... results

There were also a number of earlier LS-coupling calculations e.g. Jacobs et al. (1980), but they no longer form the basis for recommended data and so we do not consider them further.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1991)

The experimental results shown have been corrected for an error in the original data analysis (Wolf 2005, private communication). This mainly affected the position of the $n\ge 4$ peaks and the height of the unresolved Rydberg limit peak.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1969)

The results of Shore (1969) are a mistaken overestimate of the Burgess GF - see Burgess & Tworkowski (1976).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1981)

The representation of the results of the Burgess GF by Dubau et al. (1981) in their Fig. 5 is incorrect as they used a modified form due to Gabriel & Paget (1972) which omits a factor from the exponential and its results diverge increasingly from those of the true Burgess GF at lower temperatures (Burgess, private communication).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.