tells us in any particular case how far grammatical elements, in accordance with their primitive power, are able to express different shadows of meaning in the spoken language of a people.

On the advantage which philosophy or science in general derives from comparative philology, I do not venture to add anything after what was so fully and clearly explained yesterday by Chevalier Bunsen, the representative of German science in this country. Language must be considered, in its connection with nature and with the human mind, as being the natural expression of every natural impression, as being the higher unity and absolute reality of objective nature and subjective mind. Language stands in the system of the intellectual world as light stands in the system of the physical world, comprising all, penetrating all, and revealing all. There is more indeed to be read in human language itself than in anything that has been written in it.

Fourth Report on Atmospheric Waves. By W. R. BIRT.

In accordance with the resolution adopted at the last meeting of the British Association, I have the honour to report that about thirty sets of observations have been obtained from various stations in the British Islands: the extreme points of the area embraced by these stations are the Orkneys and Jersey in one direction, and Galway and Dover in the other. The observations have been executed with great care by the respective observers, and mostly at the hours named in the instructions. In some cases the observations have been continued through October, November and December; in others, they commence about the middle of October and terminate at the end of November. As instances of the increasing interest manifested on this subject, I have the pleasure to notice that I have been furnished with curves from stations in the North, where the barometric movements have been considered to result from the transit of the great November wave. These curves are referred in each case to the same period, namely, from the 2nd to the 17th of November; and the observers have invariably regarded the regular rise and fall that occurred. between these epochs as indicating a well-marked return of the great symmetrical wave.

Observations.-The following Table contains the names of the stations and observers from which observations have been received: it is right to mention that a series of observations was received from Birmingham, but the curve presents so many anomalies, that it has not (except in some minor instances) been employed in deducing any of the succeeding results.

N.B. I am indebted to the Honorable the Corporation of the Trinity House for the Lighthouse observations, and to Rear-Admiral Beaufort, R.N., for the observations made on board the surveying vessels.-W. R. B.

The return of the November wave in the year 1846 occurred under very dissimilar circumstances to any that have been noticed in former years, and when contrasted with its transit in 1842, is extremely remarkable. The barometer at London maintained an elevation above thirty inches from the 26th of October until the 16th of November, both inclusive, with a very slight depression (012 only) below thirty inches on the afternoon of the 2nd, the period of the commencement of the wave. About 9 A.M. of the 17th, the mercurial column had slightly descended below thirty inches, so that the development of the wave at London was altogether above thirty inches. In 1845 the crest was scarcely elevated above this line. In consequence of this peculiarity the London curve is flat compared with those of former years, the altitude scarcely exceeding half an inch; these circumstances, in connexion with the usual character of the weather attendant on an elevated mercurial column, and the absence of strong gales of wind at the commencement and close of the wave, had no small tendency to mask it in the south-eastern parts of our island. The projected curve however strikingly developes its essential features. The five subordinate waves, of which the great wave is composed, are well seen, although the inflexions are not strong; the central or crowning wave at London occupied five days; the subordinate waves on the anterior slope a little more than a day and a half each; the wave that immediately followed the central wave exhibited a greater development, having an amplitude of four days; and the closing subordinate wave a day and a half. The epoch of transit, November 9th, was slightly earlier than usual, but sufficiently near to regard these movements as a decided return of the wave, and strictly in accordance with the type as expressed in my last report.

Previous to proceeding with the examination of the observations received, I beg to solicit the attention of the Association to the principle that I laid down in my last report, namely, that a barometric curve, including a complete rise and fall at any one station, does not represent the form of any reality in nature. The peculiar combination of barometric ascents and descents occurring about this period of November, and exhibiting a remarkably symmetrical arrangement under very diverse circumstances and with very different barometric altitudes, is a phænomenon that appears to receive its explanation in the crossing of various systems of atmospheric waves or currents. The south-eastern portions of our island have generally presented the most symmetrical curves; and this circumstance, connected with the constant decrease of oscillation from the north-west, appears to indicate that about this part of the year, whatever may be the volumes of the individual waves of pressure, however they may be affected with regard to velocity, or may be superposed on much more extensive normal waves, the two systems so cross each other in their respective progressions towards the north-east and south-east, as to produce by their combined effects the barometric phænomena (so far as regards the curve) really observed. In the following notices of the observations, the term wave will be restricted to the protuberances on the symmetrical curve.

The most symmetrical curve has been obtained from observations made on board H.M.S.V. Porcupine, under the superintendence of Captain Frederick Bullock. The wave commenced off Walmer on the 2nd, culminated on the 10th, the ship being in Ramsgate harbour, and terminated on the 17th, the vessel in Dover harbour. This curve is characterized by a rapid fall, which took place immediately upon the culmination of the great wave, value 27 between 9 A.M. of the 10th, and 3 P.M. of the 11th. A second culmination, the crest of the first subordinate wave on the posterior slope, occurred on the 13th.

The highest reading of the barometer in the series occurred at St. Vigean's

near Arbroath, east of Scotland, at 3 P.M. of November 12th, the crest of the first subordinate wave on the posterior slope, value 30'66. This wave at this station, and indeed at all the northern stations, was so strongly developed as to appear in the form of a large protuberance on the back of the symmetrical wave. The culmination of the great wave occurred at 6 P.M. of the 10th, value 30-64; the fall immediately after this culmination was slight compared with that at Ramsgate, value '06; the barometer commenced rising at St. Vigean's earlier than at Ramsgate.

At Stornoway in the Western Isles, the curve obtained from observations made on board H.M. ketch Sparrow, under the direction of Commander Otter, presents a most striking and marked difference to that representing the observations at Ramsgate. At the commencement of the great wave the barometer at the Western Isles was more than half an inch below that at Ramsgate, wind S.W., force 9. The two anterior waves, which were scarcely perceptible at Ramsgate, were well-developed at the Western Isles, the barometer at the latter station gaining on that at Ramsgate during their transits. On the evening of the 6th both barometers began to rise, the one at the Western Isles very rapidly, so that at noon of the 8th the pressure was equal at the two stations. The curves then diverged, Rainsgate being still superior, and the movements being of an opposite character until midnight of the 9th. The fall characterizing the Ramsgate curve did not occur at the Western Isles; the curve at the latter station continued to rise until the forenoon of the 12th, when the altitude of the barometer was equal to that at St. Vigean's: these curves, Stornoway and St. Vigean's, run nearly together until the morning of the 15th, both being superior to Ramsgate; at 9 P.M. of the 15th the Stornoway curve crosses the Ramsgate, the barometer at the former station falling rapidly. At Stornoway the first wave on the posterior slope of the symmetrical wave is so strongly developed as completely to mask the central wave.

The following Table exhibits the phænomena above noticed; also the alternations of the barometric differences.

TABLE II.

Barometric altitudes and differences observed at Stornoway in the Western Isles north-west of Scotland, and Ramsgate south-east of England, during the transit of the Great Symmetrical Wave of November 1846.

Somewhat similar alternations of pressure were exhibited at the stations Stornoway and Helstone; in the following Table the altitudes have been corrected and reduced to the level of the sea. It is to be remarked that during the period of the great wave, the curve at Helstone was altogether inferior to that at Ramsgate.

TABLE III.

Barometric altitudes and differences observed at Stornoway in the Western Isles, and Helstone in Cornwall, during the transit of the Great Symmetrical Wave of November 1846.

The curves at Stornoway, the Orkneys, Largs and Applegarth Manse, the north-western stations, run nearly together during the whole of their course; their intersections are very numerous; the greatest extent of their divergence is 27. During the transit of the first posterior subordinate wave, Stornoway exhibited by far the greatest pressure. The curve at the Orkneys agrees with that at Stornoway in the maximum being identical with the crest of the first

* Reading at 4 P.M.