As a student of Charlier at Lund observatory he became a member of the stellar statistical research group founded by Charlier, applying mathematical methods for statistical analysis of large amounts of observational data.
He developed these methods resulting in for example the "Malmquist relations" which is today one of the standard methods in statistical astronomy.
In 1931 he moved to Stockholm observatory and took part in the completion of the new observatory at Saltsjöbaden.
In 1939 he was installed as professor in Uppsala where he continued his
theoretical works using the observational data from the large research on
the Milky Way which had been in progress in Uppsala since the beginning of the century.