1 Introduction

The two preceding papers (Paper I: Ginestet et al. 1994; Paper II: Ginestet et al. 1997) provided:

$\bullet$ A study of MK standards in the region $8370\,{-}\,8780$ Å, which established the validity in this region of the MK classification and provided the criteria which could be used. We showed also the interest of the near infrared spectral region for the classification of the cool companions of composite spectra (CS).

Let us remind briefly that in our studies we consider as composite spectra all those in which a hot dwarf (of spectral type B or A) spectrum appears with that of a cool subgiant, giant or supergiant (of spectral types G, K or M).

$\bullet$ An application of our criteria to a sample of 180 known or suspected to be CS: 60 stars present a hot spectrum of spectral types B, A, Am or F in the near infrared, whereas 120 present a cool spectrum of spectral type G, K or M.
These two papers are based on 33 Å/mm spectra obtained with the CARELEC spectrograph mounted on the 193 cm telescope of the OHP. We call attention to the fact that this instrument was used also to establish an Atlas of MK standards in the near infrared (Carquillat et al. 1997).

Most unfortunately in 1995 we encountered difficulties, since the spectrograph became reserved for extragalactic work and we had to change to the spectrograph AURELIE, mounted on the 152 cm telescope. Such a change had various implications for our work:

- we had to make a new calibration of MK standards and to reobserve a certain number of stars already classified to verify the joint with the previous system

- to prolongate the exposure times, with the resulting loss of limiting magnitude. As an advantage we have now a larger spectral region accessible, which includes the lines MgI 8806 and FeI 8824.

In the present paper we establish the spectral classifications of 137 objects from the near infrared, observed with the AURELIE spectrograph.

Let us mention in passing that numerous studies have been devoted in the last years to the near infrared region, from both the stellar and the extragalactic point of view. The interested reader will find the majority of these papers quoted in our Papers I and II, so that there is no need to repeat them here.