506 



The colour and polarisation of the light scattered in 

 the sea is discussed by Prof. Raman in a chapter which 

 must interest biologists as well as physicists. The 

 colour of the deep sea is not mainly due to reflected sky- 

 light, as has sometimes been suggested, but to light 

 molecularly diffused from within the water. The re- 

 flecting power of water at normal incidence is quite 

 small (only 2 per cent.), and consequently to an observer 

 flying at a great height above the surface of the water 

 the luminosity of the sea would be determined almost 

 entirely by internal scattering. 



In crystals such as quartz and rock-salt the scattering 

 of light can be observed visually, the Tyndall cone 

 being of a blue colour. The effect may be attributed 

 to the thermal movement of the atoms in the crystal 

 introducing local fluctuations of optical density. Thus 

 there is a close connexion between this phenomenon 

 and the well-known influence of temperature (" Debye 

 effect ") on the intensity of X-ray reflection as illus- 

 trated, for example, in the experiments of Sir W. H. 

 Bragg on rock-salt. It may be suggested that further 

 study of the scattering of light in amorphous solids like 

 glass would yield information of value regarding the 

 molecular structure of such bodies. 



H. S. A. 



Technical Electricity. 



(1) l'n lid pies of Electrical Engineering. By Prof. 

 W. IT. Timbie and Prof. Vannevar Bush. Pp. viii + 

 513. (New York: J. Wiley and Sons, Inc. ; London: 

 Chapman and Hall, Ltd., 1922.) 20.5. net. 



(2) High Voltage Power Transformers. By W. T. 

 Taylor. Pp. x+117. (London : Sir I. Pitman and 

 Sons, Ltd.. 1922.) 2s. 6d. net. 



(3) Electric Power Systems. By W. T. Taylor. Pp. 

 xii + 107. (London : Sir I. Pitman and Sons, Ltd., 

 1922.) 2: . (''</. net. 



(1) ' I ^HIS book is intended for students of electrical 

 J_ engineering. Although much of the ground 

 has been covered in a similar way before, yet there 

 are several novel features. The magnetic circuit — the 

 importance of the theory of which in electrical engineer- 

 ing can scarcely be over-estimated — is allotted the space 

 it deserves. The electron theory, which is admirably 

 adapted for giving easily intelligible explanations, is 

 freely used. Thermionic emission, conduction through 

 gases, and electrolytic conduction are all discussed. 

 There are also about 500 practical problems, some of 

 which give interesting engineering data. 



The authors divide electrical engineers into three 

 classes. The first class comprises the men who apply 

 scientific laws to electrical development. This includes 

 NO. 2763, VOL. I IO] 



NA 7 URE [Octob E r 14, 1922 



the research engineer and the designer. The second 

 class includes the distributing engineer who plans, 

 constructs, and operates power-transmission lines, 

 telephone and telegraph circuits, and electric railway 

 systems. There is, finally, the engineer who acts as 

 liaison officer between electrical engineering and civil 

 and mechanical engineering, including all industrial 

 applications. The consulting engineer and the pro- 

 moting or " commercial " engineer belong to this 

 class. This type of engineer must be well grounded 

 in economics, and well versed in business, law, and 

 procedure. 



The authors state that the terminology they use is 

 that recommended by the American Institute of 

 Electrical Engineers. It differs in several important 

 respects from that recommended by the International 

 Electrical Commission. The Gauss is defined as the 

 unit of magnetic induction density. In I.E.C. nomen- 

 clature the unit of magnetic induction density is one 

 .Maxwell per square centimetre. In Europe the Gauss 

 is defined practically always as the unit of magnetic 

 force. The authors take as the unit of magnetic force 

 " one Gilbert per cm.," and the unit of reluctance is 

 called the "Oersted." We do not think that the 

 Gilbert and the Oersted will be accepted internationally. 

 In our opinion also a case has not been made out for 

 the use of such words as " abohm," " abvolt," " stat- 

 volt," etc. Abohm is the unit of resistance in the 

 absolute system of units and " statvolt " is the unit 

 of electric pressure in the electrostatic system of 

 units. 



In discussing sparking voltages between spherical 

 electrodes (p. 417), it is stated that " small balls with 

 a given separation break down at a smaller potential 

 than large balls." This is not true in all cases. Russell 

 (Journ. I.E.E., vol. 57, p. 228), for example, states 

 that for half a centimetre spark at 25° C. and 76 cm. 

 the disruptive voltages for spheres of i cm., 5 cm., and 

 25 cm. radius are 17-7, 16-3, and 15-0 kilovolts respec- 

 tively. It will be seen that the voltages required for 

 a break-down in this case are appreciably higher for 

 the small electrodes. For still smaller-sized electrodes 

 for the same air-gap the disruptive voltages get smaller. 

 Similarly when the air-gap is 1 cm. it can be shown 

 that the sparking voltage has its maximum value of 

 32 kilovolts when the diameters of the electrodes are 

 each equal to 2 inches very approximately. 



(2) Mr. Taylor's book will be of interest to the electrical 

 engineer, as it discusses problems in which he takes 

 great interest. The rating of a transformer, and con- 

 sequently its price per kilowatt, depends on the " hottest 

 spot " temperature after a full-load run. Considerable 

 space is therefore devoted to methods of keeping trans- 

 formers cool. It is stated that a cast-iron case will 



