Part I.
An analysis is made of the errors which arisefor absolute levels and fine structure separationsfrom the use, for the interaction of two electrons,of the nonretarded Breit expression instead of thetrue retarded interaction. The effects on the finestructure turn out to be too small for observation.The correction to the ground state energy maybecome observable if the experimental accuracy isincreased. This is probably not the case for any excited level.
Part II.
The exact propagation function for a Diracelectron in an arbitrarily intense magnetic fieldis derived in closed form as a parametric integral.Using the exact relativistic wave functions andthe exact of order e2, due to the emission and reabsorption in closed form as double parametric integrals. These integrals are shown to possessan asymptotic expansion in the small parameter H/m2.This expansion is not a pure power series but involvesalso terms of the form Hkln(m2/H). The termsof order H agrees with the known correction to the magnetic moment. The terms of order H2and Hkln(m2/H) are exhibitedand discussed.
