PHOTONS AND ELECTRONS 29 



atom can be transposed at any time into any one of those outer orbits for 

 which the energy values are given by the formula Rhl 1 — ^ j, with 



any integer substituted for n; for these orbits are always vacant, except 

 when that electron itself has just been shifted to one of them. 



Production of spectrum lines in this manner — by a stream of electrons 

 of definite and controllable energy, just sufficient to excite one or a few 

 lines — is unusual, except for definite purposes of research. It is some- 

 times fairly easy to get the first line of a series by itself ("single-line 

 spectrimi"), but the following lines are so close together that it is often 

 difficult and before long impossible to form an electron beam of suffi- 

 ciently sharply defined speed to bring one line out clearly without bringing 

 out the next one appreciably. Then too, the electron beam must not be 

 very strong — i.e., there must not be too many electrons per centimeter 

 squared per second — or atoms will suffer impacts from two electrons in 

 quick succession and lines will be emitted which no single impact could 

 produce. Usually the gas is made the theater of an electrical discharge. 



There is a great variety of electrical discharges, and a great part of the 

 art of spectroscopy consists in knowing which to use and for which spec- 

 trum lines. In all of them, the gas is traversed by electrons having 

 kinetic energy superior to the ionizing energy of the atoms or molecules, 

 so that at any moment a certain proportion of these are ions (rarely, 

 however, more than a few per cent) while neutral particles are constantly 

 being converted into ions to keep the balance while ions are resuming 

 the neutral state by capturing electrons. Nearly always the voltage 

 between cathode and anode, when multiplied by the charge of an electron, 

 gives a product eV greater than the ionizing energy of the gas — we say 

 that the voltage is greater than the ionization potential; but there are 

 some curious exceptions to this rule, for there may be places in a discharge 

 where the potential is lower than at the cathode or higher than at the 

 anode, or there may be electrical oscillations capable of communicating 

 more energy to an electron than it could get by falling unimpeded from 

 cathode to anode. The cathode may be cold {i.e., not incandescent), 

 in which case the potential difference between it and the anode must be 

 of the orders of hundreds or thousands of volts; or it may be intensely 

 hot, in which case the potential difference may amount to only a few volts. 

 The latter discharge or "arc" is generally far hotter and more luminous 

 than the former discharge or "glow." The latter is sometimes main- 

 tained in a gas of atmospheric density, the former nearly always in a 

 tube where the gas pressure is only a few thousandths of that character- 

 istic of the atmosphere. Intermediate cases may be realized by making 

 the cathode in the form of a filament to which heat may be supplied from 

 an independent battery. 



