calm, the temperature of water may fall 2° or 3° below the freezing-point before ice forms.''* The formation of masses of ice in the atmosphere belongs rather to electrical phenomena, and would more properly have followed notice of the thunder-storms of the Andes, so common in summer. There was only one during the three years, viz: on the 13th January, 1852. It commenced suddenly about 4 P. M., after a day moderately overcast by masses of cumuli and a low temperature, (68° at 3 P. M.) and was unaccompanied by lightning. The stones were truncated cones and pyramids, with spherical bases, as though they had formed portions of spheres perhaps an inch in diameter. Their bases were of a milky yet translucent ice, whilst the upper halves were softer, whiter, and more opaque. The storm lasted about ten minutes, though the sky remained clouded over until after 7 P. M., and occasional drops of rain fell all the afternoon. Two miles west of the observatory there was a violent squall of wind, but no hail; and on the distant summits of the Andes a large body of snow, or hail, was deposited. At night the thermometer fell below 50°, the barometer remaining nearly 0.15 inch above its mean height-a very great variation from its normal elevation, in a country where the fluctuations are so small. Ordinarily, the air is calm from about sunrise until between 9 and 10 o'clock A. M., at which time a wind commences from the S.W. This increases in strength till 2 or 3 P. M., and then moderates as gradually to sunset, when it is again calm. Its violence on the plain was rarely more than what, in nautical parlance, is called "a fresh breeze;" but on the elevated summits of the Andes, over which it also extends, it is usually excessive. Lieut. MacRae wrote me that it was so strong when he arrived at the pass of the Cumbre, between Santiago and Mendoza, that, as early as 10 A. M. it almost overturned both mule and rider; and arrieros declare they have seen small stones blown away by it. Deep ravines debouching on the plain near Santiago, and many hills near it, cause so many deflections as to render it impossible to determine the true direction of this wind by estimation; nor will it be practicable to do so, except by placing a register anemometer half-way across the plain. Our guides were, the direction that smoke was moved, or, failing this, the plane in which some of the lofty and pliant poplars were inclined. From these we found that the current varied in its direction from W.S.W. to S.S.W., and when strongest was most generally from S.W. These winds are attended with a clear atmosphere, and the only clouds to be seen are formed within the valleys of the rivers, at elevations of 5,000 to 10,000 feet above the plain, and which continue ascending until they rest in heavy cumulous masses over the elevated range. Vapor is rarely condensed immediately over the valley during the day, but a line not unfrequently collects at night half-way up the near chain, and remains there until dispelled by the heat of the day. With night the entire mass above the Andes disperses, though not until after a display of lambent sheets of lightning, sometimes continuing beyond midnight. Whenever similar corruscations were seen over the central range to the N.W., as was the case on two or three occasions, they were invariably followed by rain. After sunset "el terral," or, as it is called in the south, "el puelche," a land breeze, commences. This, first perceptible on the coast, recedes slowly towards the Andes, where it is scarcely felt until near morning, thus proving itself a true wind of aspiration. Its apparent direction is modified at Santiago by causes analogous to those influencing the "travesia," as the day wind is named; and we find it one night from N.E.; the next, perhaps, or even at a subsequent hour of the same night, from S.E. It is never more than "a light breeze," and ceases entirely throughout its range by sunrise. Late in the season a sort of dry fog, resembling thin smoke, deprives the atmosphere by day of something of its transparency, though the nights are all that the astronomical observer can desire. Then the Andes, whose crests are not less than 18 miles distant in an air-line, look almost within stone-throw, and the stars rise over them with a steadiness and brilliancy known * Anales de la Universidad de Chile, Junio, 1851. in our climate only at mid-heaven. The observer will appreciate me when he is told that I have made very fair micrometrical measurements of Venus when the planet was not more than 3° above the eastern horizon, and its crescent was more than once seen with the naked eye. At times, the atmosphere was steady as the earth itself; and the colors of close double stars not greater than the twelfth magnitude were satisfactorily distinguished, though the magnifying power was 235, and the telescope fully illuminated for other observations. Such a climate places a small telescope on equality in optical capacity with a much larger one in a moist atmosphere, and there were opportunities to distinguish small objects with our 6-inch achromatic, which could only be seen with difficulty with the 20-inch reflector of Sir John Herschel at the Cape of Good Hope. Adopting Maskelyne's ratio between reflectors and achromatics (8:5), the illuminating powers being as the squares of the diameters, our 6 inches at Santiago was quite equal to 12 inches at the Cape. Early in December, wheat and barley are harvested. After the strawberry, figs and cherries are the next to ripen, the former being somewhat forced by puncturing them with an oiled needle. By Christmas day, melons, apricots, early nectarines, and one or two other fruits, are brought to market-some of them ripe, but more partially green, in which state nearly all fruits and vegetables are gathered. Garden flowers are in their perfection; dahlias, tuberoses, carnations, diamelos, (Jasminum sambac) jasmines, and a host of others, enable the ladies to exercise freely their graceful and refined custom of sending charming bouquets to friends on their Saint's day. But the hill-sides and uncultivated plain are completely denuded and desolate; the south wind drives clouds of dust from their surfaces, and the traveller avoids as much as possible the heat of the day, making his journeys before 9 A. M., or after 4 P. M. On the coast the heat is moderated by the ocean. There, the thermometer never rises as high by day, nor falls so low at night, as between the great mountain chains. In three years the temperature at the Exchange of Valparaiso, at 8 A. M., was not lower than 62°, nor higher at 4 P. M. than 78°, and the mean of all the observations was 70°.8. Owing, however, to the imperfect exposure of the instrument, these records can scarcely be regarded as true indices of out-door temperature. Moreover, each district of country has its local peculiarities; so that there is no general law by which observations made at a particular hour can be reduced to the mean temperature of that place upon the application of the correction found for any other station. Obtaining a correction for stations with whose latitude and chorography there is least contrast, the average temperature from the Valparaiso observations will not be far from 6°.5 in excess of true summer heat, or nearly 5° below that of Santiago. If it has the advantage of a lower temperature, the southerly winds quite counterbalance it by their greater violence, and the annoyance of clouds of fine sand which they whirl from hills in the rear of the town. Sometimes they are so furious as to prevent vessels from reaching an anchorage in the bay. Though it is well known that they are equally constant at a little distance from the land from Chilóe to Lima, and draw more from the westward outside of the islands of Juan Fernandez, their entire limits have never been satisfactorily ascertained. North of Valparaiso, and within a few leagues of the land, they are feebler by day, and the land breeze replaces them at night. Even in summer, fogs over the land are not uncommon. Coquimbo also enjoys at this season a cooler and more agreeable atmosphere than Santiago. The mean of the observations at the selected hours are there in excess of the mean for the day 3°.2, which, applied to Señor Troncoso's results, shows a summer temperature of only 63°.6. During the seasons of 1849 and 1850, the range of the thermometer between 8 and 9 A. M. and 9 and 10 P. M. did not exceed 16°.8; the barometer quite steady, and the atmosphere often cloudy. This last fact is obtained from his notices of earthquakes, which, omitting April and November, are more frequent than during any other months; and it is greatly to be regretted that there is no diurnal record from which to decide whether the clouds that accompany or almost immediately follow these subterranean disturbances have been only coincidences. From daily observations at Concepcion during the summer of 1850, the temperature at 3 P. M. ' was found to be 73°.5; that of evaporation at the same hour, 61°.3. At sunrise during seventeen days of February, 49°.2; sunset, 66°; and that of the wet thermometer for the same epochs, respectively, 47°.8 and 58°.2. On the 29th November preceding, Mr. Theodore Philippi, by whom the journal was kept,* tried the temperature of water in seventeen different wells, and found the thermometer range from 55°.4 to 61°.7. Selecting those where the water came nearest to the level of the earth, the lowest temperature was found in a well only nine feet deep, and at a small superficial spring. On the 7th December the temperature of the well continued the same; but by the 31st January it had risen to 57°.2, and on the 28th February was 57°. From these results he deduces that the mean temperature of Concepcion will be between 55°.4 and 57°.2. There was neither storm, hail, nor earthquake from the commencement of September to the close of February, nor did any rain fall during the last month. Poeppig † mentions an interesting fact respecting the influence of the easterly winds here. He says that when they blow in spring, they depress the thermometer in a short time from 12° to 15°; but towards the end of February, they raise it almost as much. The first he attributes to the deep snow with which the Andes are covered at that season, and the second to the high temperature to which the air upon the sandy plains of the Pampas of the Argentine republic is raised during the summer months; but neither of the three Philippis who have resided several years in the south notices the fact. It is supposed that its climate has been undergoing a gradual change ever since active destruction of the forest-trees of the vicinity commenced, and to this decrease of mean heat is attributed injury to the peach-trees during some years past. Trees of the same kind introduced from abroad, flourishing somewhat later, had proved less liable to blight. It had also been remarked that there was a great difference in the time of maturing of fruits at Tomé, on the coast, and Concepcion-less than six leagues distant in a south direction. At Valdivia the summer temperature is 60°; the lowest observed at 6 A. M., 41°; the highest (in January), 96°.2; and the mean difference between the 6 A. M. and maximum temperatures, 18°. There are, however, great differences from day to day; and the coldness of some of the nights may be judged of from the fact that there are localities of small extent where the leaves of potatoes, beans, and other plants, are occasionally frosted. From the same cause, almond-trees rarely mature their fruit, and there is a difference of more than two months in the times of flowering of similar plants here and at Santiago. The prevailing winds are from west; after that, they are most common from S. E. and S. W.-never from north, and only once from west. Rain fell on twenty-eight of the ninety-one days, and twelve others were cloudy. In all the year there were one hundred and fifty-six rainy and seventy cloudy days. In Washington the annual average number of the former is ninety. Grouping the winds with the rainy days, it is found that those which blow from any point of the compass between N.E. and N.W. are essentially rain winds; those from south to east, dry winds. In winter and spring the greatest number of rains are with N.E. winds; in summer, with west; and in autumn, from the N.W. At a little distance off the coast, northerly and N.W. winds are invariably accompanied by damp, disagreeable, and unsettled weather. When a change takes place, it is usually to the S. W., and thence to the southward; sometimes in a violent squall, accompanied by rain, thunder, and lightning; at other times it draws gradually round, and as a steady southerly wind approaches, the sky becomes clear and the weather healthily pleasant. Though usually a prelude to a clearing-up storm, lightning is always a sign of more immediate bad weather. The only barometrical observations known to have been published are those of Capt. Fitzroy, which were made in the harbor at noon from the 9th to the 22d February, 1835. On one day, (20th, when the earthquake destroyed Concepcion,) there were two other records, viz: at 6 A. M. and 6 P. M. The instrument was suspended at the level of the sea: its range, during thirteen days, was from 29.85 to 30.10 inches; and the fall recorded between 6 A. M. and 6 P. M., *Anales de la Universidad, Marzo de 1850. + Poeppig. Reise in Chile, Peru, und auf dem Amazonenstrome in 1827-'32. 1 from 29.99 to 29.92 inches. The mean of all the noon observations is 29.973 inches, and both extremes of pressure occur with northerly winds. At San Carlos, (Ancud,) from January 18th to February 4th, 1835, the barometer at noon was never lower than 29.95 inches, nor higher than 30.03 inches; mean, reduced to 32° Fahrenheit, 29.917 inches. There was a due proportion of fair weather, though moderate winds prevailed from the northward and westward, and there were only three days when it blew from S.W. The thermometric variation at the same hour was from 50° to 68°-the former temperature with a wind from S. W., and the latter with one from W.S.W. The mean at noon, 57°.3. Such temperature scarcely confirms the experience of Agüeros, who says: "Chilóe has its four seasons, but does not enjoy the benefit of those changes as do other parts of Chile; for there is neither that abundance of fruit, nor are its fields adorned with so many and such beautiful flowers and useful medicinal plants. The summer is the most pleasant season; for though, in the month of January, it is excessively hot from 10 in the morning until 3 in the afternoon, there is a sea-breeze during those hours, called 'virazon,' which refreshes the air. At this time the day is from seventeen to eighteen hours long, and conversely in winter. * * * The weather, when it is fine, cannot be depended on for any length of time; for, in the month of January I have frequently seen rains as copious and gales as violent as in the winter. During the summer months, southerly winds are more prevalent; and while they last, the weather is fine and clear, and the air particularly dry." Capt. King found the first half of December, 1829, tempestuous and wet; but it proved scarcely one-third as bad as he had experienced at St. Martin's Cove, near Cape Horn. During the preceding month the range of temperature had been from 42° to 68°.5, and the mean at 9 A. M. 53°.5. At Hobarton, (Van Dieman's land,) which does not differ much in latitude, the temperature of November, at 9 A. M., exceeds the mean annual heat 6°.65; that of February, at noon, is 70°.89 greater; and supposing these corrections approximate, the mean temperature of the northern extremity of Chilóe will not vary greatly from 48°. Autumn in the province of Santiago is not less charming than the other seasons of this so favored region-a country in whose soil and climate vegetation typical of the torrid and temperate zones, side by side, thrive equally. The native palm and pine of Araucania-the cherimoya of tropical America and the medlar of Japan-the magnolia of Florida and the olive of Asia, may all be found within the compass of a garden, not less luxuriant in their proportions and ever-verdant foliage than under the climes of their origin. All through March, and the larger half of April, unexceptionable fine weather lasts, though the atmosphere is less transparent by day than during the other seasons, and copious dews at night show its increasing relative humidity. About the close of the former month, or in the first half of the latter, there are usually from ten to fifteen days when it assumes that peculiar appearance between smoke and dry fog which is so notable at the "Indian summer" of North America. During its continuance there is scarcely any wind; and, as the temperature after noon rises to summer heat, with its fresh southerly breeze, the air is more enervating than at the latter season. Here the resemblance between the two hemispheres ceases. Unlike the North American "Indian summer," of which, its continuity once broken, there is no return until the following year, the Chilean "verano de San Juan'* is often interrupted by a renewal of the periodic winds with greater force, or by clouds; and after a day or two, there succeeds another interval when the air is tranquil and smoky. Even the moderate daily breezes from the S. W. lose their strength in autumn, and el terral is frequently replaced by one from the Western cordilleras, though the direction of the daily surface current is rarely doubtful. The mean atmospheric pressure is 28.065 inches, and the extreme of its mean daily fluctuations 0.039 inch; its periods of maxima and minima conform * St. John's summer. So named in the Argentine republic, though St. John's day is June 24th. I never heard a Chileno designate it ing to those of spring and summer during the first two months, but through May as often reversed as in winter. Between the coldest and warmest hours of the day there is an average difference of 20°; and between the extreme heat of the first and the cold of the last month, the thermometer ranges through 50°, though the difference of their mean temperature is only 10°, and that of the season 59°.4, differing very slightly from that of the whole year. The hygrometric condition of the atmosphere changes more rapidly: the mean height of a wet thermometer at 3 P. M. during March being 16° below its temperature; and in May, at the same hour, only 8°.5. The average difference at all the observation hours in the former month is 10°.1, and in the latter 4°.3; even the very last figures proving that there is still a less amount of aqueous vapor in the atmosphere than during the driest months at Washington. These would tell the reader that the rainy season was not fully commenced, had he not been prepared for the fact by the compilation from the tables of Señor Reyes. Our three years give a somewhat different result, showing an increase in the average daily time during which rain falls in May, from 1h. 7m. to 1h. 42m., and the deposit of water 6.9 inches. Every effort is made to harvest the crops of beans, capsicums, potatoes, and other vegetables, for winter use, before the rains commence; and as the grapes are ready for the vintage between the 10th and 20th of April, this is the busiest season of the agriculturist. It is also the period of the year when morning fogs are the most frequent; when the halos that almost nightly encircle the moon are most opaque, and the meteors brightest; when clouds above the Andes are in densest masses, and storms of rain, snow, and lightning are phenomena of daily occurrence among their heights. On one occasion a meteor exploded, in the direction of the greater Magellanic cloud, with a noise so audible as to command attention; and another, which was very brilliant, was witnessed by Lieut. MacRae and myself, ascending from the S.E. quarter of the heavens in a nearly vertical direction. More particularly during our last year, and in the months of May and June, there were beams of light visible on many mornings, radiating from a point of the heavens opposite the sun. Sometimes they were so broad, well defined, and distinct, as to give the intervening shadows the appearance of black streamers on a rose-colored ground. The dark spaces were usually blacker to the south than to the north of the zenith, across which many of them could be traced almost to the eastern horizon. The effect was greatly heightened when there were banks of cumuli about the mountains, for these were often brilliantly lighted up, as one witnesses at a tropical sunset. Occasionally the phenomenon was visible on both sides at once; and then it was a most beautiful sight to watch one point of radiance descend as the other gradually rose to the summit of the Andes. In twenty minutes to half an hour the presence of the sun would obliterate every trace of it. Another phenomenon, which attracted my attention soon after reaching Chile, was the red color assumed by the snow-crests as the sun approached the western horizon. The change of color began as soon as the plain had fallen under the shadow of the Western cordilleras, and increased in depth until the direct light of the sun had entirely left the peaks. Above the shadow of the cordilleras, as it crept up the Andes, there were violet and purple hues, according to their distance from the illuminated portion; and not only these, but also the red, were of greater intensity at the close of spring and in winter, when the sun was farthest north, and there had been recent deposits from the clouds. At such times the view was certainly very exquisite; and if there were radiant beams of light and shade over the Andes at the same time, as was quite frequently the case, it was a picture to which no words can render justice. To my vision the color of the snow at such times was more of a vermillion than of a red or rose color, but the assistants could only recognise it a rosy pink, even with prepared water-colors before us; another example that all eyes were never alike sensitive to colors. The Protococus nivalis, mentioned by arctic voyagers, has been found also on patches of perpetual snow of the Andes; but this shows the red color only when accidentally crushed, or a rapid thaw |