The absence of a signal from the Sandage region discussed in the last section is perhaps the most powerful indication that the signal at longer wavelengths is not back scattered starlight, and hence is extragalactic. It is not proof, however. The new information that we presented in the previous section showing that the albedo of grains continues high into the farthest astronomical ultraviolet is not conclusive, as g could be varying with wavelength to compensate. If one is desperate to avoid concluding that the radiation is extragalactic, one can postulate that for backscattered light the shape of the scattering function changes abruptly near 1216 Å. There is nothing sacred, after all, about the Henyey-Greenstein scattering form. One could also simply postulate an arbitrary additional previously-unknown dust population having the needed optical properties.
In this context it is interesting to note that Martin, Hurwitz, and Bowyer (1991) concluded (ignoring entirely the Voyager data) that what is observed at the highest galactic latitudes indicates "either the existence of a hitherto unidentified dust component, or ... a large enhancement in dust scattering efficiency in low-density gas." They reached their conclusion from the striking resemblance between their highest-latitude UVX spectrum and a lower-latitude spectrum that they believed was dust-scattered light. That is, their evidence for dust reflection at high latitudes is its exact resemblance to dust at low latitudes, but then they are forced to conclude that it is anomalous dust, and that the resemblance is therefore fortuitous.
What would it take to resolve the matters that we have discussed in this paper? A good deep ultraviolet image of the Sandage region at 1500 Å would be a great step forward. If the Sandage region is not clearly seen, that would be virtually conclusive evidence that the high-latitude background is not due to scattering by normal dust. Also, with the Hopkins Ultraviolet Telescope (Davidsen 1993) it ought to be possible to obtain a spectrum of the dust-scattered light in the direction of the Coalsack that has high enough signal to noise that a comparison is possible with the rather structured high galactic latitude spectrum of Martin and Bowyer (1990). Certainly the interstellar radiation field (Henry et al. 1980) that is incident on the putative high latitude dust does not have the observed cosmic background spectral character reported by Martin and Bowyer; the Coalsack observation, while not decisive (it would not be backscattered light), could be extremely suggestive one way or the other.
To disprove the existence of an anomalous dust component specially designed to account for the observations would be difficult. A very long exposure indeed might, if the light is scattered starlight, reveal the spectral structure (absorption lines) that is expected in an integrated B-star dust-reflection spectrum.