We model the population of double white dwarfs in the Galaxy and find a better agreement with observations compared to earlier studies, due to two modifications. The first is the treatment of the first phase of unstable mass transfer and the second the modelling of the cooling of the white dwarfs. A satisfactory agreement with observations of the local sample of white dwarfs is achieved if we assume that the initial binary fraction is ∼50% and that the lowest mass white dwarfs (M < 0.3 M) cool faster than the most recently published cooling models predict. With this model we find a Galactic birth rate of close double white dwarfs of 0.05 yr −1 , a birth rate of AM CVn systems of 0.005 yr −1 , a merger rate of pairs with a combined mass exceeding the Chandrasekhar limit (which may be progenitors of SNe Ia) of 0.003 yr −1 and a formation rate of planetary nebulae of 1 yr −1. We estimate the total number of double white dwarfs in the Galaxy as 2.5 10 8. In an observable sample with a limiting magnitude V lim = 15 we predict the presence of ∼855 white dwarfs of which ∼220 are close pairs. Of these 10 are double CO white dwarfs of which one has a combined mass exceeding the Chandrasekhar limit and will merge within a Hubble time.

Abstract.We review the properties of Galactic binaries containing two compact objects, as derived by means of population synthesis. Using this information we calculate the gravitational wave signal of these binaries. At frequencies below f < ∼ 2 mHz the double white dwarf population forms an unresolved background for the lowfrequency gravitational wave detector LISA. Above this limit some few thousand double white dwarfs and few tens of binaries containing neutron stars will be resolved. Of the resolved double white dwarfs ∼500 have a total mass above the Chandrasekhar limit. About ∼95 of these have a measurable frequency change allowing a determination of their chirp mass. We discuss the properties of the resolved systems.

Abstract.We study two models for AM CVn stars: white dwarfs accreting (i) from a helium white dwarf companion and (ii) from a helium-star donor. We show that in the first model possibly no accretion disk forms at the onset of the mass transfer. The stability and the rate of mass transfer then depend on the tidal coupling between the accretor and the orbital motion. In the second model the formation of AM CVn stars may be prevented by detonation of the CO white dwarf accretor and the disruption of the system. With the most favourable conditions for the formation of AM CVn stars we find a current Galactic birth rate of 6.8 10 −3 yr −1 . Unfavourable conditions give 1.1 10 −3 yr −1 . The expected total number of the systems in the Galaxy is 9.4 10 7 and 1.6 10 7 , respectively. We model very simple selection effects to get some idea about the currently expected observable population and discuss the (quite good) agreement with the observed systems.