480 



NATURE 



\_March 18, 1S80 



comparison. On the average of the whole period from May, 

 1876, to August, 1S7S, the atmospheric pressure at this level 

 was relatively more excessive than on the plains of Lower 

 Bengal, the mean excess being + 0-0375" at Earjiling, and 

 + 0"O29S" on the Bengal plains, which stretch away from the 

 foot of the Sikkim Himalaya. Moreover it prevailed more 

 steadily. From August, 1876, to August, 1878, or for twenty- 

 five consecutive months, there was not one in which, at the hill 

 station, the pressure did not exceed the average of the month ; 

 whereas on the plains it fell slightly below the average in 



November, 1876, and also in August and November, 1S77. 

 The registers of other hill stations (at least such as are trust- 

 worthy) extend over too short a period to furnish a good 

 average ; but, as far as their evidence goes, it is consistent with 

 that of Darjiling ; and we may therefore draw a second and 

 very important conclusion, viz., tliat the excessive pressure of the 

 two years 1876-1878 in India was mainly, if not entirely due, to 

 the condition of the higher strata of the atmosphere ; to that 



portion, at all events, v*hich lies above 7,000 feet. This is very 

 important, and while it explains the apparent anomaly above 

 adverted to, and which in other cases has been emphatically 

 insisted on by the late John Allan Broun, it points a useful 

 caution against the too frequent habit of arguing from conditions 

 of temperature (as observed at the earth's surface) as if the 

 whole thickness of the atmosphere were affected in the like 

 manner. 



Leaving now, for the moment, the Indo-Malayan region, and 

 turning to other parts of the Europe-Asiatic continent, we find 

 in Western Siberia and European Russia, evidence of a 

 cyclical oscillation of pressure, which is of the opposite 

 character to that already noticed. Of all the stationswhich 

 since 1847 have furnished the registers published in the 

 Annates de I'Obscn atone Central de Russie, Ekaterinen- 

 burg, at the eastern foot of the Ural, is that which exhibits 

 this oscillation in its most salient and regular form. But 

 it is more or less distinctly traceable in the registers 

 of Bogolowsk and Slatoust also in the Ural ; of Barnaul 

 at the northern foot of the Altai, and with considerable 

 intensity but much masked by irregular variations in 

 those of St. Petersburg. Tiflis, however, to the south- 

 west, and Nertchinsk and Pekin to the east, show no. 

 distinct trace of it. Indeed the somewhat interrupted 

 registers of Pekin give a curve which in some respects 

 rather conforms to the Indo-Malayan type. The ac- 

 companying figures 1 to 6 represent the curves of the 

 annual deviation of the mean pressure at the above six 

 stations, up to 1877, and over that of Ekaterinenburg 

 I give a dotted curve showing the variation of the sun- 

 spots. In point of amplitude the oscillation at Ekater- 

 inenburg and St. Petersburg greatly exceeds that if the 

 opposite type in the Indo-Malayan region, as indeed 

 might be expected if these oscillations are reciprocally 

 compensating, and the tropical type prevails over a 

 larger area than the Siberian. 



\Ve are thus led to the further conclusion that between 

 Russia and Western Siberia on the one hand, and the 

 Indo-Malayan region [perhaps including the Chinese 

 region) on the other, there is a reciprocating and cyclical 

 oscillation of atmospheric pressure ; of such a character 

 that the pressure is at a maximum in Western Siberia 

 and Russia about the epoch of maximum sun-spots, 

 and in the Indo-Malayan area at that of minimum sun- 

 spots. 



In tabulating the variations of the barometric means 

 of Ekaterinenburg month by month, I was much struck 

 with the greater magnitude of the anomalous deviations 

 of the winter as compared with those of the summer 

 months ; in other words with the apparent greater 

 variability of pressure during the winter season. In 

 order to verify this feature and to obtain a measure 

 of the variability, I took the mean of the deviation values 

 of each month for the whole series of (31) years, without 

 regard to algebraical sign, and dividing by 2 obtained the results 

 given in the first figure column of the following table. The 

 registers of St. Petersburg, Barnaul, Greenwich, Adelaide, and 

 Melbourne similarly treated gave the results shown in the live 

 subsequent columns, and those of Calcutta and the Mauritius the 

 figures of the two final columns. 



I am not aware whether the climatic features exhibited by this I able significance. At Calcutta the variability of pressure is 

 table have before been noticed, but they seem to have consider- | nearly the same at all seasons of the year, while at the Mauritius 



