Experiments I have recently made in collaboration with Dr. F. C. Chalklin on the one hand and with Mr. F. S. Robertson on the other, together with some observations not yet published made in the Wheatstone Laboratory by Mr. E. Rudberg, taken in conjunction with Krefft’s results for the secondary electron emission from baked tungsten, throw a good deal of light on the mechanism of the generation of secondary electrons at the surfaces of solids, particularly in the range where the energy of the primary electrons is sufficient to generate soft X-rays. I shall first consider what are the chief essential facts from this point of view as to what happens when a beam of electrons falls on a conductor. In 1908 I found that a considerable proportion of slowly moving electrons was reflected by a metal plate; in the particular case of the electrons coming from a hot platinum strip under no applied electric force on to a brass plate, I estimated the proportion reflected at roughly 30 per cent. A similar result with electrons of 2, 4 and 8 volts, equivalent energy was obtained independently about the same time by von Baeyer. Since that time a number of investigations of electron reflection at conductors have been published. Speaking broadly, it appears that with increasing energy of the primary electrons the proportion reflected, increases to a maximum, at a value which is in the neighbourhood of 11 volts for a number of metals, falls to a minimum at a value which is comparable to 30 volts, rises to a second maximum at a value which is of the order of 200 volts, and then falls off slowly and continuously with further increase in the energy. These results vary to some extent with the nature and treatment of the metal surfaces, but it is important to observe that there is generally some range of voltage in which the number of secondary exceeds the number of primary electrons. This is usually in the region in which the soft X-ray emission becomes important. Farnsworth examined the electrons emitted from a nickel plate and found that with primary electrons having 9 volts energy or less, a large proportion of the secondary electrons had an amount of energy nearly equal to that of the primary electrons, and but a small proportion had a velocity of 1 volt or less. As the energy of the primary electrons was increased above 9 volts, the proportion of low velocity electrons steadily increased and the proportion of secondary electrons having energy close to that of the primary electrons steadily fell to a very small percentage at 110 volts, a result previously obtained by Davisson and Kunsman.
Model of discharge coefficient of crest gate rubber weir at fully closed condition
