Aug. lo, 1876] 



NATURE 



315 



coasts of North America ; and in the south of Europe 

 the oscillation reaches its annual maximum just at the 

 season when the annual minimum occurs near the sea- 

 coasts, even although the general characteristics of the 

 atmosphere be substantially the same in both cases." 



I am not at present aware whether Mr. Buchan has 

 been led by these observations to any definite conclusions 

 as to the physical cause of the variation he so clearly 

 summarises in the passages above quoted. In the part 

 of his memoir which has reached me all theoretical dis- 

 cussion is deferred. But these passages afford such 

 remarkable confirmation of an explanation at which I 

 arrived some weeks since, on approaching the subject 

 from an entirely different quarter, that J do not think it 

 necessary to withhold longer the publication of my view. 

 If Mr. Buchan's conclusions are the same as mine, the 

 facts that I have to bring forward will seem to afford 

 independent confirmation of that view. 



Any person glancing over a series of curves illustrating 

 the diurnal rise and fall of the barometer cannot fail to 

 be struck with the characteristic difference of those of 

 places with a continental and those with an insular 

 climate. The case of the Mediterranean described by 

 Mr. Buchan seems, perhaps, to be an exception ; but, as I 



I. Diurnal oscil'ation of barometer at \.h n Ladakli. 2. Do. Squares 

 N. Atlantic. 



shall presently show, it is an exception of such a kind as 

 most strongly to confirm the rule. The accompanying 

 curves are striking, perhaps extreme, examples of this 

 characteristic difference. The first is that of Leh-in- 

 Ladakh,^ situated in the Indus valley (the observatory 

 being 11,538 feet above the sea), and is for the month of 

 September. The climate is characteristically dry and the 

 summer heat excessive, notwithstanding the elevation. 

 The curve for Yarkand and Kashgar, still further north, 

 and only 4,000 feet above the sea, is of similar character 

 but smaller amplitude. The second curve figured is that 

 for the northern half of square 3, of North Atlantic, pub- 

 lished by the London Meteorological Office. In the 

 former the double oscillation has almost disappeared, the 

 nocturnal fall of pressure being represented by little 

 more than a halt for some hours between two periods of 

 rising pressure ; and nearly the whole fall of the day 

 takes place between 9 a.m. and 5 P.M. In the case of 

 the Atlantic curve the day and night oscillations are 

 almost exactly alike, the night oscillation being only 

 slightly less than that of the day. These characteristic 

 differences are perhaps best expressed by the ratio of 

 the constant coefficients U' and U" in Bessell's interpola- 

 tion formula — 



x=M+ U' sininB-^t u') + U" sin («2 (9 + «") -f- , &c., 

 since the magnitude of U' determines the inequality, and 

 that of £/", though variable under different conditions of 

 climate, is so to a much less extent than the former term, " 

 and chiefly depends on the latitude. The following are 

 the values of U' and U" in English inches, and their 



* This is computed from the hourly observations, recorded during six days, 

 by Capt. E. Trotter, R.E., and of one day by Dr. J. Scully, together with 

 six days' observations by the latter at the hours 4 and to a.m. and p.m. 



ratios for the mean diurnal curves of a few stations 

 (chiefly Asiatic). The arcs «' u" corresponding thereto 

 are also given : — 



As a general rule the more humid the station and the 

 smaller the range of temperature, the smaller is the value 

 of V, and hence it has sometimes been spoken of as the 

 temperature element of the oscillations ; the double oscil- 

 lation which is superimposed on it being referred by 

 Dove, Sabine, and Herschel to the varying tension of 

 water vapour, by Lamont and Broun to some solar influ- 

 ence other than heat ; and by Espy and Kreil to the 

 oscillation of pressure produced by heat in an elastic 

 fluid expanding and contracting under the influence of 

 gravity. To me it seems that there can hardly be a 

 doubt that the last explanation is the true one, and that 

 this has not been generally recognised I attribute to the 

 fact that the consequences of the theory as a purely 

 physical problem have never yet been traced out and 

 verified by such a mass of facts as Mr. Buchan is now 

 bringing together. So long as the whole phenomenon is 

 not satisfactorily accounted for, some doubt may reason- 

 ably attach to the explanation offered of one only of its 

 elements. 



My own attention was first drawn to the subject of the 

 explanation which I am about to give by a paper of Mr. 

 F. Chambers in the P/izV. Trans, for 1873, in which that 

 gentleman showed, as the result of an analysis of the 

 diurnal variation of the winds at Bombay, that one 

 element of this variation is a double rotation of the wind 

 direction of such a character that the southerly com- 

 ponents attain their maximum value at the epoch of the 

 most rapid semi-diurnal rise of pressure, the easterly 

 components at the epoch of maximum, the northerly with 

 the most rapid fall, and the westerly with the epoch of 

 minimum. On these facts Mr. Chambers based a sug- 

 gested explanation of the barometric tides ; regarding 

 them as a phenomenon of static pressure ; and assumed 

 (as now appears, on insufficient grounds) that the pheno- 

 menon in the northern hemisphere is generally of the 

 same type as at Bombay. There was indeed one feature 

 in his explanation, which it seems difficult to reconcile 

 with mechanical laws, since he supposed air to flow from 

 both east and west towards a region where the pressure 

 is rising above the mean, and by its accumulation to 

 produce a maximum of static pressure. But apart from 

 this, the discovery was an important one, and since it 

 clearly showed that a regular horizontal transfer of air 

 corresponded to the oscillations of pressure, it held out a 

 promise that further steps in the same path might clear 

 up what appeared to be anomalous, and possibly lead to 

 a complete explanation of the diurnal oscillation. 



Some time before this paper reached me, the Rev. M. 

 Lafont had placed in my hands four years traces of a 

 Secchi anemograph, erected on St. Xavier's College, 

 Calcutta, and these having been measured off, tabulated, 

 and reduced, I was interested to find that the diurnal 

 wind variation at Calcutta showed the double diurnal 

 oscillation quite as distinctly, and relatively even more 

 prominently than that of Bombay. But one important 

 difference presented itself. The north and south elements 

 of the oscillation, while agreeing in epoch with those of 



