entitled, Recherches sur les Chaines de Montagnes et la Climatologie comparée. This table contains the mean and season temperature, as well as the temperature of the hottest and the coldest months for 305 places in the two hemispheres. I have also supplied the place (p. 228) of the little table of the limit of perpetual snows, at different latitudes, which is given. in the German book, by that which M. DE HUмBOLDT has given in the same work.

"M. KAEMTZ's table for the reduction of the barometer to zero extended only from 540 to 778 millimetres; for this I have substituted M. DELCROS', which extends from 400 to 800 millimetres (p. 241), in order that it may be useful to those who apply themselves to the determination of heights by the barometer. At page 243, I have inserted a small paragraph on the correction of the barometer due to capillary action; and I have added the table constructed by M. DELCROS for making this correction, which is of such importance when we desire to know exactly the weight of the atmosphere. The chapter which treats on the height of the barometer at the sea-shore, has been completed (p. 277) by a table, in which MM. SCHOUW and POGGENDorff have given this height for a great number of places. The paragraph on the influence of the winds on the differences of level calculated by the barometer, has been replaced by that which M. KAEMTZ gave in his preface, as presenting results more conformable to

truth. These are the substitutions, which I have taken the liberty of making in the text; they are a necessary consequence of the progress of meteorology. In imitation of the German book, I have printed the names of men in different characters from those of the context. At the end of this book will be found an alphabetical list of these names, which will facilitate the search after facts or theories, of which the most faithful memoir has often retained nothing beyond the name of the author.

"M. LALANNE having graphically represented the greater part of the tables, I have substituted for Plate 1. of the German text that of the frontispiece, which represents a halo that I observed in Sweden, with M. BRAVAIS. It appeared to us worthy of being represented; because it presents in one all the circles and arcs which have most frequently been observed, and which theory explains. At page 430 will be found a note, in which this figure is compared with the projection of a complete halo given by M. KAEMTZ, Pl. v. fig. 3.

"Plate II. of the German book was partly occupied by curves; I have replaced them by a general figure, and by the details of FORTIN'S barometer as modified by M.DELCROS,―a barometer equally suited to meteorological observations and to levelling.

"Plate III. representing the clouds, left much to be desired in point of execution; I have had it entirely done over again.

"Plates Iv. and v. have been faithfully copied.

"Plate VI. of the original work represents the isothermal and the isogeothermal lines of the northern hemisphere, laid down on a MERCATOR's projection. I have preferred giving the isothermal lines only, and on a polar projection, which has the advantage of shewing how the curves become re-entering in the high latitudes, and form the two frigid poles."

In the English translation I have included the additions alluded to by M. MARTINS, in the above extract, and have added a second Appendix of Notes, in which I have endeavoured, as far as practicable, to illustrate certain passages in the text; and to describe some of the new and important registering instruments, which recent necessities have called into existence, and modern science has enabled us to construct: among these, WHEATSTONE'S Electro-Meteorological Register holds a first rank. I have also given an account of the observatories of Kew and Greenwich; and have described the instruments employed, illustrating the description by printing a fortnight's observations made at Kew. Plates VII. and VIII. of the beautiful aurora seen by M. LOTTIN; Plate IX. of OSLER's anemometer; and Plates x. and XI. of Mr. WHEATSTONE'S Register, have been also added to this edition.

London, Nov. 1844.

CHARLES V. WALKER.

INTRODUCTION.

METEOROLOGY is that part of natural philosophy which treats on the phenomena and the modifications of the atmosphere, in order to analyse them and seek their explanation. Plunged at the bottom of the atmospheric ocean in which the earth is enveloped, we are witnesses of the changes which are incessantly going on in it. Serene or cloudy, cold or hot, calm or agitated, the atmosphere exercises a powerful influence over all organised beings. There does not exist a man who has not asked himself what the cause is of these continual variations. It is not simply the desire of knowledge which urges him to this inquiry; but these questions are often of the highest importance to the husbandman, the sailor, the artisan, and the physician. Our physical and moral well-being depends, in a great measure, on the condition of the atmosphere. When the sky has remained for some weeks covered with dark clouds, the spirits are affected, but the mind becomes serene as soon as the sun appears again: so also, during changeable, damp, and cold weather, the number of invalids is always greater than during fine weather.

From the very remotest antiquity men have been occupied in investigating the causes of these variations. This study even preceded that of natural philosophy, properly so called, because it embraces the most striking phenomena of the inorganic world. In the works of the Greeks and Romans, we find a multitude of observations and laws carefully treasured up. Among all nations, even the least civilised, travellers have met with some notions of Meteorology. It would seem, then, that this science ought to be one among the most advanced, because, for thousands of years, it has been the object of the labours of so many intelligences. Unfortunately it is not so; and a few lines will suffice to give the reason.

B

The number of observations on the modifications of the atmosphere is doubtless considerable; but they are, at the same time, observations in the most restricted sense of that word. We observe the phenomenon presented to us, but we cannot modify and vary it at pleasure; we cannot even reproduce it at will. In a word, we cannot have recourse to experiment. Our means and our powers are much too limited to give us the power of producing the least changes in the atmosphere. We are hence compelled to register facts; and, as w. Herschell has very well observed, we resemble a man who hears now and then a few fragments of a long history related at distant intervals by a prosy and unmethodical narrator. In recalling to mind what has gone before, he may occasionally connect past with present events; but a host of circumstances omitted or forgotten, and the want of connexion, prevent his obtaining possession of the entire story. Were we allowed to interrupt the narrator, and ask him to explain the apparent contradictions, or to clear up any doubts on obscure points, then might we hope to arrive at a general view. The questions that we would address to nature are the very experiments of which we are deprived in the science of atmospheric modifications.

When reduced to observation, Meteorology cannot possibly advance at a pace equal to that of the other branches of natural philosophy. To attain the power of establishing laws, we are compelled to register a long series of facts similar in appearance, that we may draw from them some general results. These being once obtained, we pass on to the study of isolated variations. We know, for example, that the barometer is ever oscillating, and that it never ascends without its afterwards descending again. We notice a certain connexion between the state of the atmosphere and these variations in atmospheric pressure. In like manner, as the barometer is always in motion, so also the temperature does not rise uniformly from the coldest degree of winter to the hottest moment of summer. Thus cold and hot days, in so far as the season is concerned, interrupt this regular march. Knowing that the weight, and consequently the equilibrium, of the different atmospheric strata, changes with the temperature, we may inquire whether barometric oscillations may not be connected with these changes of temperature. In order to determine whether this relation does exist, the thermometer and barometer are observed for a long period at the same fixed times. We examine how often and how much the two have varied in the space of twentyfour hours: then, grouping separately the observations in