J. Reader, J. Opt. Soc. Am. 62, 1336A (1972); 64, 1017 (1974).

L. J. Radziemski, Jr. and V. Kaufman, J. Opt. Soc. Am. 59, 424 (1969).

R. D. Cowan, Phys. Rev. 163, 54 (1967).

If radial wave functions had been computed for the *p*^{4}*d*^{2}*S* terms, the configuration-interaction matrix elements between ^{2}*S* terms would have been zero (Brillouin's theorem). However, we are using radial functions for the center of gravity of each configuration; hence only the average interaction is zero, and there are residual nonzero matrix elements for the individual *LS* terms.

There are, of course, four additional *J* = ½ levels in each *p*^{4}*d* configuration. Interactions with these ^{4}*D*, ^{4}*P*, and ^{2}*P* levels are appreciable only over very small ranges of the *R*^{k} scale factor, and for simplicity have been ignored in the figure.

A case similar to the present one is discussed by E. U. Condon and G. H. Shortley, The Theory of Atomic Spectra (Cambridge University Press, Cambridge, 1935), pp. 41 and 42.

Simpler cases, involving only two interacting levels, are discussed by R. D. Cowan and K. L. Andrew, J. Opt. Soc. Am. 55, 502 (1965).

U. Fano, Phys. Rev: 124, 1866 (1961).

The possibility of a Beutler—Fano profile appearing in a discrete Rydberg series as well as in a continuum was recognized by Fano (Ref. 8, end of Sec. 1, and Appendix B). An experimental instance of such a profile extending across the boundary between discrete and continuous spectrum appears in Ba I: W. R. S. Garton and K. Codling, Proc. Phys. Soc. Lond. 75, 87 (1960); W. R. S. Garton and F. S. Tomkins, Astrophys. J. 158, 1219 (1969).

Corresponding values for the transitions from the perturbed and unperturbed 3*d*^{2}*S*_{½} levels are *A* = 0.8·10^{8} s^{-1} and 2.6·10^{8} s^{-1}, respectively.

J. L. Tech, J. Res. Natl. Bur. Std. 67A, 505 (1963).

L. Minnhagen, Ark. Fys. 21, 415 (1962).

Y. G. Toresson, Ark. Fys. 18, 417 (1960). We calculate the 3*s*3*p*^{5}^{1}*P* level to interact comparatively weakly with the continuum, and to remain far above the ionization limit at ~125 kK. If this prediction is correct, it would indicate that Toresson's *sp*^{5}^{1}*P* designation for a level at 81.437 kK is in error; the proper identification may be 3*p*^{3}(^{2}*D*)3*d*^{1}*P*, which we compute to lie at 81.6 kK.

L. J. Radziemski, Jr. and V. Kaufman, J. Opt. Soc. Am. 64, 366 (1974).

A. W. Weiss, Phys. Rev. 178, 82 (1969); K. B. S. Eriksson, Ark. Fys. 39, 421 (1969); R. D. Cowan, J. Phys. Colloq. Suppl. 31, C4-191 (1970). The level computed in each of these investigations to lie just above the ionization limit is the nonphysical result of including only a finite number of 3*s*^{2}*nd* configurations, this level having been pushed off the top of the Rydberg series in the same way that *sp*^{6} is pushed off the bottom in Fig. 2; the Fano-profile phase relations are such that transitions from this level to the ground configuration acquire a very high strength, analogous to the strong continuum in Fig. 4. If continuum effects are included in the manner of the present Cl I calculation, this physically unrealistic level settles inconspicuously down among the other members of the Rydberg series (Ref. 16); cf. C. D. Lin, Astrophys. J. 187, 385 (1974) and A. W. Weiss, Phys. Rev. A 9, 1524 (1974).

R. D. Cowan, unpublished calculations (1971).