iO 



NA TURE 



\Nov. 19, 1885 



through 320 degrees of longitude. The principal low 

 centre, 2g'2oo inches, was north of the Caspian Sea ; a 

 second low centre, 29'4oo inches, was over the northern 

 part of India ; a third low centre, 29600 inches, over the 

 (}ulf of St. Lawrence ; a fourth low centre, 29'8oo inches, 

 over China; a fifth low centre, 29'8oo inches, north-east 

 of Japan; and if every part of this large portion of the 

 earth's surface had been sufficiently represented by 

 observing stations several other subordinate low centres 

 would doubtless have been exhibited. On the other hand, 

 a centre of high pressure, 30-400 inches, was found over 

 the Atlantic Ocean ; a second, 30-200 inches, over the 

 south-eastern part of the United States ; and a third, 

 over the eastern part of the Pacific near latitude 30° N. 

 The area of high pressure formed a belt closely following 

 the parallels of 30" — 35' and extending through at least 

 240 degrees of longitude, but interrupted by the Asiatic 

 Continent. 



We drew attention five years ago to the all-important 

 bearings of these areas of high and low pressure on the 

 weather in all the regions of the globe over which anom- 

 alously high arid low barorneters at any time prevail 

 (Nature, vols, xxi., xxii. and xxiii.). But the importance of 

 this department of meteorology is much enhanced when it 

 is considered that it is through a careful record of the ap- 

 pearance and disappearance in dift'erent regions of the 

 globe of these cyclonic and anti-cyclonic areas and an 

 investigation of the causes determining their form, position, 

 and intensity from time to time that we may hope to 

 reach the solution of the problem of the weather. In 

 prosecuting this large inquiry, the results of Prof. Loomis's 

 careful measurements of meteorological phenomena, as 

 detailed in the revised edition of his "Contributions" 

 now before us, form one of the best guides we at present 

 possess. 



Directioiof Mo'i'ctucnt pf Areas pfLo7u Pressure. — Areas 

 of low pressure,or cyclones.seldom remain stationaryin the 

 same position for many hours. The centre of low pressure 

 generally changes its position steadily from hour to hour, 

 and everywhere there is observed a marked uniformity in 

 the direction of this movement. Prof. Loomis gives 

 several charts showing the progressive movement of 

 cyclones in different parts of the world, including one 

 showing nearly all the different storm tracts delineated 

 on the International Weather Maps of the United States 

 Signal Service for a period of more than four years. 

 Maury's Storm Charts are also brought under review. 

 The lowest latitude reached by the centre of any cyclone, 

 which has been distinctly traced, is 6^ I' N., and there are 

 only eight cases of cyclones whose paths have been traced 

 to points south of lat. 10° N. 



Observations indicate that, both in the Pacific and 

 Atlantic, gales are of extremely rare occurrence within 

 six degrees of the equator, and, when they do occur, the 

 barometric depression is small, and the cyclonic character 

 of the winds indistinctly marked. But in low latitudes, a 

 little higher than six degrees, gales are more freciuent 

 over the Pacific than over the Atlantic Ocean. 



Tropical storms which are found to pursue a westerly 

 course are limited to two regions of the globe — viz. the 

 Atlantic Ocean, but particularly its western portion, near 

 the West India Islands, and the region south of the 

 continent of Asia. As regards the Pacific, no cyclone has 



ever been observed, except near Asia or its outlying 

 islands. 



As regards the tracks of tropical cyclones in the neigh- 

 bourhood of the West Indies, the teaching of the data 

 represented on the International Charts is that nearly all 

 the areas of low barometer which occur within the tropics 

 and advance westwards, instead of following the ordinary 

 course of the trade winds, advance in a direction some- 

 what north of west. Of these West Indian cyclones, 08 

 per cent occurred in August, September, and October, 

 thus leaving only 12 per cent, for the other nine months 

 of the year. On the other hand, of the Asiatic cyclones 

 52 per cent, occurred in September, October, and 

 November, and 43 per cent, in April, May, and June, thus 

 leaving only 5 per cent, for the other six months. There 

 is, therefore, a marked seasonal difference as to the 

 frequency of the tropical cyclones of the Atlantic as com- 

 pared with the Pacific : in the Atlantic they are almost 

 exclusively confined to the autumn, but in the Pacific 

 they are nearly as frequent in spring as m autumn. 



The average direction of the course of the Asiatic 

 cyclones, while moving westward, is 38° north of west, 

 which closely accords with that found for West Indian 

 cyclones. But, as regards the onward progress of tropical 

 cyclones, whdst Asiatic cyclones advance westwards at 

 the average rate of 8 English miles per hour, the average 

 velocity of West Indian cyclones is double that amount. 

 Asiatic cyclones come around to a due north course about 

 lat. I9°-S N., but West Indian cyclones do not assume 

 a due northeriy course till, on the average of instances, 

 lat. 30° N. is reached. In the Pacific the average course 

 of cyclones, after turning eastward, was 35° E. of N., 

 and their velocity was 9-8 miles, which is scarcely half of 

 the velocity of the West Indian cyclones. These striking 

 and vital differences between the tropical cyclones of the 

 Atlantic and the Pacific will doubtless play no unimport- 

 ant part in the development of the theory of the cyclone. 

 An examination of Prof Loomis's chart of storm- 

 tracks for the northern hemisphere, with wind charts 

 indicating the prevailing direction of the wind, shows a 

 remarkable correspondence between the two classes of 

 facts. Examining the point more narrowly. Prof Loomis 

 finds that for the middle region of the Atlantic, near lat. 

 50°, the average direction of storm paths corresponds very 

 closely with that of the average direction of the wind ; 

 but in the western part of the Atlantic the average course 

 of storms is considerably more northerly than that of the 

 wind, while in the eastern part it is more southerly. 

 These results, which fairly accord with those derived 

 from tropical storms, seem to indicate, in the opinion of 

 the author, that in the middle latitudes of the northern 

 hemisphere the direction of progress of storm-centres is 

 not the same as that of the average wind, but is sensibly 

 affected by some other causes ; and that the results de- 

 rived from observations in the China Sea indicate that 

 one of the causes is the prevalent direction of the wind 

 which immediately follows a storm. The subject is 

 further prosecuted by an examination of the prevailing 

 winds and storm-tracks during the three winter months 

 for the ten winters ending 18S2 of that portion of the 

 United States between long. 90° W. and the Rocky 

 Mountains. The result of this somewhat exhaustive 

 comparison is similar to that derived from the observa- 



