230 BHU. sysrrM teciisical jovrxal 



As an example of tlie iialuie of this approxiination. at a height of 

 about 100 kilometers, we may expect an atmospheric pressure of 

 10~' standard and a corresponding collision frccjuencx' of the order 

 of 10''. Thus for \er\' long \va\es of frequency 40,000 cycles per 



second we still have •- =.4, while at the critical frequency — is only 

 n ' n 



1/100. 



The computation of the collision frequency for electrons is rather 

 invoked because of the peculiar nature which such a collision may 

 ha\c and because it probably is not permissible to assume thermal 

 equilibrium with the molecules of the gas. The processes of ionization 

 and recombination will also lead to complications. Probably the 

 most significant information would be the number of electron free 

 paths per second for unit \'olume. 



The question of the behaxior of waxes in or below the la\er of 

 maximum absorption per ion is a somewhat tlifferent one and belongs 

 projjerly in another paper. 



Kor the case of transmission along a magnetic meridian the oppo- 

 sitely circularly jx)larized rays have the absorption constants '■ 



, _ aN (j}-f/n , _ <rN o)^ / 



Ii will be noted that, at the critical frequency, the first of these waves 

 lias till' lii.uli .ibsorption .y j ' ,■ and is therefore extinguished in a 



short (li^Iall(■c■, wliik' the other \va\e has a normal alisorption const. ml 



aN f 



— ^ • ^. Thus lor the case ol traiisniission along a meridian at the 



8w„- H 



critical frequency we might expect a receiving station, sufiiciently far 

 al)ove the ground, to recei\e a circularly polarized beam. This would 

 mean that if a loop were used for reception, the intensity of the 

 received signal would be independent of the angle of setting of the 

 loop, pro\ided one diameter of the loop was set parallel to the direction 

 of propagation of the wave. In general, of course, this ideal condi- 

 tion could not be realized due to the disturbing action of the ground 

 and of other conducting or refracting bodies and the most we should 

 expect to receive in practice would be an elliptically polarized beam. 

 In the third case, namely, that of propagation perpendicular to 

 the ilirection of the magnetic lield. we find that the wave polarized 

 with its clcciric \ccior parallel to the magnetic field has the same 



