In order to understand how this is possible, we must remember that in two respects the spectrum is not a complete phenomenon. There is, to begin with, the fact that the colour-band visible on the observation screen is only apparently confined to the surface of the screen. For, as we have seen, because of the differing co-ordination of levity and gravity at the two ends of the spectrum, the conditions of space prevailing at each are polarically opposite. Negative space opens up spherically behind the blue-violet colours on one side, while positive space, filled by the radially shining yellow-red colours, arises on the other. So we see that what we found earlier for the two poles of magnetism and electricity holds good also for the spectrum. That is, the two processes bringing about the relevant phenomena are not confined to the part of space which these phenomena seem to occupy; for the whole positive and negative realms of the universe share in them. Hence the spectrum, though apparently bounded at its two ends, proves by its very nature to be part of a greater whole. Once before we were led to recognize - though from a different aspect - that the spectrum is a phenomenon which, when rightly viewed, calls for a certain completion. In following Goethe's initial observations we realized that the known spectrum, extending from red via green to violet, has a counterpart extending from violet via peach-blossom to red. The reader may have wondered why we never returned to this other spectrum, in spite of the role it played in making Goethe aware of Newton's error. The reason was that in order to gain the understanding we needed of the spectrum, we had to observe the two border-phenomena independently - that is, without regard to their relative positions. Moreover, with ordinary optical means it is possible to produce only one type of spectrum at a time, so that each is left in need of being complemented by the other. In order to have both together in finite space, as part of one and the same phenomenon, space itself must be dynamically transformed in such a way that the continuation of the finite spectral band running through infinity enters into the finite as well. Our understanding of magnetism as a specific representation of the polarity of the second order enables us to comprehend, at least in principle, how magnetism might influence - not light itself, as present-day physics erroneously believes - but the secondary polarity of the spectral colours formed out of the primary polarity Light and Dark. To see this in all necessary detail is a task of the future, beyond the scope of this book. We have here to continue our account of Rudolf Steiner's statement by communicating what he indicated concerning the particular nature of the new source of force which would appear in the normally infinite part of the spectrum, if this were brought into the region of the finite.