This paper is based on the conviction that a greater degree of discrimination is possible in spectral classification than that now obtainable with the MK System. This is due principally to incompleteness in the system itself. The MK System furnishes a satisfactory frame of reference for stars of Population I in the solar neighborhood; but when we consider the variety of stellar spectra encountered in our galaxy and in other stellar systems we are unable to classify many stars satisfactorily from the MK standards alone.
What we now require is a more general frame of reference - or frames of reference - that will accept with precision a greater variety of ``specimens'' (stellar spectra). To bring these into existence, we propose to supplement the MK System with additional systems that will be constructed according to what we shall label the MK Process.