REVIEWS 697 



or Pearson's coefficient, since it was first given by Bravais in 1846, and in 

 1896 was shown by Prof. Pearson to be the best value of the correlation 

 coefficient. Prior to 1896, Galton used average deviations in the array from 

 their means. This average deviation, when divided by the probable error, 

 was termed by Prof. Weldon Gallon's function. 



It seems rather a mistake to use the symbol v for the probable error, when 

 Y at present always signifies the correlation coefficient. 



Ethel M. Elderton. 



Textbook on Wireless Telegraphy. By Rupert Stanley, B.A., M.I.E.E. 

 [Two vols. : Vol. I, pp. xi + 471, with 253 figures ; Vol. II, pp. viii 

 + 357> with 227 figures.] (London : Longmans, Green & Co., 1919. 

 Price 15s. per volume.) 



This greatly enlarged edition of Major Stanley's well-known book is a welcome 

 addition to the literature of wireless telegraphy. He starts from the electron 

 theory of matter and develops his elementary section, dealing with the flow 

 of current, on this basis. This, in itself, is an interesting experiment, and 

 the experiment seems justified by the lucidity of the treatise. The first 

 volume is very similar to the original book that Major Stanley published. 

 It has been improved and minor corrections made. The chief interest of the 

 new edition, however, lies in the second volume. 



The author was in touch with all wireless developments during the war, 

 and for the last four years was chief instructor in wireless to the British 

 Expeditionary Force in France. He has thus been able to keep abreast of all 

 the most recent developments, and the description he gives of the various 

 types of valves, and of their circuits, is of the highest value. 



The second volume opens with a chapter on electrons, in which the theories 

 of current flow and electrostatics are expressed in terms of the electron unit. 

 The second chapter deals with the theory of the hard valve. It starts by 

 explaining the action of the original Fleming valve and develops the theory 

 of the action of the three-electrode valve. Characteristics are given, and par- 

 ticular attention is paid to the well-known French valve, which has been used 

 so largely by all countries during the war. 



In chapter iii, the subject of valve detectors and detector relays is dis- 

 cussed, and one is glad to note that due justice has been done to Prof. 

 Fleming for the valuable pioneer work that he did on the valve. The author 

 states, " Fleming was the first to use the unilateral conductivity effect of an 

 electron discharge, from a heated filament to a plate at a higher potential 

 than the filament, so that his patent of 1904 may be considered as the parent 

 one of all valves." The analogy of the tilted water- vessel, which is used to 

 explain the action of the three-electrode valve, when receiving a series of wave 

 trains, is exceedingly ingenious. The next chapter deals with the use of the 

 three-electrode valve as an amplifier, describing not only the simple series 

 of valves that are used as amplifiers, but the valve in which a reaction coil is 

 employed to make the valve nearly " self -ex citing. " 



The " heterodyne " action now used so largely in valve receivers is also 

 explained. In chapter v, the action of the valve for generating oscillations is 

 dealt with very fully, and the various kinds of coupling between plate and 

 grid circuit discussed. The following chapter deals with the general arrange- 

 ment of low-frequency and high-frequency amplifying apparatus ; then follows 

 a short chapter on the sensitiveness of valve reception, and the next three 

 chapters deal with the characteristics of the French valve, and other special 

 continuous wave types of valves designed for varying conditions. Some of the 

 Standard C.W. transmitter sets are next described, and the following three 

 chapters give a detailed account of the various systems that are now used in 



