COMMISSION 27 OF THE I.A.U. INFORMATION BULLETIN ON VARIABLE STARS Number 1482 Konkoly Observatory Budapest 1978 October 6 DETECTION OF EMISSION LINES OF HOT PLASMA IN FIVE PECULIAR ECLIPSING BINARY SYSTEMS In the course of our observations of the far ultraviolet spectra of eclipsing binary stars with the IUE satellite, we detected strong emission lines in the spectra of the five binary systems (other than beta Lyr) listed in Table 1. The emission spectrum is richest in SX Cas. For RX Cas and W Ser the spectra are surprisingly similar, while fewer emission lines are observed in W Cru and V367 Cyg. Nevertheless, the basis pattern is the same, and is well represented by the most prominent emission features seen in W Ser, listed in Table 2. In addition to these lines, the spectrum between lambda1200 and lambda2000 A shows over 30 other emission features. Most of these features are blends of two or more components, since the resolving ower of the low-dispersion IUE spectrograph is only 6 A. The high level of ionization indicated by strong emission lines of Si IV, C IV, and N V is remarkable. It is not yet clear from which region or regions of the systems these emissions come. One possibility is a high temperature, low density plasma surrounding the hotter component. It should be noted that the hot component may well be the spectrographically invisible one if it is surrounded by an optically thick disk of gas. All five systems listed in Table I are most likely at a phase of rapid mass transfer or mass loss, so the presence of a large cloud as well as a thick disk is quite plausible. The hotter component may in turn be a collapsed object or a more "normal" accreting star. Emission lines of N V are known to be associated with stars earlier than O9. Although we have detected blue continua in the systems listed in Table 1, they do not seem to correspond to the continua of O stars. Additionally, the total luminosity of each system does not suggest the presence of a normal O star. Therefore the ionization is probably not due to blackbody thermal radiation, but it may still be radiatively induced, if we assume absorption of X rays gener- ated near the secondary star. Their origin is straightforward if we surmise a collapsed object. In the case of a non-collapsed star later than O-type, one can conjecture X radiation generated by accretion shocks. It is also possible that the emitting plasma lies in a very extended chromosphere and/or corona of the mass-losing component, perhaps created by an unusually strong stellar wind. Table 1 Binary Systems Showing Emission Lines of Hot Plasma Name Period (days) f(M)(SUN) Spectral type RX Cas 32.3 0.16: A5e III + (G3 III)? SX Cas 36.6 0.31: A6e III + (G6 III)? W Cru 198.5 5.82 G1e Iab V367 Cyg 18.6 1.56: A5e Iab? (shell!) beta Lyr 12.9 8.5 B8e II W Ser 14.2 0.35: F5e Ib (shell) Note: beta Lyr was not observed in this program. However, previous observations with Copernicus (Hack et al., 1975, Astrophys J. 198, 453) justify its inclusion in our list. Table 2 Major Emission Lines Detected in W Ser 1200A - 1950A Observed Identification Possible Comments (A) Ion (A) Contributors 1213 H I (Ly alpha) 1216 geocoronal 1240 N V(1) 1239, 1243 1263 Si II(4) 1265 1306 O I(2) 1302-1306 Si II(3), [O V]?? 1335 C II(1) 1335, 1336 Si III(34) 1394 Si IV(1) 1394 1401 Si IV(1) 1403 O IV 1401, 1405 1549 C IV(1) 1548, 1551 1641 He II 1641 very broad 1672 Al II(2) 1671 1766 Al II(5) O III, Fe II 1783 Si III(35) Ni II(5) reseau-contaminated 1808 Si II(1) 1808 Ni II(2) 1816 Si II(1) 1817 1856-62 Al III(1) 1855, 1863 Si III, A1 II, Fe III very broad 1888-93 Si III(1) 1892 Fe III, N II?, P IV? very broad 1910 C III 1909 Fe III reseau-contaminated 1925 C II 1927, 1928 Fe III reseau-contaminated Decision about the location of the emitting plasma is not easy since the low dispersion does not permit accurate measurement of the radial velocities. Also, the phase dependence of the phenomenon is unknown at this time. All five systems listed in Table 1 are very peculiar, and, from previous observations with Copernicus, beta Lyr is reasonably included in the list. The presence of a black hole in beta Lyr, W Ser, W Cru, and V367 Cyg is a distinct possibility. On the other RX Cas and SX Cas are probably stronger cases for the stellar wind alternative. It may be dangerously misleading that we concentrated on peculiar systems. Certainly the occurrence of the phenomenon in various classes of binary stars is yet to be tested, however, we have not observed this type of emission spectrum in epsilon Aur, mu Sgr, TT Hya, RZ Sct, U CrB, to mention just a few of the systems already surveyed. MIREK PLAVEC ROBERT H. KOCH University of California University of Pennsylvania Los Angeles Philadelphia