COMMISSION 27 OF THE I.A.U. INFORMATION BULLETIN ON VARIABLE STARS Number 1444 Konkoly Observatory Budapest 1978 June 28 A DISCUSSION OF PERIOD CHANGES IN THE WHITE DWARF ECLIPSING BINARY SYSTEM V471 TAURI Times of minimum have been obtained for the white dwarf eclipsing binary system V471 Tauri (BD +16d516) between 1973 and 1976. Seven eclipses were observed with the 76.1 cm telescope of the Rosemary Hill Observatory (RHO) and four with the No. 4 40.6cm telescope of the Kitt Peak National Observatory (KPNO). The data are listed in Table 1 where the modified Julian Date (as defined by: MJD = JD - 2400000.5) of mid-eclipse is given as well as the O-C in seconds of time. The O-C's were computed from the light elements of Young and Lanning (1975) transformed to modified Julian Date: 40609.56490 + 0.52118346*E. Plotted in Figure 1 are the O-C's from Table 1 (crosses) as well as those from Lohsen (1974) - open circles, Young and Lanning (1975) - dots, and Cester and Pucillo (1976) - triangles. The times of minima obtained by Ibanoglu (1976) have not been plotted because they were published to 0.001 day. Similarly, the average O-C has been plotted for the four new times obtained with the small telescope at KPNO. It is apparent in Fig. 1 that the period of V471 Tau has changed more than once since its discovery in 1969. Young and Lanning suggested that the changes were due to mass loss or mass transfer in the binary system. Herczeg (1975) objected on the grounds that the system is detached with no other evidence for mass flow. He suggested a light-time effect in an eccentric orbit around a third body. The trend of the recent O-C's in Figure 1 indicates that if there is a light-time orbit, its period is greater than the five years suggested by Herczeg. It should be noted that Herczeg's objection to mass flow may not be entirely valid: there are a number of binary systems in which large scale mass flow is suspected even though the systems are detached, e.g. the RS Canum Venaticorum systems. [FIGURE 1] From Figure 1 we see that since epoch 3500,the period has been relatively constant. It appears simplest to assume a period change at about epoch 1500 followed by a second change at about epoch 3000. (Apparently there was an earlier change just after the discovery of the system but there is too little information available for much discussion of this event.) We have fitted three straight lines to the available O-C's as indicated in Fig. 1. The coefficients of these lines and the resulting linear light elements are given in Table 2. The third segment (C) should be useful for prediction of eclipses in the near future. For prediction of the first contact, 0.01699 day should be subtracted from the predicted time of mid-eclipse. The O-C diagram consisting of linear segments would imply rather short time intervals when the period was changing separated by longer intervals of period constancy. Unfortunately, the time scales of the period changes can not be estimated from the available material because there are not enough timings at intervals when the linear segments join together. We can obtain the upper limit to such a time scale, however, assuming that the segment B is actually parabolic, implying a continuous period change between segments A and C. The time-scale (e-folding time) of the period change estimated in that way is 2x10E6 years. Such a time-scale, being comparable to the thermal time scale of the K0V component, suggests that this component (and its variable moment of inertia in particular) might cause the period changes observed in V471 Tauri. We wish to thank Dr. Cafer Ibanoglu for kindly sending us some of his data in advance of publication. Partial support of this research came from NSF grant INT 76-80588. JPO wishes to thank Professor S.L. Piotrowski for his kind hospitality at the Warsaw University Observatory. Dr. T.R. Flesch assisted with the observations at Rosemary Hill Observatory. The observations at Kitt Peak National Observatory. were obtained when SMR held the Research Associateship of the NRC of Canada. Computations for this paper were made on the PDP 11/45 computer of the N. Copernicus Astronomical Center in Warsaw. Table 1 Site UT Date t_mid (MJD) Epoch O-C (seconds) RHO 30 Nov 73 42016.23908 2699 + 1.9 RHO 26 Oct 74 42346.14794 3332 -21.5 RHO 2 Dec 74 42383.15197 3403 -21.1 RHO 22 Dec 75 42768.30621 4142 -50.2 RHO 11 Jan 76 42788.11112 4180 -55.5 RHO 17 Feb 76 42825.11509 4251 -60.4 RHO 11 Mar 76 42849.08949 4297 -63.7 KPNO 26 Oct 76 43077.3678 4735 -68\ KPNO 27 Oct 76 43078.4100 4737 -82 \ 78.5 KPNO 28 Oct 76 43079.4523 4739 -88 / KPNO 29 Oct 76 43080.4948 4741 -76/ Table 2 Segment Range in Epoch Linear O-C Fit Linear Light Elements (MJD) A 500-1800 -24.7s+0.0259sxE 40609.56462+0.52118376xE B 1800-3000 +42.3 -0.0151 xE 40609.56539+0.52118329xE C 3000-5000 +131.2 -0.0445 xE 40609.56642+0.52118294xE J.P. OLIVER S.M. RUCINSKI* Rosemary Hill Observatory Warsaw University Observatory University of Florida and Dominion Astrophysical Observatory, Victoria References: Cester, B. and Pucillo, M. Astron. & Astrophys., 46, 197, 1976 [BIBCODE 1976A&A....46..197C ] Herczeg, T. IBVS No. 1076, 1975 Ibanoglu, C. IBVS No. 1088, 1976 Lohsen, E. Astron. & Astrophys., 36, 459, 1974 [BIBCODE 1974A&A....36..459L ] Young, A. and Lanning, H. Publ. Astron. Soc. Pacific, 87, 461, 1975 [BIBCODE 1975PASP...87..461Y ] * Visiting Astronomer, Kitt Peak National Observatory, which is operated by the Association of Universities for Research in Astronomy, Inc., under contract with the National Science Foundation.