Kormendy, J., 1977 c, Astrophys. J. 218, 333. Sandage, A., 1961, Hubble Atlas of galaxies, Shobbrook, R. R., 1963, The ObservatolY, 83, lonsdale, C. J., Helou, G., Good, J. C. and Carnegie Institution of Washington Publi­ 36. Rice, W. 1985, "Catalogued galaxies and cation 618. Varnas, S. R., Bertola, F., Galletta, G., Free­ quasars observed in the IRAS Survey", JPl Sandage, A., 1975, in Galaxies and the Uni­ man, K. C. and Carter, D., 1987, Astrophys. publication D. 1932. verse, Sandage, A., Sandage, M. and Kris­ J. 313, 69. Matthews, T.A., Morgan, W. M. and Schmidt, tian, J. eds. University of Chicago Press; de Vaucouleurs, G., 1953, The Observa­ M., 1964, Astrophys. J. 140,35. p.1. tory,73, 252. Möllenhoff, C., 1982, Astron. Astrophys. 108, Sandage, A. and Tammann, G.A., 1981, A de Vaucouleurs, G., 1959, Handbuch der 130. revised Shapley-Ames cata/og of bright Physik 53, 311. Neugebauer, G. et al., 1984, Astrophys. J. galaxies, Carnegie Institution of Washing­ de Vaucouleurs, G., 1987, lAU Symp. 127, 278, l1. ton Publication 635. p.3. Nieto, J.-L. and lelievre, G., 1982, Astron. Schneider, D. P., Gunn, J. E. and Hoessel, Westerlund, B. E. and Smith, L. F., 1964, Aus­ Astrophys. 109,95. J. G., 1983, Astrophys. J. 264, 337. tra/ian J. Phys.. 17, 340. Phillips, M. M., 1979, Astrophys. J. 227, Schweizer, F., 1980, Astrophys. J. 237, 303. l121.

Long-Term Walraven Photometry of Cataclysmic Variables s. VAN AMERONGEN and J. VAN PARAOIJS, Astronomicallnstitute "Anton Pannekoek", University ofAmsterdam, the Netherlands

Ouring the last seven years we have lerian orbit around the white dwarf. It is in a spasmodic way. Long time intervalls been observing cataclysmic variables generally believed that viscous pro­ of the order of months of a low accretion (CV's) with the Walraven photometer cesses give rise to exchange of angular rate (and therefore luminosity) are in­ behind the gO-cm Outch telescope at La momentum, by wh ich an initially formed terrupted by outbursts, lasting for a few Silla. An advantage of this photometrie ring spreads and forms a flat disk-like days to about two weeks, during which system for strongly variable blue sour­ configuration, the accretion disk, the accretion rate increases by a large ces is that it allows simultaneous detec­ around the white dwarf. Matter trans­ factor (see Fig. 2). tion in five passbands, extending from ferred from the secundary gradually spi­ the optical (5400 to the ultraviolet rals inward along quasi-Keplerian orbits. A) Intermediate Polars (3200 A). We have observed these CV's Ouring this diffusion of the particles, the by monitoring them for several hours, other half is radiated away. In these cataclysmic variables the and reduced the measurements diffe­ This radiation from the accretion disk white dwarf has a magnetic field that is rentially with respect to nearby compari­ generally dominates the optical and UV strong enough to dominate the motion son stars. luminosity of cataclysmic variables. The of the inflowing matter within a certain Results obtained so far include mea­ kinetic energy of the inflowing material is distance from its centre. Within this surements of changes in the rotation eventually dissipated close to the white "magnetospheric radius" the matter is periods of accreting white dwarfs in dwarf surface, e. g. in a transition region channeled onto regions near the CV's, quite strong upper limits to secu­ from the accretion disk to the white magnetic poles of the white dwarf. In­ lar changes in optical brightness of dwarf surface. side this magnetosphere an accuration dwarf novae (ON) between outbursts, Based on observational characteris­ disk cannot exist; what fraction of the and the recognition that accretion in­ tics, cataclysmic variables have been accretion disk remains depends on the stabilities similar to ON outbursts also divided into several groups. Ouring our size of the magnetosphere, relative to occur in systems not classified as such. observing project, we have paid most of that of the Roche lobe around the white our attention to the following types of dwarf. cataclysmic variables: Introduction Brightness Variations Cataclysmic variables are close bi­ DwarfNovae nary stars consisting of a white dwarf Most cataclysmic variables with an and a low-mass « 1 MG) near-main­ In the dwarf-nova systems the accre­ accretion disk show brightness varia­ sequence companion. This "secondary" tion rate onto the white dwarf changes tions with the orbital period, wh ich are star fills its so-called Roche lobe, i. e. the critical equipotential surface, outside which matter is no longer bound to the star (see Fig. 1). At the point on the line through the two stellar centres, where the Roche lobes of the secondary and the white dwarf touch each other, mat­ ter from the secondary can easily fall into the gravitational potential weil of the white dwarf. Such matter, flowing from the secon­ dary star, has angular momentum with respect to the white dwarf, due to the orbital revolution of the binary. It will Figure 1: A sketch of the structure ofa cataclysmic binary. Matter from the secondary falls into therefore settle into a more or less Kep- the potential weil of the white dwarf and hits the accretion disk at the bright 01' "hot" spot. 44 3 ~--r------.------r-,-"-~ 3 novae, the SU UMa systems, that show '. two rather different types of outbursts 2 2 (superoutbursts and "normal" out­ bursts). Superoutbursts are different ! . from normal outbursts - in their rate of lL i I· 11 il I I I; --"- ---'- .1....- .-1 oeeurrenee (Iess often), their duration 2445900 2445950 (Ionger), and their maximum luminosity 2446000 (higher). Apart from this, it is found that Figure 2: An overview of the optical observations of the SU UMa-type dwarf nova VW Hyi during superoutburst, the optieal bright­ during July-November 1984, when the source went into a normal outburst 4 times and finally ness is strongly modulated; these so­ went into a superoutburst. Drawn is the Walraven-B-flux in log F", with F,. in units of milli­ Jansky's, against Heliocentric Julian Day. ealled superhumps reeur with aperiod that is slightly longer than the orbital period; also this superhump period is due to a so-ealled "bright spot" at the through the disk to the white dwarf. not exaetly eonstant, but inereases outer edge of the disk, where it is hit by During quieseenee, when the aeeretion through the superoutburst. The the stream of matter flowing in from the rate onto the white dwarf equals the superhumps have always been a big seeondary (see Fig. 1). In systems with a mass transfer rate from the seeondary, problem, whieh resisted theoretieians' magnetized white dwarf the light ean no seeular ehanges are expeeted in the attempts to model them. However, re­ also be modulated with the white dwarf aeeretion disk's state. eently Whitehurst has proposed that the rotation period, due to the higher One of the aims of our long-term superhump phenomenon might be the emissivity from the magnetie pole re­ photometrie projeet of eataelysmie var­ result of a mass-transfer burst from the gions and their general non-alignment iables was an attempt to distinguish be­ seeondary, leading to a slowly preeess­ with the rotation axis of the star. In these tween these two types of model, by ing, more or less eeeentrie disk, as the systems the light ean also be modulated studying the possible seeular brightness result of tidal distortion by the seeon­ with a third period, the beat period be­ variations of dwarf novae in quieseenee. dary of the orbits of partieles in the outer tween the white dwarf rotation period For this purpose we have observed the disk regions [4]. His model sueeessfuily and the binary revolution period, whieh system VW Hyi during 42 nights over a aeeounts for many of the strange prop­ is the rotation period of the white dwarf time interval of 4 months, during whieh 5 erties of the superhumps. in the referenee frame of the binary. outbursts oeeured. These observations Our Walraven observations of the Modulations with the beat period are were eoordinated with observations superoutburst of VW Hyi, wh ich were due to proeessing of eleetromagnetie with EXOSAT, Voyager, IUE, and Walra­ done in five ehannels simultaneously, radiation from the rotating white dwarf ven [3] (see Fig. 2). The main outeome of have led to the diseovery of another by parts of the binary system that this study is that there is no evidenee for strange property of superhumps: the are not distributed axially-symmetrie a signifieant trend in the optieal bright­ whole superhump-modulation light around the white dwarf, e. g. the seeon­ ness from the system between out­ eurve arrives later when seen at longer dary and the bright spot on the aeere­ bursts (see Fig. 4). The same was found wavelengths. The total shift from 3200 A tion disk's outer rim (see Fig. 3 for an in the X-ray and UV observations (if any to 5400 A is + 0.06 in phase of the example of these modulations). change is diseernable it is a seeular superhump period [5]. We eannot offer deerease, not an inerease). This result is an explanation for this finding. not easy to reeoneile with simple ver­ DwarfNovae sions of the disk-instability model. Intermediate Polars At the end of our eampaign VW Hyi Two types of models have been pro­ went into a so-ealled superoutburst. VW In intermediate polars the influenee of posed to explain the outbursts of dwarf Hyi belongs to a subgroup of dwarf the white-dwarf magnetie field does not novae. In the disk-instability model [1] the aeeretion disk ean exist in two diffe­ rent modes, eorresponding to a high and a low value of the mass aeeretion 1 . 3 1 . 3 rate through it. In the quieseent state the aeeretion rate through the disk is small­ ++ er than the the mass transfer rate from 1 . 2 1 .2 the seeondary, so that the density in the disk builds up steadily. Coupled to this seeular inerease of the density one ex­ 1 . 1 1 . 1 peets an inerease of its brightness (mainly in the optieal and UV). When the density has reaehed a partieular value, 1 .0 1 .0 the disk eannot remain in this state of low aeeretion, but suddenly switehes to astate with a high aeeretion rate. During 0.9 + 0.9 this outburst the mass aeeretion rate onto the white dwarf is higher than the '------'-----'---~----'---- mass transfer rate from the seeondary. o. 8 __'----J 0 . 8 When the density has declined suffi­ 5083.7 5083.8 5083.9 eiently, the disk returns to its quieseent Figure 3: Modulations of the Walraven-B-band brightness of the intermediate polar V 1223 Sgr state. In the mass-transfer-instability on 1982 April 24. A long-term modulation with the binary revolution period of 3.4 hours (0.14 model [2], the outbursts are explained day) and a modulation with the beat period between the white rotation period and the orbital by mass-transfer bursts from the seeon­ period, of 13.2 minutes (0.0092 day), can be seen. The flux is given in flux units of a local dary, wh ich are direetly transported comparison star; time is given in HJD-2440000).

45 F­ time scale of the binary; therefore these 0·24 0·24 r + changes in the mass transfer rate 15 0.22r C-22 should occur very often in the lifetime of o.to 0·20 the binary [6]. As there is no reason why 0.18 + + C.18 this should only happen in intermediate 10 0.16 0.16 polars, it is likely to occur in all cataclys­ O-C 0.14 O. 14 mic variables. 5 0.12 0.12 A possible explanation for long-term 0.10 0.10 variations in the mass transfer rate from t 0.08 0.08 the secondary, might be the occurrence 0 o.oG + 0.06 of thermonuclear flashes on the surface 0.04 + 0.C4 of the white dwarf (classical nova out­ + 5 0.02 C.02 bursts), in which about 10- MG is lost. -5 0.00 These classical novae possibly lead to an increasing separation of the two stars -O.Ctr ~:::: and a temporary disconnection of the -1 0 -C.04r tttt+/++tJ+ secondary from its Roche lobe, so that -c.CGi -O·OG the mass transfer process stops (or at 0.0 r -C.C8 least becomes much smaller) [7]. In the o.lol~ G. 10 following stage of "hybernation" of the -1 0 0.[" 0.' 1 .0 cataclysmic variable, the two stars will 10- 5 N Figure 4: VW Hyi. The residual flux in the gradually come closer to one another as Figure 5: The deviations (O-C, in units of Walraven-B band, averaged over one orbital the result of loss of orbital angular 0.0001 day) of the optical beat pulse arrival cycle, relative to the overall average orbital momentum, due to emission of gravita­ times in the intermediate polar V1223 5gr light curve, is given against the relative phase tional radiation and to magnetic braking, from a linear fit against their cycle-number in time between two successive normal out­ (N). The parabolic curve drawn gives an esti­ bursts. Apart from the decline of the flux after and mass transfer will restart when the mate ofthe rate ofincrease of the white dwarf an outburst, no rising or falling trend is seen secondary again touches its Roche rotation period (spin-down). The spin period in the brightness of VW Hyi during the quies­ lobe. is 0.0092 day. cent state of the dwarf nova. 00 All Cataclysmic Variables Burst? extend to the secondary. Consequently, Ouring recent observations of the in­ Uv Amsterdam) collected spectra of the matter spiralling inward in the outer termediate polar TV Col we encountered second outburst. From the spectral line accretion disk region becomes con­ a surprising feature, only seen once be­ information obtained, it became clear nected to the magnetic field (and will fore in this system. Within one week this that this outburst developed in a diffe­ corotate with the white dwarf) only with­ source showed two outbursts, that rent way than normal dwarf nova out­ in the magnetosphere of the white looked remarkably similar [8]. Both bursts do. dwarf. When it is connected, it either bursts lasted less than about half a day, It is attractive to relate the differences looses angular momentum to the white and reached an amplitude of 2 mag­ in the outburst characteristics of the in­ dwarf or gains angular momentum from nitudes; they last therefore much shorter termediate polars and dwarf novae, with it. This accretion torque accelerates or and are less luminous than dwarf nova the difference in their accretion disk decelerates the white dwarf's rotation, outbursts. We have in 1984 also de­ structure. It looks as if the absence of an until the white dwarf has reached an tected a similar short burst in V1223 inner disk in the intermediate polars pre­ equilibrium period at which the net Sgr, with the same characteristics as the cludes the full development of the burst torque vanishes (this equilibrium period bursts in TV Col (see Fig. 6). In addition as areal dwarf-nova outburst. depends on the magnetic field strength to the photometry, the observers (H. It appears valuable to follow the and the mass accretion rate). Schwarz of ESO and M. Heemskerk of cataclysmic (physically) variable From our long-term observations of intermediate polars we found that the white dwarf in V1223 Sgr spins down, while in H2252-035 it spins up (see Fig. 5); Pakull & Beuermann also found a down-spinning white dwarf in H2215­ 086. Previous to our observations, only spin-ups of accreting magnetized white dwarfs had been found. From models which link the accretion torque to the 1'1it" rate of change of the spin period of the "f" . white dwarf, we found that the white i ' dwarfs do not spin close to their equili­ brium spin rate. One way in wh ich these white dwarfs could be unable to reach their equilibrium spin period, is by large Figure 6: 5pectral continuum flux, close, to the Hydrogen-alpha emission line, of the intermedi­ changes in the mass accretion rate. We ate polar V1223 5gr during six successive nights of observations with the Boiler & Chivens estimate that such equal changes in the spectrograph behind the 1.S-m E50 telescope. Flux is given in linear arbitrary units against mass transfer rate take place on time time in units of days. The decline of a burst is visible in the 4th night of observations (on 1984 scales less than 105 - 6 years, wh ich is Aug 31). Ouring the declination of the flux, the more or less sinusoidal orbital modulation of the very short compared with the evolution light is visible. 46 binaries, to see how they behave on a [3] Pringle, J. E., Bateson, F. M., Hassall, Honey, W. B., Charles, P.A., Whitehurst, R., long time scale. The spasmodic emis­ B.J.M., Heise, J., Holberg, J.B., Polidan, Barrett, P. E., and Smale, A. P., 1988. Mon. sion of radiation due to the unstable R. S., van Amerongen, S. F., van der Not. R. Astr. SOG., 231, 1. mass transfer/accretion process, makes Woerd, H., van Paradijs, J., and Ver­ [5] van Amerongen, S., Bovenschen, H., and it necessary to open the observing win­ bunt, F., 1987. Mon. Not. R. Astr. SOG., van Paradijs, J., 1987. Mon. Not. R. Astr. 225,73. SOG., 229, 245. dow a bit more in order to detect crucial van Amerangen, S., Damen, E., Graot, M., [6] van Amerongen, S., Augusteijn, T., and phenomena at the right moment. From Kraakman, H., and van Paradijs, J., 1987. van Paradijs, J., 1987. Mon. Not. R. Astr. the long-term behaviour of these sys­ Mon. Not. R. Astr. SOG., 225, 93. SOG., 228, 377. tems we have probably been able to Verbunt, F., Hassall, B.J. M., Marang, F., [7] Shara, M. M., Livio, M., Moffat, A. F. J., derive information on their evolutionary Pringle, J. E., and Warner, B., 1987. Mon. and Orio, M., 1986. Astrophys. J., 311, scenario. Not. R. Astr. SOG., 225, 113. 163. Polidan, R. S., and Holberg, J. B., 1987. Mon. [8] Schwarz, H. E., and Heemskerk, M. H. M., Reterences Not. R. Astr. SOG., 225, 131. 1987. lAU. Circular No. 4508. van der Woerd, H., and Heise, J., 1987. Mon. Schwarz, H. E., van Amerangen, S. F., [1] Osaki, Y., 1974. Publ. Astron. SOG. Japan, Not. R. Astr. SOG., 225, 141. Heemskerk, H. M., and van Paradijs, J., 26,429. [4] Whitehurst, R., 1987. In preparation (pre­ 1988. Astr. Astrophys., submitted. [2] Bath, G. T., 1975. Mon. Not. R. Astr. SOG., print). 171,311.

IRAS Molecular Clouds in the Hot Locallnterstellar Medium

1 1 1 1 P. ANOREAN/ ,2, R. FERLET , R. LALLEMENT ,3 andA. V/OAL-MAOJAR

1 Institut d'Astrophysique, CNRS, Paris, France 2 ESA Fellow on leave trom Dipartimento di Fisica, Universita di Roma, La Sapienza, Roma, Italy 3 Service d'Aeronomie, CNRS, Verrieres le Buisson, France

1. Introduction 2. The Observations associated to the cold interstellar mate­ The picture of the nearby (d :s 100 pe) rial with temperatures ranging from 14 From the IRAS HCON 1 survey (see interstellar medium, as resulting from to 40 K. IRAS Explanatory Supplement, 1984), the observations in the wavelength In order to understand the physical we selected those clouds at 100 f.lm, range from X-ray to radio, consists of a properties of the local medium and their which were already detected at the CO very hot, low-density gas, forming a hot effect on other observations, it is inter­ band (Magnani et al., 1985) and located bUbble, filled with many cloudlets of esting to tackle the problem of the coex­ at high galactic latitude (I b I ~ 25°). ionized and neutral material and located istence of a cold and neutral gas with a The MIDAS software package (ESO near to an interstellar condensation (see hot medium and the mixing between Operating Manual No. 1, 1984) has been the "Local Interstellar Medium", NASA these two gaseous phases. used to analyse the IRAS maps. The CP-2345, 1984; and Cox and Reynolds, The distance, the morphology and whole procedure of IRAS images analy­ 1987). other properties of the clouds, belonging sis will be published elsewhere (An­ Only recently the existence of a colder to the local interstellar medium, can be dreani et al., 1988). counterpart in the nearby gas was determined by mapping the interstellar Table 1 lists the known properties of argued by Magnani, Blitz and Mundi absorption lines toward stars at different the clouds from infrared and CO mea­ (1985), who reported the detection of a distances and projected along the line surements. The position, photometry, large number of "Iocal" molecular of sight to these c1ouds. spectral type and distance are taken clouds at high galactic latitude and in­ The feasibility of this procedure has from the Bright Star Catalogue for the ferred from statistical arguments an av­ already been demonstrated by Hobbs et brightest programme stars, and from erage distance of about 100 pe. al. (1986) who estimate the distance of the HD Catalogue for the others. Stars A comparison with the IRAS maps the cloud Lynds 1457/8 at about 65 pe. were chosen to be bright, hot, and of showed that all the high latitude We selected several high latitude early spectral type. molecular clouds can be identified with elouds detected by IRAS and/or at CO High-dispersion Ca 11 K and Na ID the cores of the new discovered infrared wavelength (2.6 mm) by Magnani, Blitz spectra were gathered during (60 and 10 ~lm) features: "the IRAS and Mundi (1985) and gathered eehelle 1986-1987 with the Coude Echelle cirrus" (Weiland et al., 1986). spectra of a few stars with the ESO CAT Spectrometer fed by the 1A-m CAT Nearby molecular clouds are then telescope at La Silla, Chile. telescope and equipped with either a

TABlE 1. Infrared and CO Properties of the Clouds

Cloud Coordinates IRAS CO

# ~l ~l Name a b I b 12 ~l 25 ~l 60 100 Td Ta (h) (0) (0) (0) (MJy/sr) (K) (K)

20 l1642 433 -1420 210.9 -36.5 - - .3 ± .2 11.2 ± 2.8 20 6.8 126 Q-Oph 1616.3 -1948 355.5 -21.1 - - 10 ± 3 12 ±3 29 9.2 113 - 1517.1 -2925 337.8 -23.04 - - 13 ± 4 12 ±4 32 6.1

47