by solar rays in its course, and also a slight mixture with less cool strata of air.

"This being granted, one easily understands that, as at the extremity of South America, in latitudes where temperatures of 12°, 15°, or even of 20° below zero (centigrade) are occasionally possible in winter on the surface of the ground and near the level of the sea, a strong wind, that is to say, a great mass of air moving W.S.W. to E.N.E., beat against the mass of the southern Andes, where, by its acquired velocity, it ascended, still keeping its E.N.E. direction. . . . Then, in the higher current, its northerly movement was retained, and by terrestrial rotation it gradually lost its easterly direction, until, after a long westerly movement, it finally became a south wind. For this frozen wind to gain at once the latitude and level of the plateau of Barbacena, it is now sufficient . . . to meet favourable circumstances to extend northwards. . . . Thus we see that for this phenomenon there was needed a rare combination of numerous and fortuitous circumstances over a considerable journey."

I will next translate a few extracts concerning other meteorological phenomena; but space prevents my giving more than very short summaries, and excludes my detailing the causes, for which I refer those interested to M. Liais's exhaustive work.

HAIL.

"The hailstones are large, very hard, and I have seen them take three or four minutes to melt. In 1862, I observed four falls of hail in November. There are, according to the inhabitants, on an average twenty in a year. At Rio de Janeiro falls of hail are rare. I have only known four from 1858 to 1864, of which I saw three; and two others from 1865 to 1871. The first fall was on February 22, 1859, when there were only a few hailstones mixed with a heavy storm of rain. Two others were on October 22 and 30, 1863, during heavy storms, accompanied by thunder. The hailstones were lenticular. I measured some, eighteen millimetres in diameter, and one millimetre thick. They produced a general surprise; and I have seen persons of sixty. years of age who never remembered having seen the like. But the fourth fall was the most remarkable. It occurred October 10,

1864, during a terrible hurricane, five days after the great hurricane of Calcutta, and after the extraordinary cold in France from October 2 to 4, 1864, consequently during a considerable atmospheric disturbance, whose action had extended to very distant parts of the globe. This fall was extraordinary, accompanied by a violent storm, and a wind by which, at certain places, venerable trees were uprooted. I was not then at Rio, but I knew that there occurred in that place hailstones as thick as one's thumb. Since that time there have been no heavy falls of hail (up to 1871), but only twice a few small hailstones in some storms. Hail may thus be considered as an exceptional phenomenon at Rio de Janeiro, and on the plateau of Minas Geraes as an habitual phenomenon. At the north of the empire, falls of hail are almost unknown.”*

STORMS AND RAIN.

"Storms are excessively frequent in summer at Rio de Janeiro and in the province of Minas Geraes. There is sometimes magnificent lightning, not only bifurcated, but with a considerable number of branches; and the discharges are repeated occasionally, from the same point, seven or eight times in a second. The frequency of storms diminishes considerably on approaching the north. At Pernambuco, during eight months, I only saw lightning twice; and I have never heard thunder.

"In Rio de Janeiro, and on the coast of Espirito Santo, it rains every season of the year; but, as a rule, much more in summer, and less in winter. Generally, the dryest months are June, July, and August. In the whole of the interior of Brazil, these months are always invariably dry, and the seasons divide into two the time of rain from October to March, the dry season from April to September. . . . On the coast of Pernambuco, the rains are specially abundant in the months of June, July, and August, which are the dry months in the south." M. Liais enters at length into an explanation of this curious inversion of climate, which is briefly as follows:-From the lofty table-lands of the interior, when heated by the midsummer vertical sun, arise cur

* During the ten months that I was in Minas, I never remarked any hail accompanying the tremendous thunderstorms.

rents of hot air, and the moisture-laden currents from the sea rush in to take their place, passing over the lowlands and discharging themselves in storms on the elevated plateaux, which are near the sea at Rio, and far from it at Pernambuco. In winter, the reverse is the case.

MISTS.

"Mists" (brouillards; but I cannot call them fogs, as these are so mixed up with the idea of our smoky fogs)" are constant in the morning on the great rivers of the interior, and equally frequent in the morning, during the winter, near the coast, especially in the valleys and bays, as in the bay of Rio de Janeiro. Those on the banks of the rivers in the interior occur in the dry as well as in the wet season, and are due to the great excess of the temperature of the water above that of the morning air, an excess of 3° or 4° centigrade, the former sometimes rising to 60° centigrade. These mists sustain the vigour of the vegetation on the banks during the dry season, and the trees there retain their leaves. They lose them, on the contrary, away from the river-banks, owing to the dryness: and this circumstance even occasions a special riparian flora nearly approaching in its characteristics that of the virgin forests."

Y

METEOROLOGICAL NOTES

In Minas Geraes, from August 18, 1883, to May 27, 1884. At Botafogo, Rio de Janeiro, from May 30 to
July 27, 1884.

TAKEN BY H. C. DENT.

NOTE. The great pressure of professional work and other circumstances prevented regular observations before the beginning of October.

REMARKS.

Showers from 6.30 a.m. 7.30 a.m. 60°. No sunshine. Much

rain, especially after 3.30 p.m. 8 p.m. 60°.

Cold and cloudy.

7 a.m. 56°. Cloudy all morning. Fine afternoon.
Very hot day, but breezes.

Showery all night. Pouring incessantly from 9 a.m. till 2 p.m.
Very fine day.

Fine day.

Perfectly cloudless day till 5 p.m. Heavy showers 5 to 6 p.m.
Fine evening. Thunder and lightning.

Magnificent cloudless day.

Cloudless sunrise. Grilling from 10 to 2; then a few clouds
and cool breeze.

Cloudless day. 86° in tent at II a.m.

Cloudy and very cool morning. Sunshine 12 to 2; then cool
afternoon. 10 p.m. 67°. Lightning as usual.

At 4 p.m. 78° in tent. Cloudy and cool.

Cool, cloudy. Thunder and slight rain. About an hour's
sunshine during the day.

104° at 8.30 a.m., 88° in tent midday, 70° in tent at 8.30 p.m.,
but cool. Great thunderstorm, deluge, during night.
Very sultry. Slight breeze.

Dull, drizzle. Sun at intervals. 6 p.m. 65°, cool.

Much thunder,

Very cold morning. Scotch mist. From 10a.m. rather warmer.
Strong breeze, occasional sunshine all day.
lightning, and rain all night. 62° at 5 p.m.
Dull, overcast morning, with rain.
Much rain, night and day. Average, 64°. Much wind, very

cold and miserable.

Cool.

8 a.m. 58°. Cool, cloudy day. No sunshine. 8.30 p.m. 60°.
54° at 7 a.m. Cool, dull, windy, rain. Drizzle all day. 6
p.m. 54°.