If we assume a parallax of say 0"·01--or one-hundredth of a second of arc--which would still keep the nebula within the bounds of our sidereal system--we have the dimensions of the nebula increased 17 times, and hence its mass nearly 5000 times greater (17{3}) than that found above. The mass would then be 40,000 million times the sun's mass! This result seems highly improbable, for even this small parallax would imply a light journey of only 326 years, whereas the distance of the Milky Way has been estimated by Prof. Newcomb at about 3000 years' journey for light. In Dr. Roberts' photograph many small stars are seen scattered over the surface of the nebula; but these do not seem to be quite so numerous as in the surrounding sky. If the nebula lies nearer to us than the fainter stars visible on the photograph, some of them may be obscured by the denser portions of the nebula; some may be visible through the openings between the spiral branches; while others may be nearer to us and simply projected on the nebula. To add to the difficulty of solving this celestial problem, the spectroscope shows that the Andromeda nebula is not gaseous. The spectrum is, according to Scheiner, very similar to that of the sun, and "there is a surprising agreement of the two, even in respect to the relative intensities of the separate spectral regions."[366] He thinks that "the greater part of the stars comprising the nucleus of the nebula belong to the second spectral class" (solar), and that the nebula "is now in an advanced stage of development. No trace of bright nebular lines are present, so that the interstellar space in the Andromeda nebula, just as in our stellar system, is not appreciably occupied by gaseous matter."[366] He suggests that the inner part of the nebula [the "nucleus"] "corresponds to the complex of those stars which do not belong to the Milky Way, while the latter corresponds to the spirals of the Andromeda nebula."[366] On this view of the matter we may suppose that the component particles are small bodies widely separated, and in this way the _mean_ density of the Andromeda nebula may be very small indeed. They cannot be large bodies, as the largest telescopes have failed to resolve the nebula into stars, and photographs show no sign of resolution.