On the U. V. Spectrum of O2