Oh Paul, I can be much more needlessly obscure than that.

> "GOTO 137.03599765"

The question is: to what accuracy is 1/alpha known?

When I last studied studied this (around 1996)

there were 2 answers:

1) Good old experimental skepticism:

1/alpha

= 137.0359979

+/- 0.0000032

2) Theory-laden arrogance:

Combining the expermiental value of the electron's

g-2 (anomalous magnetic moment) and the theoretical

prediction as a perturbation expansion in alpha

one may extract

1/alpha =

= 137.03599941

+/- 0.00000056

They agreed, but the theorists' sigma was ten times smaller

than the experimetalists'.

The July 2000 Particle Data Booklet now has

1/alpha =

= 137.03599976

+/- 0.00000050

which is presumably exptl and indicates that we theorists

did a good job.

So a reasonable command would be GOTO 137036/10**3

(which ought to be a integer line number in Fortran :-)

[OT [OT]]

David