Clayton points out the influence of quantity of sun spots on atmospheric pressure and conse- 

 quently on the distribution of precipitation. The fluctuations in level of certain lakes, depending 

 mainly on quantity of precipitation, show a striking concurrence with the changes in sun spots . It 

 is remarkable that Lake Ladoga, for example, has more water at the time of the sunspots minimum 

 than at the maximum, while Lake Victoria in Middle Africa is the reverse. This is due entirely to 

 the different pressure pattern which occurs in connection with these lakes as a result of solar 

 activity. 



Memery goes somewhat further in his investigations. He notes that since 9 sun spot periods 

 equal 100 years and if the quantity of sun spots affects the weather, approximately every 100 years 

 (we note that 100 years almost corresponds to three of the 35-year periods of Bruckner) the weather 

 should repeat itself. Memery confirms this assumption by means of thirteen sharp deviations in 

 seasonal weather from the norm for the period 1888 to 1928 which corresponds to similar seasonal 

 weather deviations for the period 1788 to 1828. In connection with this, in his work which was 

 printed in 1928 Memery warned of severe winters expected in 1929 and 1930, which, as we know, 

 was very strikingly confirmed for the winter of 1929-30. 



On the other hand Memery attempts to explain why, since we have an 11-year sun spot period, 

 we do not have an 11-year weather period. Memery shows that the quantity of sun spots varies 

 extremely irregularly during one single year. It sometimes increases, sometimes decreases, and 

 the yearly maximum sometimes falls in the summer, sometimes in the winter. In 1928 the maxi- 

 mum of solar activity fell in August and this caused positive deviations of temperature in that 

 summer. Whenever the minimum of sun spots falls in the winter we must expect negative devia- 

 tion of temperature in that winter. 



In studying the long term changes in ocean level we discover two factors: first, the average 

 yearly levels of separate parts of the ocean and particularly the levels of individual semi-closed 

 seas differ from each other by more or less considerable amounts. Secondly, the average yearly 

 levels increase or decrease over a large expanse of the shore. This phenomenon is most typical, 

 of course, in the semi-closed seas. 



Thus, for example, along the whole coast of the Baltic Sea, including its bays, the average 

 yearly level was lower than the average long-term level in 1891, 1897, 1901, 1904, 1908, etc. and 

 higher than the average long-term level in 1893, 1899, 1903, etc. 



The latest American investigations show that the increase or decrease of average annual level 

 occurs along the entire coastline of the U.S.A. , both on the Atlantic and Pacific. In all Pacific 

 Ocean ports the increases and decreases in average yearly level fall in the same years. In all 

 Atlantic ports these variations of the average yearly likewise fall on the same years, but these are 

 different from the corresponding years on the Pacific coast. According to Marmer, for example, 

 the highest levels on the Atlantic coast of the U.S.A. fell in 1902, 1910 and 1919. 



L'Allemand and Prevaux's special research on the results of French levelling work shows 

 that the long-term fluctuations in level are periodic and are connected with lunar periods . We may 

 thus consider it certain that the long-term variations in ocean level, in any case in a certain part 

 of it, are connected with long-term variations in the tide-producing forces of the moon and sun. 



But variations in level, especially when occurring simultaneously over a large ocean 

 expanse, are caused by great shiftings of the appropriate water masses. These shiftings are 

 reminiscent in character of the shiftings connected with wind-caused phenomena. The surface 

 water, moving into the shore or into a separate sea, raises the water level. At the same time the 



460 



