530 



NA TURE 



[February 2, 1905 



were made for several years during the 'eighties. Com- 

 parison of the results, however, showed that the adopted 

 classification was inadequate, and it became necessary to 

 agree on a more complete subdivision of cloud forms. 

 This task proved to be by no means an easy one, but 

 eventually our present international classification of clouds 

 into ten main types was adopted, and some years later, 

 early in 1896, the international cloud atlas, which contains 

 twenty-eight coloured plates illustrative of cloud forms, 

 together with e.\planatory text in three languages, was 

 published. 



At the request of the committee, cloud observations were 

 carried out at a large number of stations during the 

 period from May i, 1896, to the end of 1897. At the 

 more important stations the height and the direction of 

 motion of clouds were determined by means of the photo- 

 grammeter or with theodolites ; at the remainder, direction 

 only was observed with the help of nephoscopes. 



The materials thus accumulated, as well as a large 

 number of trustworthy observations of earlier date, are 

 discussed by Prof. Hildebrandsson in the present report. 

 The method adopted has been to work out, for each region 

 of the earth's surface, the direction of the average monthly 

 drift of the atmosphere at various heights with a 

 " resultantometer " devised by Mr. Sandstrom. The 

 results are set out in tables and diagrams, and in what 

 follows attention will be directed to some of the most 

 important points. 



I. — Tropical Zone. 



Observations at stations near the equator agree in show- 

 ing a drift of the upper atmosphere from some easterly 

 point at all seasons of the year. At Paramaribo (Dutch 

 Guiana, lat. 53° N.), out of 270 observations of upper 

 clouds, only 6 were from south-east and five from north- 

 east. This well marked easterly current in the upper- 

 most regions of the air near the equator was revealed in 

 a most singular manner during the eruption of 

 Krakatoa in 1883. The optical effects produced by the 

 fine dust, which was carried up to great heights, travelled 

 round the earth from east to west in about twelve or 

 thirteen days, indicating an upper east wind moving with 

 the prodigious velocity of 83 miles per hour. 



II. — Trade-wind Zone. 



The generally accepted theory of the origin of the trade 

 winds formulated by Halley and completed by Hadley 

 teaches us to expect upper anti-trade winds from south- 

 west or north-west in the northern and southern hemi- 

 spheres respectively, and this expectation was found to be 

 fully confirmed. At Mauritius, which lies in the centre 

 of the region over which the south-east trade wind pre- 

 vails, the cloud observations show a steady upper wind 

 from the north-west throughout the year. VVe may there- 

 fore assume the existence of an upper wind from the 

 south-west' at corresponding latitudes in the northern hemi- 

 sphere. 



As more temperate regions are approached this south- 

 westerly wind becomes deviated to the right, and at 

 Teneriffe, and still more decidedly at San Fernando and 

 Lisbon, the average drift at the cirrus level is from almost 

 due west. No support is afforded to the assumption made 

 by James Thomson and by Ferrel in their schemes of the 

 general circulation of the atmosphere, that the anti-trade 

 wind continues its course as an upper south-westerly wind 

 until the Arctic regions are reached. 



Special interest attaches to the observations from the 

 region between the upper equatorial east wind and south- 

 westerly or north-westerly anti-trade winds. On the 

 northern side of the equator, at surface level, a broad 

 band on the earth's surface is alternately covered in 

 winter by the north-east trade wind and in summer by 

 the tropical belt of calms. At higher levels a similar 

 alternation is shown. In winter, when the trade wind 

 prevails at the surface, the anti-trade from south-west 

 blows above, but in summer the tropical upper east wind 

 is found above the calm region at the surface. The 

 observations from square No. 39 of the Atlantic Ocean, 

 which is situated in io°-20° lat. N., 2o°-30° long. W., 



NO. 1840, VOL. 71] 



form the most complete example of this alternation ir» 

 Prof. Hildebrandsson 's report; some further very striking 

 instances are to be found in the cloud results for the 

 West Indies recently published by the U.S. Weather 

 Bureau {Monthly Weather Review, vol. xxxii., No. 4, 

 p. 166). 



\U.— India. 



The wind circulation over India is exceedingly complex 

 at the surface, but at higher altitudes a much simpler 

 state of affairs is found to prevail. Prof. Hildebrandsson 

 divides his observations into two groups, those from the 

 north (Lahore to Calcutta) and those from the more 

 central districts between Bombay and Cuttack. He finds 

 that in the former the upper currents blow steadily from 

 the west from December to April, but during the remainder 

 of the year they tend to become easterly. Over Central 

 India the upper westerly wind prevails throughout the 

 year, except in August and September. Since the appear- 

 ance of the report, Sir John Eliot has dealt with the de- 

 tailed cloud observations taken at six Indian stations 

 during the years 1S96-1900 (Indian Meteorological 

 Memoirs, vol. xv., part i.). These show a much steadier 

 upper westerly current in the north. At Simla and Jaipur 

 the average upper wind is westerly throughout the year ; 

 at Lahore and Allahabad an easterly component appears 

 in the averages for August and September only. Further 

 to the south we find an alternation similar to that de- 

 scribed above. At Madras the equatorial upper current 

 from the east prevails during the summer ; in winter the 

 upper currents vary between south and south-west. 



IV. — Temperate Zone. 



Throughout the temperate zone the direction of the 

 average upper currents is from some westerly point all 

 the year round in both hemispheres, though few observ- 

 ations are available from the south of the equator. In 

 Europe and in North America there is thus substantial 

 agreement between the general drift of the atmosphere 

 at all levels, but when we turn to eastern Asia this is not 

 the case. The excellent observations taken at the Observ- 

 atory of Zikawei (Shanghai) show that at the surface and 

 at the level of the lower clouds the prevailing direction 

 is from the north during the winter and from the east, 

 i.e. towards the low pressure system over the continent 

 of Asia, during the summer ; but already at the level of 

 the intermediate clouds, and still more at higher levels, a 

 steady drift from the west is found at all seasons. Similar 

 results are shown by the observations from Japan. 



Though there is substantial agreement in the mean 

 direction of air motion over Europe at all levels, a general 

 tendency for a component from the north to make its 

 influence increasingly felt at higher altitudes is clearly 

 shown. Thus at Upsala, during the winter months, the 

 surface wind is from the south-west ; the lower clouds 

 travel from west-south-west and the intermediate ones 

 from west-north-west, while at the cirrus level the direc- 

 tion of motion is from north-west. Further north, at 

 Nora, in Swedish Lapland, cirrus moves from north-west 

 throughout the year. Some particularly interesting results 

 have been obtained from those of M. Teisserenc de Bort's 

 balloon ascents in which the level of the highest cloud 

 forms was exceeded. In all these cases the balloons were 

 carried towards the south-east, showing that they met 

 with a north-westerly wind in the uppermost layers of 

 the atmosphere. 



North-westerly winds at the cirrus level are also very 

 prominent at Perpignan, Pola (Austria), Tiflis, and Madrid,, 

 stations which lie on the northern side of the tropical belt 

 of high pressure, over which, as we have seen above, the 

 direction of the anti-trade winds has become deviated 

 from south-west to west. 



Prof. Hildebrandsson sums up the results he has arrived 

 at under the following headings : — 



(i) .'\bove the thermal equator and the equatorial calms- 

 there exists throughout the year a current from the east 

 which appears to have a very high velocity at great 

 altitudes. 



(2) Above the trade winds, anti-trade winds from south- 



