Abstract Section I. Methods are described for exposing samples of hard rubber to light and measuring their resistivity, both under controlled conditions of humidity. Section II. The effects of the following experimental conditions were investigated in preliminary tests: surface polish on specimen; type of material—loaded and unloaded; humidity of exposure; humidity of measurement; nature of light—direct sunlight and north daylight; thickness of wall of glass exposure tube; type of glass—ordinary and Vitaglass. Section III. The variation of surface resistivity at 75% R.H. with time of exposure at 75% R.H. was investigated in the case of a high grade unloaded (A) and a low grade loaded (B) hard rubber. It was found that there were three well-defined stages in deterioration: (a) an “induction period,” during which the resistivity remained practically unchanged; (b) a period during which the resistivity fell very rapidly, and (c) a “saturation period,” during which the resistivity, having reached a low value of about 107−108 ohms, continued to fall relatively slowly. The log time/log resistivity curve was found to be reducible to a simple mathematical form, in accord with which all the results in the report are analyzed. Quality A, although initially superior to B, deteriorated more rapidly and to a greater extent. The resistivity of B increased slowly on removal from the light. Section IV. The surface resistivity of quality A was sensitive to the humidity during measurement, except with samples in the “induction” and “saturation” stages; that of quality B was more sensitive to humidity at all stages between 85 and 100% R.H., but less so at lower humidities. At 97 and 100% R.H., B was in all respects inferior to A, its apparently greater stability at 75% R.H. being due to its lower hygroscopicity at normal atmospheric humidities. Section V. When the humidity of exposure was varied there was a maximum rate of deterioration at 70–90% R.H. in the case of quality A, both dry and saturated conditions having a retarding effect. Section VI. In artificial light (at 15 cm. from a 100-watt gas-filled lamp) the deterioration was slower than in bright north daylight, but for quality A the rate was the same in both cases if expressed in terms of standard hours as measured by a photographic exposure meter. Quality B deteriorated relatively less rapidly in the artificial light, a phenomenon probably connected with that of recovery of resistivity; and it is pointed out that accelerated or artificial light aging tests cannot yet be accepted as giving a true measurement of relative stability in sunlight. Section VII. The mechanism of formation of a conducting film is discussed. It is considered that (a) a complete layer of acid cannot be formed before the beginning of the “saturation” period; (b) the assumption that acid formation is due to an autocatalytic reaction is neither necessary nor sufficient to explain the results obtained; (c) the first and second stages of deterioration correspond to a period in which acid nuclei are formed, and increase in number and/or size to form conducting paths; (d) the changes of resistivity with humidity are explainable in a similar way; and (e) the difference in behavior between qualities A and B are due to the presence in B of water-soluble and/or acid-soluble substances which take part at a limited rate in secondary reactions with the acid as it is formed.