction of clouds of the night made by Kämtz and Howard appears to have no foundation; I have never been able

to perceive them. So true is this, that these two savants are nowise agreed upon this point, which has given rise to the new variety of Kämtz. With Kämtz, the cloud of night is the strato-cumulus, while with Howard it is the stratus. On the other hand, since the stratus is not a true cloud according to Howard himself, but simply a mist or hoar-frost, the distinction between the cloud of night and the cloud of day must become wholly superfluous. I will close this proof of the non-existence of strato-cumulus by reminding the reader that Kämtz himself told me before his death, without discussion, that he no longer attached any importance to his cloud of night, and authorized me to erase it from the nomenclature of Howard. In the exposition of my new classification of clouds, published in 1855, in the Annual of the Meteorological Society of France, while pointing out the identity of the cumulus and cumulo-stratus, I retained both these determinations, because the term cumulus did not designate its peculiar cloud-type as well as cumulo-stratus. But, as cumulo-stratus is derived from two species of clouds, it is preferable to keep the generic name of cumulus for the two identical clouds described by Howard, attributing to them the character given above.

DEFINITION OF CLOUDS.

Every country, according to its geographical position, topography, &c., has its own type of clouds. Here the cirrus predominates; there the cumulus. All these different appearances of clouds are everywhere intimately connected with some particular condition of climate, which powerfully influences health, agriculture, navigation, and a thousand other objects of importance to humanity. They show us at every instant the direction, the velocity, and the altitude of the superior currents which afterward determine the winds at the surface of the earth. We may regard the clouds as a weathercock in the sky, constantly indicating changes so long as a single one, however small, exists, and therefore a profound study of them, in their diverse, scientific, and social applications, becomes of high importance. For this we should consider their nature, form, quantity, direction, velocity, and azimuthal rotation. Despite the scientific interest and the practical value which is attached to this subject, the study of clouds is unhappily in its infancy. It is rarely we find clouds included in the meteorological registers, and when they are so the characteristics above mentioned are omitted. Some observers simply write "clouds; " others denote the form, or, it may be, the quantity, the direction, or perhaps all these three elements, but neglect the velocity, and especially the azimuthal rotation to which I was the first to call attention and which is not yet understood.

We now proceed to present the basis of a new classification more in harmony with the actual facts of the science, and which is the fruit of twenty years' assiduous study of clouds in the Antilles, Mexico, the United States, and Europe. From the beginning of my meteorological investigations in the tropics, where the entire phenomena of the atmo

sphere assume a character of simplicity unknown in higher latitudes, I have more and more felt the necessity of a reform of Howard's nomenclature. I was unable to understand his stratus, nimbus, cumulostratus, and strato-cumulus. It was not until I had an opportunity to consult Howard's original work that I perceived the errors into which Kämtz and other meteorologists had fallen. I had then to introduce into Howard's classification the modifications which the continued progress of meteorology requires, in order that the nomenclature may be more in harmony with our advances in this line. I acknowledge with pleasure that Howard's classification, which has existed without a rival for more than half a century, was originally based upon profound study, directed by great, acuteness of observation; unhappily, however, it is too plainly stamped by the locality where his studies were prosecuted. I refer to the gray and cloudy sky of Great Britain, whence result his strato-mist, his imperfect distinction of the two great strata cirrus and cumulus, or his nimbus, (the rain-cloud,) the difference which he has established between cumulus and cumulo-stratus, besides many other faulty details of description in relation to cirrus, cirro-stratus, and cirro-cumulus.

I shall now proceed to give the derivation of my three new clouds. When certain clouds spread out uniformly over the whole face of the heavens and assume a gray or ash color, under which state rain may occur for hours or whole days, what name do we give them? They are not Howard's nimbus, as we conceive them, and as they are generally described; they are neither stormy nor electrical; they yield only a fine and continuous rain. Under this stratum we see constantly other clouds of more or less extent, but always isolated, becoming lost in it and increasing its thickness. But just before this stratum begins to break up, and during this operation, we see these same formless fragments detach themselves and fly to other regions. This inferior stratum is not alone; for when its disruption is completed we see through it another stratum of clouds, whiter and less dense, which breaks up in its turn, and ends by disappearing in an opposite direction to that of the inferior stratum. Have we a name for this variety of cloud so common in time of rain from the inter-tropical regions to high latitudes, especially in winter during the fall of snow? Does Howard's term nimbus and his description of it answer for its designation? Certainly not. We apply the name nimbus to the single storm-cloud, as well as to this inferior stratum, or to the united strata, and this without electrical manifestations. To this cloud I give the name pallium. When the superior stratum is formed of cirrus it constitutes the pallio-cirrus, and when the inferior stratum is formed of cumulus it constitutes the pallio-cumulus. The fragments of clouds, which differ entirely from the cumulus or cumulostratus, are the fracto-cumulus.

From what has been said, the necessity of distinguishing these two strata by different names is evident; but this necessity results, more