1 Introduction

The light variability of the eclipsing binary YY CMi (HD 67100) was discovered by [14, Morgenroth (1934)]. It is classified as a Beta Lyrae type eclipsing binary in the variable star catalogue of [9, Kukarkin et al. (1969)] who gave a period of 1.0940253 day and a spectral type of F5. Complete photoelectric light curves in UBV passbands were obtained by [1, Abhyankar (1962a,b)] who gave a period of 1.0940197 determined on the basis of observed minima times spread over a period of 24 yrs. Applying [22, Russel & Merrill's method (1952)] he found the primary eclipse to be a transit and concluded that it is a detached binary composed of an F8III primary and an A5V secondary. [7, Giuricin & Mardirossian (1981)] have reanalysed [1, Abhyankar's (1962a,b)] three-colour photoelectric observations using [27, Wood's (1972)] model and found that their photometric elements differed appreciably from those of Abhyankar's, leading to a contact configuration. They have expressed the view that Wood's model (which treats the stars as triaxial ellipsoids) used by them for the analysis may not handle the contact systems very well and hence the derived elements may be questionable. In view of the existing discordant opinions on the general picture of YY CMi, we felt it worthwhile to reanalyse the three light curves of [1, Abhyankar (1962a)] using the synthetic light curve analysis method of [28, Wilson & Devinney (1971)] according to their improved version of 1993. In the following, we give details of our analysis and its results.

In the meantime [18, Niarchos et al. (1998)] have obtained new photometric observations of YY CMi in V passband and analysed them by the Wilson-Devinney method. Both theirs and our results confirm the contact nature of the binary and allow us to compare the solutions obtained from the two sets of observations separated by a span of 30 years.

To our knowledge no radial velocity studies of YY CMi are yet available in the literature.