The two-telescope interferometer (I2T) (Koechlin 1985)
located at Observatoire de la Côte d'Azur
(OCA), France, works with two 26-cm telescopes movable along a North-South baseline. The I2T
built by Antoine Labeyrie in 1972 for observing interference fringes on 
Lyrae with a 12 m baseline
(Labeyrie 1975), provided high angular resolution measurements in the visible and near infrared
domains (Di Benedetto & Rabbia 1987;
Koechlin & Rabbia 1985;
Faucherre et al. 1983;
Bonneau 1979).
In 1993, a new beam combining table, the Active Stabilization in Stellar Interferometry (ASSI)
table (Damé et al. 1988; Sorrente et al. 1991),
developed at Office National d'Etudes et de Recherches
Aérospatiales (ONERA),
France, was installed at I2T (Robbe et al. 1994).
The goal of this instrument
is to improve the quality of interferometric measurements in routine operation of the interferometer.
ASSI is a recombining table for automatic acquisition of the interference fringes and compensation for
the optical path fluctuations, a task for which it is crucial to minimize the wavefront distortions caused
by atmospheric turbulence. Most importantly, the mean slope of the wavefront over each aperture must
be stabilized and the remaining corrugations have to be minimized. Such stabilizing is achieved by
correcting the angle of arrival. To keep the loss of fringe visibility below 5%, the required single beam
stability must correspond to residual fluctuations of less than 0.12 of the size of the Airy disk radius
(Buscher 1988).
The first fringes were observed and stabilized on the ASSI table in June 1994 with a
11 m baseline.
The instrument is described in Sect. (2), providing specific requirements linked to the operating
conditions for the system so as to reach the expected tracking accuracy. In Sect. (3), we recall what is
the available information from observed data, namely the characteristics of the servo-system, the
estimation of the Fried parameter r0 and of the temporal spectrum of the angle of arrival fluctuations.
The measurements are discussed in Sect. (4), and Sect. (5) provides results concerning observed
tracking accuracy, estimation and statistics of seeing parameters, and observed temporal power spectra
of the angle of arrival fluctuations. The latter is compared with the behavior predicted by Kolmogorov
theory.