(This is a most remarkable observation. Its probable meaning is that the electric oscillations of light are photo-electrically effective in so far as they are normal to the surface on which they act; while electric oscillations tangential to the surface are scarcely operative. Different angles of incidence must be tried before the proof is complete.--O. J. L.) (3) Electric oscillations of very short period, such as are given by a Hertz oscillator, are commutated by illumination in the presence of alkali metals in rarefied gas, so as to be able to set up a constant electric tension in the gas. (A Zehnder tube[39] was used, and the momentary phases of the oscillation during which the metal is negatively charged are apparently taken advantage of by the illumination.) (4) The photo-electric dissipation showed by powdered fluorspar is dependent on the colour of the mineral, in such a way that the deepest blue, violet or green specimens are the most sensitive. [39] See Fig. 13, p. 15. APPENDIX V. PHOTO-ELECTRIC RESEARCHES OF M. AUGUSTE RIGHI.[40] M. Righi has observed the following facts: (1) That ultra-violet rays reduce to sensibly the same potential two metals placed near each other (plate and gauze parallel and close); (2) That several photo-electric couples of this kind can form a battery; (3) That a simple metallic plate charges itself positively under the influence of radiation; (4) That a voltaic arc formed with a zinc rod gives the strongest effect, while the sun gives none. [40] _Comptes Rendus_, vol. 107, p. 559. Besides these facts he finds:--(_a_) That certain gases and vapours, such as coal-gas and CS₂, absorb the active rays strongly. (_b_) That if the discharging body is easily movable it recedes like an electric windmill. (_c_) A film of gypsum interposed between gauze and plate charges itself negatively on the side facing the negatively charged plate. (_d_) Radiation produces its discharging effect even on non-conductors (ebonite and sulphur). With glass, resin and varnishes the action is feeble, or nearly nothing. (_e_) If the experiment is made with a copper gauze and a zinc plate, the phenomenon nearly disappears on varnishing the gauze. His hypothesis is that radiation produces convection of negative electricity, the carriers being molecules of air. (_f_) The carrying molecules move along the lines of force, and throw electric shadows. To show this he varnishes a zinc cylinder, all except a generating line, charges it negatively to 1,000 volts with a dry pile, and places it parallel to a large earth-connected plane, which has a narrow rectangular portion insulated from the rest and communicating with an electrometer. Light only acts on the uncovered line of the cylinder, and on turning the cylinder round the electrometer is only deflected when it is exposed to some of the (circular) lines of force emanating from the active line of the cylinder.