9i8 



NA TORE 



[December 22, 1923 



Variations in the Level of Lake George, 

 Australia. 



ON May 18, 1876. a Jcttcr appeared m Nature 

 from Canon R. Abbay on the subject of the 

 changes in level of I^kc George, in the south-cast of 

 New South Wales, which in the past hundred years 

 has varied from a small swamp to a depth of 25 feet 

 or more. We have now received from Canon Abbay 

 a letter and a diagram showing the variations of level 

 in the lake from 1817 to 1918. The latter, which is 

 reprotluced in Fig. i, was drawn from information com- 

 piled by the late Mr. H. C. Russell, Government 

 Astronomer of New South Wales, up to 1904, and 

 since that date by the Commonwealth Meteorological 

 Bureau. It also shows the " residual rainfall curves " 

 for Goulbum, the nearest station, and. for Sydney, 

 150 miles distant. A residual rainfall curve is 

 obtained by finding the difference of rainfall for 

 each year from the average for the whole period 

 and adding up the differences for successive years, 

 so that the figure plotted for any year represents 

 the total excess or deficit of rainfall from the beginning 

 of observations until that year. The curves show, 

 as Canon Abbay points out, that while rainfall is 



L (Actual level, inches from base) ■sO'36 R +0'36 1 



The years in which the lake was dry have 

 otnittecf from the calculations. The results confin 

 those obtained from I^kc Victoria, that variatUHi 

 of evaporation are prob r^ ' .<• important tJha 



rainfall variations in dct- the level of 



and that the rate of c^.ijnM.uion is appreciaU 

 greater when sunspots are lew than when they 

 numerous. 



The diagrani shows that the rainfall at Goulbui 

 agrees fairly closely with that at Sydney, but xi 

 rainfall at several stations over the lake basin 

 been available for a long period, there is no doub 

 that the correlation with their average would ha\ 

 been appreciably higher than that with Sydne 

 rainfall. It also appears that the evaporation i 

 Lake George is not determined by sunspots to 

 same extent as that at Lake V'^ictoria. As Canon Abba 

 points out, the frequency of west and nortb-w« 

 winds would be of great importance in this cor 

 nexion, and this would be governed by the pressui 

 gradient between, say, Melbourne and Brisbane. 



In 1875 Canon Abbay thought that the rise in thi 

 lake, which had proceeded fairly steadily from abou 

 1849 until that date, was connected with the do 

 struction of " bush," allowing the rni"^ ■" ' 



Kit;. I. — Fluctuation of water in Lake George, N.S.W. (I,at. S. 35° 4' ; Ixjng. E. 149° 23'). The fluctuations of the water in the lake as shown in black , 

 from 1817 to 1904 were compiled from various sources by .Mr. Russell, and from 1904 to 1918 by the Commonwealth M eteorokigkal Bureau. 



Residual mass curve of Sydney rainfall Mean rainfalV at .Sydney for 76 years, 48*35 in. 



Residual mass curve of Goulburn rainfall Mean rainfall at Goulbum for 60 years, 25-34 in. 



evidently an important factor in the level of the lake, 

 there must also be other influences at work. 



The lake is without outlet, and we may accordingly 

 regard its changes of level as determined by the 

 balance between the rainfall and evaporation in its 

 basin, the loss by seepage probably being negligible. 

 As a measure of rainfall the long series of observa- 

 tions at Sydney has been employed from the com- 

 mencement of the official observations in 1840. The 

 question of evap>oration is more difficult, but it has 

 recently been found that in the Central African 

 lakes, Victoria and Albert, the amount of evapora- 

 tion bears a very close inverse relationship to the 

 number of sunspots, the correlation coefficient 

 between lake level (Lake Victoria) and sunspot 

 number, after elimination of rainfall, being as high 

 as + 0-90, and much higher than the correlation witii 

 the average rainfall in the basin. In the case of 

 Lake George, a few years of heavy rainfall and slight 

 evaporation result in a considerable rise of level, 

 and if they were succeeded by a series of dry hot 

 years the lake level would fall gradually until it was 

 dry or until another wet period supervened. It was 

 accordingly found best to correlate the changes in 

 level between the beginning and the end of each 

 year (L^) with Sydney rainfall (R), and average 

 sunspot number (S), during the same year, and the 

 following results were obtained : 



Correlation between change of level and rainfall ; 

 influence of sunsp>ots eliminated : r = -i-o-35. 



Correlation between change of level and sunspots ; 

 influence of rainfall eliminated : r = + 0-39. 



The regression equation is : 



L' (in inches) = 0-96 R (in inches) +0-43 S (Wolf's No.). 



The corresponding equation for Lake Victoria, 

 Central Africa, was : 



NO. 2825, VOL. 1 12] 



into the basin with little loss, but the subsequen 

 fall in level showed that this could not be the c^ 

 since the destruction of the bush continued -' 

 the level of the lake was falling. The nineteen 

 year periodicity which has been advocated in con- 

 nexion with Australian weather occurs in the lake 

 levels, though not very definitely, and there is also 

 an eleven-year periodicity connected with the sun- 

 spot effect. The two chief maxima in the level, 

 about 182 1 and 1875, and the two chief p>eric>ds when 

 the lake was dry, about 1848 and 1905, are separated 

 by intervals of 54 to 57 years, and may represent a 

 quasi-periodicity of about 56 years caused by the 

 interference of these two periodicities ; but weather 

 cycles are treacherous things, and it would not be 

 safe to base a forecast on them. C. E. P. B. 



Geological Progress in India. 



T T is satisfactor\' to notice that, in spite of financial 

 *- stress in India, the Government has continued to 

 add to the staff of the Geological Survey, which, with 

 the recruits recently selected, now includes 26 out of 

 the sanctioned 30 officers of the senior grade. The 

 progress of work also during the last few years, since 

 the return to normal duties of those officers who were 

 on active service, has resulted in an approach to 

 completion of many lines of work that had been for 

 some time necessarily left indefinite. Among these 

 the classification of the Tertiary beds of Burma, and 

 their correlation with the Tertiaries of Western India 

 and the standard stratigraphical scale of Europe, are 

 now showing distinct signs of stability. 



The untimely death of Mr. E. Vredenburg (Nature, 

 April 14, p. 505) prevents the completion of the hea\-y 

 task of summarising the palaeontological results, but 



