The amplitude of a wave of light, in a place where it is most intense, namely near the sun where its energy amounts to 2 ergs per c.c., comes out only about 10⁻¹⁷ of the wave-length. The maximum tangential stress called out by such strain is of the order 10¹¹ atmospheres. The hypothetical luminous circulation-velocity, conferring momentum on a wave-front, in accordance with Poynting's investigation, comes out 10⁻²² cm. per sec. These calculations are given in the concluding chapter of the new edition of _Modern Views of Electricity_. The supposed magnetic etherial drift, along the axis of a solenoid or other magnetic field, if it exist, is comparable to ·003 centim. per sec., or 4 inches an hour, for a field of intensity 12,000 c.g.s. But it is not to be supposed that this hypothetical velocity is slow everywhere. Close to an electron the speed of magnetic drift is comparable to the locomotion-velocity of the electron itself, and may therefore rise to something near the speed of light; say 1/30th of that speed: but in spite of that, at a distance of only 1 millimetre away, it is reduced to practical stagnation, being less than a millimicron per century. In any solenoid, the ampere-turns per linear inch furnish a measure of the speed of the supposed magnetic circulation along the axis--no matter what the material of the core may be--in millimicrons per sec. [1 micron = 10⁻⁶ metre; 1 millimicron is 10⁻⁹ metre = 10⁻⁷ centimetre, or a millionth of a millimetre.] To get up an etherial speed of 1 centimetre per second--such as might be detected experimentally by refined optical appliances, through its effect in accelerating or retarding the speed of light sent along the lines of magnetic force,--would need a solenoid of great length, round every centimetre of which 1000 amperes circulated 3000 times. That is to say, a long field of four million c.g.s. units of intensity. In other words, any streaming along magnetic lines of force, such as could account for the energy of a magnetic field, must be comparable, in centimetres per second, to one four-millionth of the number of c.g.s. units of intensity in the magnetic field. APPENDIX 3 FRESNEL'S LAW A SPECIAL CASE OF A UNIVERSAL POTENTIAL FUNCTION The modern view of Fresnel's Law may be worded thus:-- Inside a region occupied by matter, in addition to the universal ether of space, are certain modified or electrified specks, which build up the material atoms. These charged particles, when they move, have specific inertia, due to the magnetic field surrounding each of them. And by reason of this property, and as a consequence of their discontinuity, they virtually increase the optical density of the ether of space, acting in analogy with weights distributed along a flexible cord. Thus they reduce the velocity of light in the ratio of the refractive index μ:1, and therefore may be taken as increasing the virtual density of the ether in the ratio 1:μ².