174 ANNALS NEW YORK ACADEMY OF SCIENCES 



SLsk, therefore, if the amplitudes of the barometric oscillations vary in 

 harmony with the pleionian cycles. 



Conclusions 



1). The hypothesis attributing most climatic variations to volcanic 

 dust veils, although verified in some cases^ presents more a purely theoret- 

 ical than practical interest. 



2). The dust veil produced by the Krakatoa eruption affected atmos- 

 pheric temperature very greatly. The violent volcanic eruptions of 1902, 

 as well as the Katmai eruption of 1912, influenced the yearly mean tem- 

 peratures but very slightly or not at all. 



3). The pleionian variations of temperature have nothing in common 

 with the presence or absence of volcanic dust veils. 



4) . An influence of the sun-spot variation upon the changes of atmos- 

 pheric temperature is undeniable. 



5). In some exceptional cases, at far-distant stations, such close corre- 

 spondences of the pleionian variations may be observed that one is justi- 

 fied in admitting that missing crests are due to locally restricted anom- 

 alies. Of course, in most cases the anomalies are prevalent. 



6). A correlation between pleionian and macropleionian temperature 

 variations and the occurrence and frequency of volcanic eruptions may 

 be presumed, although, at present, it is difficult to imagine how such a 

 correlation could be explained. 



