The relation between far-UV and visible extinctions

For directions of sufficcient reddening (E(B-V) >~ 0.25), there is a simple relation between the slope of the extinction curve in the far-UV and E(B-V). Regardless of direction, the far-UV extinction curve is proportional to 1/lambda^n exp(-2e(B-V)/lambda) (lambda in micron, n = 4), in accordance with the idea that reddened stars spectra are contaminated by scattered light (Zagury, 2001b). This relation is not compatible with the standard theory of extinction which states that far-UV and visible extinctions are due to different classes of grains. In the standard theory model the two (far-UV and visible) extinctions vary thus independently according to the proportion of each type of grains. In preceding papers I have shown that the standard theory cannot explain UV observations of nebulae, and is contradicted by the UV spectra of stars with very low reddening: for how long shall the standard theory be considered as the interpretation of the extinction curve?