Spiral galaxies are usually taken to be structured lenticular objects with spiral arms depending from a bulging star-dense central hub. The whole arrangement is essentially planar. Since our view of these objects is projected upon our fixed celestial sphere, their shape has to be inferred from the various projected shapes of many similar galaxies. Three- dimensional features of a single galaxy must appear flattened in projection. The spectacular Hubble photo of the barred spiral NGC 1073 just released by NASA nevertheless looks to me as if its spiral arms might not be coplanar with its central bar, which looks a bit like a star-dense lenticular disc inclined to the plane of projection. Is this possible? Or dynamically forbidden?