370 BELL SYSTEM TECHNICAL JOURNAL 



ments at rates as high as 20 miles per minute. Other experiments 

 and curves are mentioned which show absorption to be one of the 

 important factors causing poor daylight transmission for wave-lengths 

 around 214 meters. A discussion is given to show that both electro- 

 magnetic waves from the sun and jS particles must be assumed to 

 produce ionization to explain radio transmission phenomena observed. 

 The ionization is pictured as beginning at an altitude of about 16 

 miles and extending upward, and as experiencing diurnal and seasonal 

 variations. The electromagnetic or day ionization occupies a wide 

 region, and is fairly steady except for the diurnal variation. The 

 /3 particle ionization which is the principal ionization at night occurs 

 continuously. It is, however, less dense than the other ionization 

 and is very variable. 



Diffraction of Electrons by a Crystal of Nickel.^ C. Davisson and 

 L. H. Germer. The scattering of electrons by nickel has been reported 

 on recurrently since 1921. This paper gives the latest results which 

 indicate that a wave-length is in some way connected with the 

 electron's behavior. A rather complete summary of the experiments 

 and conclusions appeared in the Bell System Technical Journal for 

 January, 1928, which makes unnecessary a fuller account here. 



A General Operational Analysis J W. O. Pennell. The paper 

 outlines the elements of a general operational analysis. Using p as 

 the symbol for an operator, the author, starting with a general typical 

 defining equation such as p{ax^) = \l/{a, x, h), proceeds by definitions 

 and theorems to demonstrate the use of operational methods in a 

 large variety of common mathematical operations. 



Precision Determination of Frequency.^ J. W. Horton and W. A. 

 Marrison. The relations between frequency and time are such that 

 it is desirable to refer them to a common standard. Reference 

 standards, both of time and of frequency, are characterized by the 

 requirement that their rates shall be so constant that the total number 

 of variations executed in a time of known duration may be taken as 

 a measure of the rate over shorter intervals of time. Frequency 

 standards have the further requirement that the form of their vari- 

 ations and the order of magnitude of their rates shall be suitable for 

 comparison with the waves used in electrical communication. 



Two different types of standard which meet these requirements are 



^Phys. Rev., Vol. 30, No. 6, pp. 705-740, Dec. 1927. 



7 //. of Math, and Phys. of M. I. T., Vol. 7, No. 1, pp. 24-38, Nov. 1927. 



8 Proceedings of the L R. E., Vol. 16, No. 2, Feb. 1928. 



