March iS, 1880] 



NATURE 



481 



apparently on the comparative frequency of cyclones ; 

 but at all the extra tropical stations it is from two to three times 

 a; great in the whiter months as in the summer, and especially 

 S3 in the two months December and January. Hence (excluding 

 the case of the Mauritius) the less direct the action of the ?un, 

 the greater the vicissitudes of the atmospheric pressure. 



The question now presents itself, " How far is that cyclical 

 variation of pressure which conforms to the sun-spot period, 

 dependent on the variation of the summer and winter \ ressures 

 respectively?" It is obvious that the reply to this question 

 must have a very important bearing in indicating the physical 

 cause of the oscillations. If it be essentially a phenomenon of 

 the summer mouths, we may be justified in regarding it as a 

 possible effect of the more or less direct action of the >un on 

 the continental land surface; but if it be solely or even mainly 

 dependent on the winter variations, no such explanation is 



admissible. It mtut, then, rather be regarded as an effect 

 produced under negative conditions to compensate an opposite 

 effect which is due to the direct action of the sun elsewhere. 



lain this I have taken separately the means of the 

 four months November to February, and May to August, for 

 the stations St. Petersburg, Ekaterinenburg, and Barnaul, in 

 each successive pair of yean or year ; and since the figures thus 

 obtained showed irregularities which somewhat masked the 

 periodical variation, I smoothed the original values by substi- 

 tuting the mean of three consecutive summers or winters for the 

 original mean forming the middle term of the triad. I have also 

 done the same for the tropical stati us Singapore, B tavia, and 

 , and give the results in the following tables. The 

 smoothed means are illustrated by the six pairs of curves in Fig. 2, 

 the summer and w inter curves being drawn to the same horizontal 

 lines of reference, the former dotted, the latter continuous. 



Russian Stations 



Indo-Ma lay an S tat 1 o n s 



In thee tables and figures the true nature of the oscillation is j seasons of the year show an oscillation of the same kind and 

 sufficiently obvious. At Singapore, nearly or. the equator, both nearly of the same amplitude. At the other intertropical 



