Time lag in photoelectric cells

Abstract

§ 1. This work continues that of Campbell and Stoodley and of Campbell. § 2. Frequency-response curves are compared with oscillograph traces of the current through the cell when light is suddenly thrown on or off it. A theory of the comparison is given; it is shown that the comparison provides a test of the assumption of independence, i.e. the assumption that time lag arises wholly in processes following a single instantaneous light-impulse, and not from the interaction of succeeding impulses. The experiments are wholly adverse to the assumption; low-frequency time lag must arise in great part from such interaction. The evidence on which the assumption was based previously is re-examined; it is found to be exceptional and not typical, and therefore quite inconclusive. § 3. Attempts to determine directly how time lag is affected by the size of the cells, other things being as far as possible equal, failed because it was evident that factors other than those contemplated were being varied accidentally. But measurements on a single long cell with several anodes indicated that low-frequency time lag is not due directly to the finite velocity of the ions. Incidentally the positive ion current was found - as is to be expected - to show greater time lag than the whole current. § 4. The experiments on cells cooled in liquid air, described in paper II, were continued; the previous results were confirmed; temperature has an effect separate from that on the density of the gas. § 5. The very small time lag of argon-filled cells with sensitized potassium cathodes was found to be due to hydrogen evolved from the cathode. When this is removed the time lag increases greatly. Further investigation of this remarkable result is deferred. § 6. In a brief discussion it is pointed out that the distinction previously made between low-frequency and high-frequency time lag must be maintained; the latter is undoubtedly due to the finite velocity of the ions, the former is not. For the low-frequency lag metastable states are still the most plausible explanation; but there are difficulties in the failure of the assumption of independence and in the effect of hydrogen. No alternative theory is proposed at present; more facts are required.