Broadband ultraviolet photometry from the TD-1 satellite and low-dispersion spectra from the short wavelength camera of IUE have been used to investigate a long-standing proposal of Böhm-Vitense that the normal main-sequence A and early-F stars may divide into two different temperature sequences: (1) a high-temperature branch (and plateau) comprised of slowly rotating convective stars, and (2) a low-temperature branch populated by rapidly rotating radiative stars. We find no evidence from either data set to support such a claim, or to confirm the existence of an "A-star gap" in the B-V color range 0.22 ≤ B-V ≤ 0.28 due to the sudden onset of convection.

We do observe, nonetheless, a large scatter in the 1800-2000 Å colors of the A-F stars, which amounts to ~0.65 mag at a given B-V color index. The scatter is not caused by interstellar or circumstellar reddening. A convincing case can also be made against binarity and intrinsic variability due to pulsations of δ Sct origin. We find no correlation with established chromospheric and coronal proxies of convection, and thus no demonstrable link to the possible onset of convection among the A-F stars. The scatter is not instrumental. Approximately 0.4 mag of the scatter is shown to arise from individual differences in surface gravity as well as a moderate spread (factor of ~3) in heavy metal abundance and UV line blanketing. A dispersion of ~0.25 mag remains, which has no clear and obvious explanation. The most likely cause, we believe, is a residual imprecision in our correction for the spread in metal abundances. However, the existing data do not rule out possible contributions from intrinsic stellar variability or from differential UV line blanketing effects owing to a dispersion in microturbulent velocity.