Paper detail

The 2200 A bump and the UV extinction curve

The 2200 A bump is a major figure of interstellar extinction. Extinction curves with no bump however exist and are, with no exception, linear from the near-infrared down to 2500 A at least, often over all the visible-UV spectrum. The duality linear versus bump-like extinction curves can be used to re-investigate the relationship between the bump and the continuum of interstellar extinction, and answer questions as why do we observe two different kinds of extinction (linear or with a bump) in interstellar clouds? How are they related? How does the existence of two different extinction laws fits with the requirement that extinction curves depend exclusively on the reddening E(B-V) and on a single additional parameter? What is this free parameter? It will be found that (1) interstellar dust models, which suppose the existence of three different types of particles, each contributing to the extinction in a specific wavelength range, fail to account for the observations; (2) the 2200 A bump is very unlikely to be absorption by some yet unidentified molecule; (3) the true law of interstellar extinction must be linear from the visible to the far-UV, and is the same for all directions, including other galaxies. In extinction curves with a bump the excess of starlight (or the lack of extinction) observed at wavelengths less than lambda=4000 A is due to a large contribution of light scattered by hydrogen on the line of sight. Although counter-intuitive this contribution is predicted by theory. The free parameter of interstellar extinction is related to distances between the observer, the cloud on the line of sight, and the star behind it (the parameter is likely to be the ratio of the distances from the cloud to the star and to the observer). The continuum of the extinction curve or the bump contain no information concerning the chemical composition of interstellar cloud.