The use of the coherer to investigate adsorption films
In spite of the former widespread use of the "coherer" or loose-contact detector of electric waves, and the numerous researches on its mode of action, no systematic investigation appears to have been undertaken to ascertain the effect on the detector of the gas surrounding the contact. Branly and others suggested that the cohering effect under electric stress across the contact might be due to the dispersion of a badly-conducting film originally lying between the metal surfaces at the point of contact. This hypothesis appears to be confirmed by the results described below, which indicate not only a highly specific effect for each gas, but an effect in agreement with the known phenomena of adsorption of gases at solid surfaces. The possibility that additional information might be obtained about the stability and formation of these films was the incentive to the present work. The theory of Eccles, founded upon the thermo-electric properties of the metals in contact and depending upon temperature changes at the point of contact, seems to be untenable when it is found that cohering action can certainly occur with a frequency of 10 6 times a second. In the investigation described below, contacts between fine tungsten filaments taken from an old lamp, between platinum filaments, and between carbon-tungsten surfaces were studied. The first named material is admirably suited for these experiments, which are held to depend upon the formation of films of adsorbed gas, as by reason of its very refractory properties it can be raised to a high temperature in order to clean the surface. In addition there is the advantage that the filaments must have been exceptionally well glowed-out during the working of the lamp of which they formed a part. The same advantage applies to the carbon filament used.
