2016PASP..128f4202S

A history is given of the discovery between 1914 and 1935 of stars of intermediate luminosity between giants and dwarfs with spectral types between G0 to K3. The Mount Wilson spectroscopists identified about 90 such stars in their 1935 summary paper of spectroscopic absolute magnitudes for 4179 stars. Called "subgiants" by Stromberg, these 90 stars defined the group at the time. The position of the Mount Wilson subgiants in the HR diagram caused difficulties in comparisons of high weight trigonometric parallaxes being measured in the 1930s and with Russell's prevailing evolution proposal, and critics questioned the reality of the Mount Wilson subgiants. To show that the 1935 Mount Wilson subgiants are real, we compare, star-by-star, the Mount Wilson spectroscopic absolute magnitudes of the 90 stars defining their sample against those absolute magnitudes derived from Hipparcos trigonometric parallaxes. We address concerns over biases in the Mount Wilson calibration sample and biases created by the adopted methodology for calibration. Historically, these concerns were sufficient to discredit the discovery of subgiants in the Mount Wilson sample. However, as shown here, the majority of the Mount Wilson stars identified as subgiants that also have reliable Hipparcos trigonometric parallaxes do lie among the subgiant sequence in the Hipparcos HR diagram. Moreover, no significant offset is seen between the M_V brightnesses derived from the Mount Wilson spectroscopic parallaxes and the M_V values derived from Hipparcos trigonometric parallaxes with σ_π/π0.10, which confirms in an impressive manner the efficacy of the original Mount Wilson assessments. The existence of subgiants proved that Russell's contraction proposal for stellar evolution from giants to the main sequence was incorrect. Instead, Gamow's 1944 unpublished conjecture that subgiants are post main-sequence stars just having left the main sequence was very nearly correct but was a decade before its time.