The Mass of a Millisecond Pulsar

We report on nearly two years of timing observations of the low-mass binarymillisecond pulsar, PSR J1909-3744 with the Caltech-Parkes-Swinburne RecorderII (CPSR2), a new instrument that gives unprecedented timing precision. Dailyobservations give a weighted rms residual of 74 ns, indicating an extremely lowlevel of systematic error. We have greatly improved upon the previous parallaxand proper motion measurements of PSR J1909-3744, yielding a distance of$1.14^{+0.04}_{-0.03}$ kpc and transverse velocity of $(200^{+7}_{-6})$ kms$^{-1}$. The system's orbital eccentricity is just 1.35(12)$\times10^{-7}$,the smallest yet recorded. Since their discovery, the masses of the rapidlyrotating millisecond pulsars have remained a mystery, with the recyclinghypothesis arguing for heavy objects, and the accretion-induced collapse of awhite dwarf more consistent with neutron stars less than the Chandrashkarlimit. Fortuitously, PSR J1909-3744 is an edge-on system, and our data haveallowed the measurement of the range and shape of the Shapiro delay to highaccuracy, giving the first precise determination of a millisecond pulsar massto date, $m_p = (1.438 \pm 0.024) M_\odot$. The mass of PSR J1909-3744 is atthe upper edge of the range observed in mildly recycled pulsars in doubleneutron star systems, consistent with the the recycling hypothesis. It appearsthat the production of millisecond pulsars is possible with the accretion ofless than $0.2 M_\odot$.Comment: 12 pages, 3 figures, accepted for publication in ApJ