170 



ANNALS NEW YORK ACADEMY OF SCIENCES 



Since then many researches have been made. A recent publication by 

 W. Koeppen^^ gives the most convincing data. Koeppen utilized the 

 annual temperature means collected by J. Mielke,^^ as well as the figures 

 which served him in previous publications on this subject, and completed 

 the tables in order to have the results of practically all available obser- 

 vations made, all over the world, from 1811 to 1910. The final result 

 may be advantageously reproduced in a form slightly different from that 

 given by Koeppen.'^^ 



The following figures represent the mean departures from the general 

 means for the years of sun-spot maxima and of those before and after 

 these years : 



—3 



+0.09 



—2 ! —1 



—0.03 —0.15 



O Max. 



—0.23 



+1 



+2 



+3 years 



—0.15 ! —0.15 I +0.07 °C. 



The means of the years of sun-spot minima, as well as the means of 

 the years preceding and following these minima, are given below : 



—3 



—0.11 



+0.07 \ +0.24 



O Min. 



+1 



-hO.22 I +0.11 



+2 



+0.04 



+3 years 

 0.00 °C. 



The conclusion is that the solar cycle of about 11 years' mean duration 

 has a well pronounced influence upon atmospheric temperature, and that 

 the difference between maxima and minima equals on the average nearly 

 0.5 °C. 



But the diagram representing the succession of the preceding figures 

 graphically shows that the 11-year period is complicated by the existence 

 of a shorter variation of less importance. It seems that during the sun- 

 spot cycle, temperature undergoes a double oscillation; that between the 

 consecutive temperature minima, corresponding to sun-spot maxima, 

 there are two minima and three maxima, one of which corresponds to the 

 sun-spot minimum. The diagram given by Koeppen for the Eussian 

 temperature data is perhaps the best illustration of this fact. 



^^ W. Koeppen : "Lufttemperaturen, Sonnenflecke und Vulkanausbriiclie," Meteor. 

 Zeitsch., vol. 31, p. 305. 1914. 



^ Johannes Mielke : "Die Temperaturschwankungen 1870-1910 in ihrem Verhaltnis 

 zu der 11 jahrigen Sonnenfleckenperiode," Aus dem Arcliiv der deutschen Seewarte, vol. 

 36, no. 3. 1913. 



^■^ Koeppen adds the annual temperature departures of tlie years of sun-spot minimum, 

 one year after, two a. s. f. This gives him 11 columns of figures. The disadvantage of 

 this metliod is that the year of sun-spot maximum occurs once in column 4, 3 times in 

 column 5, 4 times in column 6 and once in column 9. His first column of figures com- 

 prises the departures for the years 1811, 1822, 1833, 1844, 1855, 1867, 1878, 1889 and 

 1900. Only four of these years are years of sun-spot minima; the others precede or 

 follow such a year. However it would have been difficult to make a better adjustment 

 than Koeppen did. 



