We present abundance analyses based on high dispersion and high signal-to-noise ratio Magellan spectra of two highly microlensed Galactic bulge stars in the region of the main sequence turnoff with T _ { eff } \sim 5650 K. We find that MOA–2008–BLG–310S has [ Fe/H ] We adopt the usual spectroscopic notations that [ A/B ] \equiv~ { } log _ { 10 } ( N _ { A } / N _ { B } ) _ { * } - log _ { 10 } ( N _ { A } / N _ { B } ) _ { \odot } , and that log [ \epsilon ( A ) ] ~ { } \equiv~ { } log _ { 10 } ( N _ { A } / N _ { H } ) +12.00 , for elements A and B . = +0.41 \pm 0.09 dex and MOA–2008–BLG–311S has +0.26 \pm 0.09 dex .
The abundance ratios for the \sim 20 elements for which features could be detected in the spectra of each of the two stars follow the trends with [ Fe/H ] found among samples of bulge giants .
Combining these two bulge dwarfs with the results from previous abundance analysis of four other Galactic bulge turnoff region stars , all highly magnified by microlensing , gives a mean [ Fe/H ] of +0.29 dex .
This implies that there there is an inconsistency between the Fe-metallicity distribution of the microlensed bulge dwarfs and that derived by the many previous estimates based on surveys of cool , luminous bulge giants , which have mean [ Fe/H ] \sim - 0.1 dex .
A number of possible mechanisms for producing this difference are discussed .
If one ascribes this inconsistency to systematic errors in the abundance analyses , we provide statistical arguments suggesting that a substantial systematic error in the Fe-metallicity for one or both of the two cases , bulge dwarfs vs bulge giants , is required which is probably larger than can realistically be accommodated .