322 BELL SYSTEM TECHNICAL JOURNAL 



varieitions in amplitude. Tlie amount of this effect is given, and the 

 usual methods for reducing it by keeping the pendulum swing small 

 are discussed. A method which is being used successfully by Mr. 

 Loomis for controlling the amplitude at a fixed value is described. 



The effect of atmospheric pressure on the rate is discussed and it 

 is shown that the four chief rate controlling factors involved can be 

 made to annul each other at a critical pressure for a given pendulum. 



Factors that affect the length of a pendulum, such as temperature 

 coefficient of the material, aging, etc., and some factors that affect 

 the restoring forces, are discussed since they affect the period directly. 



The effect of the phase of applied driving force in an oscillator is 

 also an important factor. In the case of a pendulum the impulse 

 should be delivered at the instant when the velocity is maximum, 

 that is, at the center of the swing. If it is applied earlier, the rate is 

 momentarily increased, and conversely. 



A brief description is given of the Shortt clock, which has established 

 an enviable record for timekeeping in some of the world's outstanding 

 time observatories. The installation of three of these clocks in The 

 Loomis Laboratory at Tuxedo Park, N. Y. is described and illustrated. 



The crystal clocks used in Bell Telephone Laboratories, N. Y. City, 

 are described briefly, as well as some of the outstanding features of 

 their use. These "oscillators" were built primarily as a precise 

 standard of frequency, but have been found in addition to serve 

 exceedingly well as timekeepers. 



Performance data are given for the crystal clocks, for the three 

 Shortt clocks in The Loomis Laboratory, and for clocks in a number 

 of representative national time observatories. 



In addition a brief account is given of a continuous comparison 

 which is being made between the crystal clocks in Bell Telephone 

 Laboratories and the Shortt clocks in The Loomis Laboratory. 



The Nature of Metals in Relation to their Properties.^ Earle E. 

 Schumacher. The methods of extracting metals from their ores and 

 fabricating them constituted the art of metallurgy in the older sense, 

 an art whose development has closely paralleled the rise of civilization. 

 The modern science of metallurgy, on the other hand, concerns itself 

 to a large extent with the explanation as to why metals behave as they 

 do and in particular why the methods employed in the art produce 

 the effects they do. This science, as distinct from the art, is of com- 

 paratively recent origin. It is, indeed, only since the development of 

 the tools used in modern research, notably the microscope and x-ray, 



* Scientific Monthly, January, 1932. 



