In the 1920s he argued that the lack of significant rotational motion in some ‘spiral nebulae’ (galaxies) indicated their extreme distance. Using redshift data for over 40 galaxies, he derived an early form of distance–velocity relation (now known as the Hubble law). He was one of the first to point out that many of the brighter galaxies were distributed around a great circle on the celestial sphere, later shown by G. H. de Vaucouleurs to reflect groupings in clusters and superclusters. In 1925 Lundmark introduced an ‘upper class’ of novae (later named supernovae) for the brightest novae observed in other galaxies.