4 Conclusions

We have shown that HD 174005 is a spectroscopic binary and the observed spectrum is highly variable with the phase. The confusion between the spectrum of a single weak-lined star and the composite spectrum of a binary composed by two similar, but not equal sources, is demonstrated.

A careful check of high dispersion data and the analysis based on more than a single spectrum appear to be necessary to obtain reliable abundance analyses of $\lambda$ Boocandidates; as a first step it is necessary to distinguish composite spectra from those of single weak-lined stars and this can require a quite subtle analysis of mean-high resolution spectra.