A moderate prevalence of S. W. winds extending also to S. E., S., and W seems to be the real state of the case when local influences are eliminated. (See Plate 7.)

Until within a few years we knew next to nothing in regard to the winds in the basin of the Jenisei. Middendorff had expressed the opinion that the S. W. winds of Europe extended to the lower Jenisei, but there were not facts enough to sustain his opinion. The observations at Krasnojarsk showed that this was the case on the middle part of the river, while Jenisseisk, situated more to the N., has prevailing S. E. winds. This is caused by a change in the direction of the Jenissei from the mouth of the Angara; it flows from S. E., and, as at other points, the winds from the upper part of the river are prevailing.

It will be seen that the winds of the summer are very different from those of winter. The flow of air towards the depression of central Asia is the principal feature at this season. In Siberia we have the influence of the Arctic Ocean, which is principally felt. It is especially the Kara Sea with the Obi Bay, extending further southward than other parts of the Polar Sea, which we must consider. It must be remembered that the steppes and deserts of central Asia are not separated by any barrier from the Arctic Ocean, in the meridian of western Siberia, so that the air of the Arctic flows freely towards those countries with their high temperature and low pressure. In comparing the table given here for western Siberia with that for European Russia, it will be seen that N., N. E., and N. W. winds are much more frequent in the same latitudes in Siberia. In this the influence of the Arctic Ocean is to be seen, although westerly winds from the Atlantic Ocean also extend there. Pressure is not steady on the Arctic Ocean, its fluctuations are great even in summer, and when a storm-centre passes over it, the air from the Atlantic Ocean and southern Europe will be drawn in to supply the deficiency, as a S. W., W. or N. W. wind.

In summer central Asia has the same winds as western Siberia, W. and N. W., while in winter the difference is great. This is clearly shown on Plates 5 and 6. Semipalatinsk, being situated in the division of zone 50°-53°, has a system of winds intermediate between western Siberia and central Asia, the E. being the most frequent in winter, but southerly winds also occur.

Further south, on the lower Syr-Daria, at Taschkent and at Krasnovodsk (on the eastern shore of the Caspian) N. E. winds largely prevail in winter. That this is also the case in other parts of central Asia, where no long-continued observations have been made, is the report of nearly all the scientific travellers who have visited this country.1

In the prevalence of easterly winds Central Asia resembles the steppes of Southern Russia, but there are two important differences. First, the winds are more northerly; second, they prevail to a much greater extent. In Central Asia the mean direction in winter is between N. and E., while in southern Russia it is between S. and E., Astrachan and Orenburg excepted, but these places are already on the border of central Asia. The reason of this difference of the two regions seems

92

1

I refer, for example, to Khanikof, Basiner, Helmersen, Severtzof.
July, 1875.

to be that in central Asia the belt of highest pressure lies clearly north, while it is N. E. from southern Russia, where it is also at a greater distance further and its influence less felt. (See Plate 14.)

It was also Wesselowski who proved the existence of a zone of N. E. winds in Central Asia, though the observations at the time when he published his work (1857) were very few.

Below are the percentages of winds for spring and autumn:

The proportion of westerly winds is larger in autumn than in winter, except in Krasnojarsk, where the great frequency of S. W. winds in winter has a local cause. Westerly winds are the most frequent at Jenisseïsk and Tobolsk, which is not the case in winter. The westerly winds in autumn are stronger than in winter, and local conditions not so important.

South of 50° easterly winds prevail largely. Pressure has risen in central Asia in autumn, and the region of high barometer is again found to the northward, yet not so much as in winter, as I have shown in the case of Orenburg. (See Plate 7.) The Austro-Hungarian polar expedition has given us an insight into the winds of the region between 75°-80° lat. N., between Nova-Zembla and the newly discovered land of Francis Joseph. As the observations have not yet been reduced, I can but mention some remarks about the winds made by Capt. Weyprecht.' In the first winter, when they were drifted from near Cape Nassau to about 78° L. N. and 73° Long. E., they had S. E. and S. W. winds, in the spring the number of N. E. increased. At this time they had drifted to the westward. In the second winter (October, 1873, to May, 1874), they were about 791° L. N., and 59° Long. E., not far from Francis Joseph Land, and had largely prevailing E. N. E. winds (more than 50 per cent. of all winds).

It seems that in the polar sea, north of western Siberia, as well as in that north of Europe (Bear Island and Spitzbergen), the polar winds are far from prevailing to such an extent as in the same latitude on the North American continent and the islands north of it.

The observations in northern Nova-Zembla2 show also a considerable number of

2

Petermann's Mittheilungen, year 1875, No. 2.

By Capt. Tobiesen, calculated by Prof. H. Mohn, see Petermann's Mitth. 1874, No. 5.

southerly winds in winter. The following are the percentages in winter on the northern coast of Nova-Zembla,

Here it seems that the winds blow from the land towards the partially open sea, with its low pressure and high temperature. By winds from the land I mean here local winds from the island itself, as also those from the cold Siberian continent.

We have seen before that prevailing westerly winds extend to the Jenisei. Farther north and east we have but very few observations. It seems that we have here the region of polar calms in winter. The number of calms increases towards the interior and N. E. of Siberia, till at last there can be said to be no prevailing wind. This is the region of highest pressure in winter, as shown on Plate 14, and of also the greatest cold. Here, unlike the American polar regions, the cold of winter is very permanent, and also high pressure. The cold is not brought by winds, but is generated on the spot by radiation.

I give below the percentages of winds as observed at some few stations.

In the first three places, situated in the vicinity of the Arctic Ocean, there is a decided prevalence of monsoon winds-from the land in winter, from the sea in The mean direction at Nijnikolymsk1 is in—

summer.

Year, S. 11° E. .25.

Summer, N. 58° E. .48: Winter, S. 16° W. .48: The direction of the winds in autumn and spring is probably nearest to that of winter, as may be expected from so high a latitude, where the land is colder than the sea a great part of the year. Thus the mean yearly direction is nearly S. The direction of winds on the northern coast of Siberia is about the same as on the shores of the White Sea (Archangel and Kem).

It is difficult to determine the reason of the frequent N. winds at Yacoutsk, if the air flows towards the Pacific Ocean and is deflected from its true course by the direction of the valley. At any rate, calms are the prevailing feature in win

1

The detailed calculations on the winds at this place were published by Spassky in his "Sibirski Vjestnik," year 1823. I have used here only the figures given by Wesselowski, p. 231, as I could not obtain the original.

ter. In the summer, winds from N., E., S., and W. are about equally frequent. It seems that in September and October, when westerly winds are so prevailing in Western Siberia, warm and moist currents of air from the Atlantic can extend to Yacoutsk. At least westerly winds reach the maximum of their frequency in October (20 per cent.). In this month the flow of air towards Central Asia has ceased, while pressure has not risen high enough at Yacoutsk to prevent westerly winds from the Atlantic. October is also the cloudiest month of the year, the amount of clouds being 6.9, while March has only 2.6. The number of rainy days then is also the greatest in the year.

At the mines of Nertschinsk calms are more prevalent than at any other station we know of. In the winter months 65 to 70 observations out of 100 show no movements of the air, and the recorded winds are generally weak. In spring and summer there are less calms and more strong winds. The basin of the Upper Amoor is thus shown to belong yet to the region of Siberian calms (in winter).

While this is the case in the lowlands and valleys, it seems that the conditions are different in higher regions of the atmosphere. At Mount Alibert, 200 miles west of Irkutsk, and over 7000 feet high, a very constant and strong W. N. W. wind is observed. This place was inhabited some years on account of rich mines of ✓graphite, and it was necessary to erect a wall to protect the inmates from the violence of this wind. The mean temperature was found to be much higher in winter than in the same latitude in lower levels. This wind is probably the upper current flowing towards the Siberian pole of highest pressure. It has been supposed that such upper currents flowed towards all regions of high pressure, but this has been proved only for the polar limits of the trades,

MONSOON REGION OF EASTERN ASIA.

Southeastward from the coldest space of Siberia, towards the Pacific Ocean, we have the region of Asiatic monsoons. I have already explained the cause of the movement of air in this region, and it is only necessary to show how far it extends and how small our knowledge of the northern part of the monsoon region was until the last year. The percentages of the winds in winter and summer are given in the annexed table:

The mass of air which is drawn towards the Asiatic continent in summer is so great that the ordinary conditions prevailing over extensive areas of the oceans must be disturbed, as shown on Plates 5 and 14. As there is also a great mass of air drawn towards India and Indo-China, we must here consider Eastern and Southern Asia together.

The summer monsoon of Asia is a deflection of air already in motion, that is of part of the S. E. trade of the Indian Ocean and part of the N. E. trade of the Pacific Ocean. It is easy to prove this for the Indian Ocean, as the observations there are numerous and well discussed. This is not the case for the Pacific Ocean. Yet seeing a region of high pressure about 30° N. to the E. of China, it is impossible to conceive how the air from above it should not be drawn towards the heated Asiatic continent with its low pressure. Probably at the beginning of the summer monsoon, only the air over the nearest parts of the ocean is drawn towards Asia, and the circle extends as long as the pressure continues to sink over the continent. The direction of the winds in summer on the coast of E. Siberia, as well as in China and Japan, shows that they cannot have come from the southern hemisphere, as they otherwise would have a direction from the S. W. as in India, and not E., S. E., or S. It seems that the air from the Pacific supplies the northern part of this region, from about 25° to 60° N. In Southern China the prevailing winds are already S. W., so that this is probably air from the southern hemisphere. (See Plates 5 and 6.)

As in summer the Asiatic continent attracts the winds, so, on the contrary, in winter a continuous stream of cold dry air pours out from it towards the surrounding seas. It takes mostly two directions: towards the depression in the northern part of the Pacific as S. W., W., and N. W. winds, and towards the equatorial region as a N. E. On the coast of E. Siberia, in northern China and northern Japan the winds are mostly N. W., in southern Japan and middle China they are N., and near the tropics they have a direction from the N. E.

The climate of the whole monsoon region is characterized by a great regularity. This is not only the case in the tropics, but also in the temperate zone. The periodicity of the change of monsoons is the leading feature, taking place at more or less fixed periods, with slight changes from year to year. The N. monsoon of winter is the dry time of the year, the summer or S. monsoon the time of clouds and rain. So, for example, at Pekin the amount of clouds is 2.5 in January and 6.3 in July, at Ochotsk, Ajan and Nikolaievsk (Amoor) 2.5 in January and 5.0 in August (an entirely clear sky =0, an entirely overcast = 10). At Pekin the quantity of rain in July is more than fifty times greater than in January.

As this distribution of rain and clouds is caused by the monsoon, which brings the dry, cold air of the continent in winter, and the vapor-laden air of the sea in summer, thus causing the above-mentioned periodicity, we have means of judging of the character of the climates of this region even without having observations of winds. For a great extent of country, in China and Mantchooria as well as in eastern Siberia, we have no long-continued observations, yet the general character of the climate is known. Thus we must include in the monsoon zone, besides the tropical countries of India and Indo-China, all of China and Japan, Corea,