Lithium always poses an interesting problem because of its unique
sensitivity to various mixing processes in stellar envelopes. This is
also the case on the hot side of the Li gap. Theoretical
calculations of Li diffusion predict a Li cloud will form under the
superficial convection zone. This should result in a Li bump -
overabundance on the hot side of the Li gap at around 7000 K accompanied
by a high Li deficit for hotter stars (Richer & Michaud 1993).
Modern CCD detectors enabled
Boesgaard (1987), Burkhart et al. (1987) and
Burkhart & Coupry (1989, 1991, 1997) to study Li in Am stars
for the first time. They concluded that in general Li is normal in Am and
normal A stars in the range 
(with 
) with a tendency to a slight deficit in Am
stars
.
In addition, they found that the scatter of observed abundances is rather
high, but that there might be a predominance of Li deficient
Am stars among the cooler ones. The star
16 Ori is a unique exception as it has by far the largest
lithium deficit among the Am stars, as well as large anomalies of other elements
(Burkhart & Coupry 1989). Having an orbital period 
and eccentricity e=0.67 it belongs to a very rare
group of long period Am binaries. The main aim of this paper is to explore
Li in three other long period Am binaries, namely HD108651, HD116657 and
HD138213.
It is well known that Am stars occur in binary systems with a frequency
much higher than that of normal stars (Abt 1961, 1965 and Abt
& Bidelman 1969). Recently Budaj (1994, 1996, 1997),
pursuing further the pioneering ideas of Abt in Am binaries, showed that:
the metallicity tends to increase with
up to at least 
, probably even up to 
; the maximum rotation velocity and
the curve of constant metallicity depend on ;
and the orbital period distribution (OPD) revealed an interesting gap
within 
. To account for these features he formulated the
tidal mixing+stabilization hypothesis (consult the references above for more
details), which assumes the tidal effects to stretch to large 's
and to play a dominant role in driving Am phenomena. The new hydrodynamical
retardation mechanism of Tassoul & Tassoul (1992) might well
fit into this pattern. The other goal of this paper is to inspect the metal
abundance anomalies (mainly Ca, Fe) of the long period Am
binaries in the most questionable region of orbital
periods ranging from 50 up to 200 days.
Finally, it is aimed to put our results, along with previously published data, in a more general context to look for further evidence of tidal effects in Am stars. Namely, we will study the suspected relation of lithium and calcium/iron abundances on the eccentricities and orbital periods of Am binaries.