New TL calculations for the production of the long-lived
  1s2s2p 4P state have been performed in the IPM within
  a factor of 2 of existing data for O5+ and almost in
 agreement for F6+
 New idea (old idea):  the inclusion of cascade feeding
  from higher lying 1s2snl 4L states also produced by TL
 Since cascade feeding rates and branching ratios are
  still unknown we have assumed that ALL 1s2snl 4L states
  produced are eventually converted to 1s2s2p 4P states
  thus establishing an upper limit for this contribution
 Whether yet another mechanism is needed to account for
  the missing factor of 2 remains an interesting possibility