NO. 8 HAROMETRIC PRF.SSURF. — DIF.I.L AND DUKLL I 7 



mer and in years with higli sun activity, and also after many ultra- 

 violet invasions, than in winter and in years with low sun activity. 

 Perhaps that is one ot tiic reasons for the fact that an influence of 

 the i)article invasions on sea-level pressure could he dein(jnstrated only 

 for winter and for years with low sun activity. The heatinj; of those 

 layers which ahsorh the i)articles is likewise not inconsiderable, as 

 has heen shown by theoretical considerations and hy computations of 

 H. Petersen (ii). R. M. Deeley (12) regards this heating as a suffi- 

 cient cause for the decreases of sea-level pressure which he observed 

 in Arctic regions during the culmination of solar-activity centers. 



.\n electric polarization of the high atmospheric layers, the prob- 

 ability of which has been stressed by several authors, could be im- 

 portant for several reasons. In the first i)lace electroconvective proc- 

 esses, i.e., "ion winds," could follow such polarization. It has been 

 proved experimentally by V. F. Hess (13) that these ion winds are 

 connected with relatively strong dynamic effects. In the second place, 

 a penetration into the troposphere of the lines of equal force origi- 

 nating in the ionospheric-electric field is possible under certain cir- 

 cumstances (J. Scholz (14)). In that case the colloidal stability of 

 clouds, and therefore the size of droplets and the precipitation 

 tendency, may be influenced (A. Schmauss and A. VVigand (15)). 

 Furthermore, there is a possibility that certain chemical comi>ounds, 

 and consequently condensation nuclei, are produced by electric 

 discharges between the polarized layers. However, such chemical 

 compounds may be produced also during the ordinary bombardment 

 by solar particles of the oxygen-nitrogen mixture, especially in the 

 presence of water vapor or hydrogen. Such particles, e.g., protons, 

 are furnished by the solar particle invasions themselves. This possi- 

 bility of formation of ammonium nitrate and ammonium nitrite-con- 

 densation nuclei by solar particles, especially by electrons, has been 

 emphasized particularly by P. I.enard (16). Industrial processes in 

 the course oi which ammonia is produced by the action of electrons 

 ujMjn a mixture of nitrogen and hydrogen are known ( l'uch-.\nder- 

 sen (17)). 



The numerous observations of a coincidence between the apjx'ar- 

 ance of intensive auroras and the sudden formation of cirrus 

 clouds (11. I-'ritz (18), 11. J. Klein (19), F. Thienemann (20), 

 A. Paulsen (21 ) ) likewise seem to point to the origin of condensation 

 nuclei during i)article invasions. Further supix)rt for that iiyixjthesis 

 was given by G. .\rchenhoId (22), who could demonstrate that there 

 is a certain probability for the geomagnetic character figure being 

 higher on days with sun halos than it would be on ordinary days. 



