the name of a wind, which is mentioned in almost every account that we have of Provence, and which is remarkable for blowing almost the whole year from north-west or west-north-west, in a climate where the wind should be variable. It is said to contribute to the salubrity of the air, by dispersing the exhalations of the marshes and stagnant waters, so common in the south of Languedoc and Provence; but at times it is also very injurious, or at least very troublesome. It is not, however, on either of these accounts that it is introduced into this Work, but for the sake of the causes assigned by Saussure for its constancy, which may be applied to other winds that nearly resemble it; and which he found might be reduced to three.

"The first and most effectual cause (he says) is the situation of the Gulf of Lyons, the banks of which are the principal theatre of its ravages. This Gulf, in fact, is situated at the bottom of a funnel, formed by the Alps and Pyrenees. All the winds blowing from any point between west and north, are forced by these mountains to unite in the Gulf. Thus, winds which would not have prevailed but at one extremity of the Gulf, or even much beyond it, are obliged to take this route, after having undergone the repercussion of these mountains; and the middle of the Gulf, instead of the calm which it might have enjoyed, is exposed to the united efforts of two streams of wind, descending in different directions. Hence arise those whirlwinds which seem to characterize the mistral, and appear to have induced the ancients to call it Circius, a turbine ejus ac vertigine. See Zul. Cellius, i. ii. cap. 22.

"The second cause is, the general slope of the grounds, defending from all sides towards the Gulf; which becoming all at once lower and more southerly than the lands extending behind it, in these joint circumstances, rendered the hottest point of all the adjacent country: and, as the air on the surface of the earth always tends from the colder to the warmer regions, the Gulf of Lyons is actually the centre towards which the air from all colder points between east and west must press. This cause, then, alone would be productive of winds directed to the Gulf, even if the repercussion of the mountains did not exert its influence.

"Finally, it is well known, that in all gulfs the land-winds blow more forcibly than opposite to plains and promontories, whatever be the situation of those gulfs. I apprehend, indeed, on strict examination (says our author), that this cause is blended with the preceding; but as the fact is generally admitted, and in some cases can be explained only by reasons drawn from the effects of heat, it may not improperly, perhaps, be distinctly mentioned. It is, at least, necessary to suppose, that several causes produce the mistral, in order to understand why, notwithstanding the variability of the seasons and temperatures, that wind is so singularly constant in Lower Languedoc and Lower Provence. A very remarkable instance of this constancy is recorded by the Abbé Papon, in his Voyage de Provence, tom. ii. p. 81. He affirms, that during the years 1769 and 1770, the mistral continued for fourteen months successively. But the three causes which I have stated, taken separately, will explain its frequency, and, united, will account for its force."