Am stars are sometimes classfied acc2 the scheme kSP1hSP2mSP3 and similar, f.ex. HD 159560 is of type kA3hA7mF2 = Am(A3/A7/F2). These k h m have eluded me for a couple of years, but the links explain that k a.k.a. K is really CaII K-line, h a.k.a. H is really Hydrogen Balmer lines, and m a.k.a. M is really "metallic lines". Since Am stars are slow rotators, their "surfaces" have a uniform temperature in comparison to f.ex. the fast rotator standard star Vega, who have a non-uniform surface temperature. The said paper calls Vega abnormal, and the Am stars normal, which is kind of easy to understand. The so called peculiar spectra of Am stars have a tripartite spectrum version to classify them, although this kind of spectrum would also be applicable to any A to early F star. ... said: Rursus (bork²) 09:16, 13 January 2009 (UTC)
page 171 tells us that Vega α Lyr and α Dra are underabundant in MgII.