458 BELL SYSTEM TECHNICAL JOURNAL 



every investigator when taking account of either influence simply 

 ignores the other. On the whole, the mathematical developments 

 have far outrun the data. Yet there are important connections 

 between experiment and theory, including for instance the proof that 

 the ions which principally reflect the signals are free electrons. 



After the theory come what I will call, for contrast, the speculations. 

 The analysis of the ionosphere being made and accepted, a host of 

 questions arise. Must we assume additional agents of ionization, 

 other than the ultra-violet light of the sun? The answer to this 

 question being certainly "yes," one must inquire how to distinguish 

 that part of the ionization which is due to sunlight from the rest, 

 and what are the causes of the rest. Why the distinctive distribution- 

 in-height of the ions, amounting to what is called "the stratification 

 of the ionosphere"? What assumptions must we make about the 

 composition of the atmosphere in its dependence on height? or (as 

 the question is more commonly put) what information can we derive 

 about the composition of the atmosphere? How far can we go in 

 interpreting the fluctuations of terrestrial magnetism, and (as later 

 will be apparent) in mapping out the earth's magnetic field? The 

 possible questions even rise to the realm of astronomy, and the sug- 

 gested answers form a part of the theory of the sun as a potent source 

 of radiations of all kinds, luminous and electrical and material. The 

 implications of the ionosphere seem to be almost limitless, but a 

 severe limit will nevertheless be set by space and time upon this 

 article. 



Methods of Experiment, and a Simplified Picture of the 

 Ionosphere Adduced for Illustrating Them 



The ionosphere is a canopy of ions overarching the earth, and in 

 Fig. 1 it is represented by a model, very simplified indeed and yet 

 instructive. Here it is shown as consisting of two "layers" marked 

 E and F, with an ion-density which is uniform in each, and greater 

 in F than in E. It is time to become familiar with the symbol N 

 used for number of ions per unit volume: this picture shows N having 

 the constant values Ne and Np ( > Ne) in £-layer and F-layer respec- 

 tively, and the value zero between. 



The lines which are broken at the layer-edges are paths of wireless 

 signals or waves sent out from the source at S — sent out obliquely, 

 for transmission over long distances. There is a path reflected from 

 E, a path reflected from F and a path which penetrates both. These 

 correspond to relatively low, medium, and high frequencies respec- 

 tively: as examples I will give the values 1, 10 and 100 mc. (megacycles, 



