March 23, 191 1] 



NATURE 



I II 



been put up to protect some sheep and lambs from the 

 wind, when they noticed a fox come scarcliing along the 

 hedgerows. They tcept perfectly still and watched, and, 

 when he got nearer, they saw that he \\'as collecting the 

 cheep's wool caught on the thorns and brambles. When 

 he had gathered a large bunch he went down to a pool at 

 Xhe junction of two streams, and, turning round, backed 

 slowly brush first into the water, until he was all sub- 

 merged except his nose and the bunch of wool, which he 

 held in his mouth. He remained thus for a short time, 

 and then let go of the wool, which tloated away ; then he 

 came out, shook himself, and ran off. 



Much astonished at this strange proceeding, they took 

 a shepherd's crook, went down to the water's edge, and 

 pulled the wool out. They found that it was full of fleas, 

 which, to save themselves from drowning, had crept up 

 and up the fox's brush and body and head and into the 

 ivool, and that thu.s the wily fox had got rid of them. 



Cambridge, March 20. T. McKenxy Hughes. 



THE CIRCULATION OF AIR IN THE 

 SOUTHERN HEMISPHERES 



IN this investigation of the circulation of the atmo- 

 sphere in the southern hemisphere, the author 

 has taken a new course. Instead of proceeding in 

 the usual way from tables of wind-direction and 

 foice, he has taken as the groundwork of his re- 

 searches the atmospheric whirls themselves. He does 

 not deal with cyclonic systems, as one might at first 

 suppose, but with the anticyclonic, the travelling high- 

 pressure systems. The reason of this is plainly due to 

 his previous work, "A Discussion of Australian 

 Meteorology " (London, 1909). xVfter a four-year period 

 in the variations of air-pressure over India, South 

 Africa, and South America, and their relations to the 

 four-year cycle in the solar variations had been success- 

 fullv demonstrated, it was necessary to investigate the 

 weather conditions in Australia with that object. In 

 the subtropical continents of the southern hemisphere 

 weather conditions are chiefly influenced by baro- 

 metric maxima almost constantly travelling from west 

 to east. This was first shown to be so for Australia 

 bv the astronomer H. C. Russell, of Sydney, to whoin 

 the meteorology of that continent is so much indebted. 



Russell already held the opinion that these travel- 

 ling b.'u-ontctric maxima (with the V-shaped depres- 

 sions acconi])anying them on their south side) do not 

 originate on the continent itself, but approach from 

 the South Indian Ocean. In Dr. Lockyer's extensive 

 work, above quoted, it was shown more conclusively 

 that in the Australian area, in latitudes 20° to 40° S., 

 anticyclonic systems travel with great velocitv from 

 west to east, and that this also holds good for South 

 America, South Africa, and Mauritius, in the same 

 belt of latitude. In all i)robal)ilitv, what holds good 

 for 130° of longitude would also obtain for the rest 

 of the earth's circumference. A proof of this would 

 be of great importance for the weather prediction of 

 these southern continents. The inquiry was there- 

 Tore extended over the whole southern hemisphere, in 

 order to obtain at the same time a more secure basis 

 for the dctcnniiKition of the cITects of the solru' varia- 

 tions (111 the circulation of the air of the soiitltern 

 hemivplicrc. 



Till' collection of the inaicriais for this widcK' ex- 

 tended investigation naturallv gnvc ihe .-uithor mtich 

 tiouble and difliculty. The dcierniiiiation of the 

 T^mplitudes of the waves of aiinosi)heric ])ressure over 

 ♦:he whole of the district in ciurstiou formed the pre- 

 liminary part of the work in vi( w . The author 



_ 1 Solar Physics Committee. Souttiprn Hein'<iiln;re Surfacc-.iiv t'ircula- 

 lion : I'einn; a Study of the Mean Moiiilih I'rf^^iiic Amplitudes, the Track'; 

 of the Anticyclones and Cyclonr^, .lu.l tin Metooro'oglcal Recorils of 

 ■Sf'veral Antarctic Expeditions. I!v I'r W. I. S TrH•L•^•,■r, under tlie 

 dirt-ctlon of .Sir Norman Lockyer. K.( I'... K.K S. ' xv plates. 



(London: H.M.S.O., Wyman nii.l Sm,,, It!. I Oliver and 



Boyd. Dublin: E. Ponsonhy, Ltd., luio.) I'licc l.,. 



NO. 2160, VOL. 86] 



rightly confined himself to the southern winter half- 

 year (April to September). It is quite clear that in 

 calculating the mean height of the pressure waves, 

 all waves, including even the smallest, cannot be 

 taken into account, but only those of a certain mag- 

 nitude. The author finds the amplitude of the pressure 

 wave (Schwellenwerth) for these by selecting the 

 three greatest wave heights for each station and takes 

 one-fifth of the mean as the lower limit. This value 

 (Schwellenwerth) is naturally difl'erent for different 

 places in the various latitudes. 



Dr. Lockyer calculates in this way the mean 

 heights of the waves of air-pressure for fifty-five 

 places in the southern hemisphere, between the equator 

 and the Antarctic continent, and enters the values in 

 the chart. That leads further to the drawing of lines 

 of equal wave heights of oscillations of air-pressure. 

 The author denotes these lines by the somewhat 

 mysteriously sounding Greek compound " Isanakata- 

 bars " : lines of equal up and down movements of air- 

 pressure. The mean amplitudes of the waves of air- 

 pressure naturally increase from the tropics towards 

 higher latitudes. In latitude 0° to 12° S. they reach 

 I to 2 mm. ; from 12° S. they increase very rapidly 

 and attain a maximum of 18 to 19 mm. in 53° to 60° 

 S., and then decrease again to 14 to 15 mm. in South 

 Victoria Land. The Isanakatabar of 16 mm. occa- 

 sionally fringes the Antarctic continent. The increase 

 of the wave heights towards the south is explained by 

 the fact that from the belt of the travelling barometric 

 maxima, with still relatively small amplitudes, we 

 first enter the region of V-shaped depressions which 

 accompany them, and then, finally, that of the large 

 cyclones of higher latitudes, the mean tracks of which 

 may probablv be taken as between 55° and 60° S. 

 At the southern limit of these, towards the permanent 

 Antarctic anticyclone, the amplitudes again decrease. 

 But, generally speaking, the Isanakatabars run fairly 

 parallel to the parallels of latitude. They exhibit, 

 however, the peculiarity that on the mountain ranges 

 of the west sides of South Africa and South America 

 they trend downwards in higher latitudes, but leave the 

 east coasts in lower latitudes. This may be ascribed 

 to the westerly ranges of mountains in these con- 

 tinents. 



These Isanakatabars form the starting jwint of 

 further verv interesting deductions by the author. 



It may here be remarked that Kiimtz, in his " Lehr- 

 buch der Meteorologie " (vol. ii., p. 339), has en- 

 deavoured to draw lines of equal non-periodical oscil- 

 lations of air-i)r( ssure. ]\c calculated for numerous 

 stations of the northern hemisphere the mean value 

 of the luouthly variaiiou of air-pre>sure, and called 

 his lines based thereon somewhat improperly " i so- 

 barometric " lines.- It is certainly noteworthy that 

 lines of equal barometric variation were drawn (^1832) 

 long before it was thought of constructing li!ic> (^f 

 equal air-jjrt'ssure (iNohars). Thoe were tirsi drawn 

 bv- Reriou (iS()4), ;iiid then ])articidarl\- hv Huchan 

 (iSt)!)). KiinU/ also reniariced that his lines did not 

 run wliollv with tlie |)arallels of latitude, but_ tliat, 

 (•.<,'. the liiie of S par. Iines=i8 mm., is met witli on 

 the e;isi co;ist <if the InitcHl States in 3(>^ N. latitude, 

 but ill wesiern luirope in 42^. .\t a nuich later ptM'iod 

 I'^ehlherg and Koppen again investigated the variations 

 (-if air-prcsstire 1 »n a much broader basis, hut also for 

 the interval of a month (Aus dfin Arrhiv d. Ih'iitsrlicn 

 Scc7i.UTrfi\ 1S7S, and M rlrnitdi'^isrlic Zrilsrliril't , 

 iSX^). Theve Di.tullilv h:ironirter \ariations ai'e natur- 

 allv a nuuli r<Mig!ier nie,i-.iii-e of ilie irregular 

 variations of iire^sure lli.in ilie mean heiglvi of the 

 indi\-idii;il pre^.^me waxes calcnliied 1)\- Dr. Lockyer. 

 Kiipiien has alre,id\- rein;irked th.it the lines of equal 

 \-ariations of .-lii'-pressiire should be in relalion with 

 the direction of (lie tracks of the l);irometer minima. 



