5 Conclusions

The assertion of Chollet and Sinceac (1999) mentioned in Sect. 1, that all solar astrolabes, except Santiago, give identical variations in time of the solar radius and in opposite phase with Sun's magnetic activity, is at issue with what we have shown in this paper. In fact, according to our results, the CCD astrolabes of Rio de Janeiro and Antalya, and the visual astrolabe of Santiago, give almost identical variations in time of the solar radius correlated with the variation of sunspot numbers. The exception is the astrolabe of CERGA which according to Fig. 1, and following the analysis of Pap et al. (2001), gives a rather marginal solar radius variation in opposite phase with the solar cycle. On the other hand, concerning solar radius measurements made with other instruments, it should be noted that results obtained by Wittmann & Bianda (2000) between 1990 and 2000 did not show any longterm variation in excess of  $\pm0\hbox{$.\!\!^{\prime\prime}$ }05$. This was confirmed by the latest results of these authors made until April 26, 2002 (Wittmann 2002).

We did not include in this discussion the CCD astrolabes of San Fernando (Spain) and São Paulo (Brazil), since we do not know long term data of these astrolabes obtained during the observational period of the astrolabe of Santiago.

Finally we would like to emphasize that the agreement of the astrolabes of Antalya, Rio de Janeiro and Santiago, is supported by recent results based on different observational techniques that give also empirical evidence that the apparent solar radius would vary in phase with the cycle of the Sun. For instance:

• From observations made at Mount Wilson Observatory during cycles 21 and 22, following a method described by LaBonte & Howard (1991), Ulrich & Bertello (1995) have shown that the apparent radius of the Sun viewed in the spectral line of neutral iron at 525 nm varies in phase with solar activity.

• Based on radio-observations of the Sun made at 48 GHz during 1991-1993 at Itapetinga Radio Observatory, São Paulo, Brazil, Costa et al. (1999) have detected a decrease of the measured radius as the solar cycle declined.

• In a three year experiment (1996-1998) to detect solar radius fluctuations with the Michelson Doppler Imager (MDI) on board the Solar and Heliospheric Observatory (SOHO), Emilio et al. (2000) found that the solar radius increases approximately in a linear way with the number of sunspots (see also Gough 2001).

• An analysis made by Basu (1998) of homogenised data sets of solar radius measurements made during the last three centuries gives a statistically significant relationship between larger diameters of the Sun and higher levels of solar activity.

Acknowledgements

Our thanks to Dr. Fredrik Rantakyro, Department of Astronomy, Universidad de Chile, for his kind assistance concerning the use of facilities for PostScripts graphics.