Absolute transition probabilities have been measured for 38 lines from 5 levels in Fe II. The branching ratios for the transitions which depopulate each level were determined by observations of the relative line intensities in a hollow cathode discharge. The relative transition probabilities, which are proportional to the branching ratios, were normalized by measurement of the lifetime of the upper level using a beam-foil, time-of-flight technique. The lifetime is the reciprocal of the sum of the transition probabilities. Our transition probabilities are compared to values from five other measurements, the most comprehensive of which are those of Warner. The work of Warner is shown to contain a significant error which comes from his use of the, now discredited, transition probabilities of Corliss and co-workers for Fe I in the determination of the temperature of his source. A method for correcting the values published by Warner is given. Our transition probabilities for Fe II were used to calculate a value of the solar iron abundance of 7.35 ± 0.18, which supports other recent determinations based on Fe I lines.
