On a method for making the wave length of sodium light the actual and practical standard of length

