"... In PAGE 6: ... The pulsar names will eventually be extended to four declination digits when better positions become available. For all pulsars except PSRs J0621+1002 and J1022+1001 (see Table4 ), the quoted positions are based on the discovery observations and are uncertain by about 50 in each coordinate. For some pulsars we have collected limited timing data at Arecibo (using the search system and folding the data o -line) and at Jodrell Bank (as described in x 4.... In PAGE 8: ...Weisberg 1989). Best- t parameters for the pulsar and its orbit are given in Table4 . The mass function of PSR J0621+1002 is f1(m1; m2; i) = (m2 sin i)3 (m1 + m2)2 = x3 (Pb=2 )2 1 T = 0:027 M ; (1) where m1 and m2 are the pulsar and companion mass in solar units, i is the orbital inclination, Pb is the orbital period, x is the light travel time across the projected orbital semi-major axis, T GM =c3 = 4:925490947 s, G is the Newtonian gravitational constant, and c is the speed of light.... In PAGE 9: ... The shape changes induce systematic biases in the timing measurements on the same time scale. The uncertainties quoted in Table4 allow for this e ect and for a possible contamination due to unmeasured proper motion. We believe they represent realistic \1 quot; uncertainties.... In PAGE 9: ... The post- t weighted rms timing residual for our solution, based on daily-averaged TOAs, is 5 s. The orbital parameters in Table4 yield a mass function for PSR J1022+1001 of f1 = 0:083 M , implying a minimum companion mass m2 = 0:73 M and a probabilistic median m2 = 0:87 M , for an assumed pulsar mass of 1:4 M . Some implications of this unusually large companion mass are discussed in the next section.... In PAGE 13: ...0 2.4 aThese pulsars are in binary systems: see Table4... ..."

"... In PAGE 55: ... PSR 2127+11C. This system appears to be a clone of the Hulse-Taylor binary pulsar, with very similar values for orbital period and eccentricity (see Table7 ). The inferred total mass of the system is 2:706 0:011M .... ..."

"... In PAGE 55: ... PSR 2127+11C. This system appears to be a clone of the Hulse-Taylor binary pulsar, with very similar values for orbital period and eccentricity (see Table7 ). The inferred total mass of the system is 2:706 0:011M .... ..."

"... In PAGE 10: ...elation can be tested by observation. Lorimer et al. (1995), Rappaport et al. (1995) and Burderi, King amp; Wynn (1996) show that while the relation is consistent with the data for a majority of the 25 relevant systems, there are several systems (currently 3 or 4; see Fig. 4 and Table3 ) for which the white dwarf mass is probably too large to t. While one might possibly exclude B0655+64 because of its long spin period Ps (but see below), the other three systems are clearly genuine millisecond pulsars.... In PAGE 10: ... Table 4 and Fig. 4 show that early massive Case B evolution with qi larger than the critical value can explain the discrepant pulsar systems of Table3 . Indeed it appears that the process can end with very short orbital periods, o ering an alternative to the usual assumption of a common-envelope phase (Bhattacharya, 1996; Tauris, 1996) The limiting factor for this kind of evolution may be the so{called `delayed dynamical instability apos; (Webbink, 1977; Hjellming, 1989).... In PAGE 11: ... The short-dashed part of this curve is realised only if the neutron star can expel the transferred mass despite the binary being subject to the delayed dynamical mass transfer instability. Also shown are the (P, M2) values of the four `discrepant apos; ms-pulsar binaries listed in Table3 , and the current position of Cygnus X-2. c... In PAGE 12: ...e.g. Webbink 1984) with the values for M2i and M2f = M2 given in Table 4, M1 = 1:4M and CE = 0:5, where 0:5 is a structural parameter and CE the common envelope e ciency parameter de ned by Webbink (1984), we nd that the nal orbital period is 0:04 lt; Pf(d) lt; 0:2 if mass transfer set in when the donor was already near the Hayashi line, and smaller still if mass transfer set in earlier or if CE is smaller than unity. On the other hand, unless CE is signi cantly smaller than unity, common envelope evolution starting from a system with the donor on the asymptotic giant branch would end with periods much longer than those of the systems listed in Table3 and shown in Fig. 4.... In PAGE 12: ... We conclude therefore that even very rapid mass transfer on to a neutron star does not necessarily result in a common envelope (cf King amp; Begelman, 1999). We note nally that all of the pulsars of Table3 have spin periods much longer than their likely equilibrium periods (i.e.... ..."

"... In PAGE 12: ... 3. Radio pulsar masses The parameters of 47 radio pulsar binary systems are listed in Table1 . At least ve of these pulsars have neutron star companions, so a total of 52 neutron stars are known in radio pulsar binaries.... In PAGE 33: ... Table1 |Continued PSR name Pspin (s) Pb (d) e Notes J1804?2717 0.009343 11.... ..."

"... In PAGE 45: ... The variation in the moment of inertia a ects the spin rate of the pulsar, while the variation in the mass can a ect the orbital period in a manner that can add to or subtract from the direct e ect of a variation in G, given by _ Pb=Pb = ?1 2 _ G=G. Thus, the bounds quoted in Table5 for the binary pulsar PSR 1913+16 and the pulsar PSR 0655+64 are theory-dependent and must be treated as merely suggestive. 3.... In PAGE 52: ... This e ect was previously considered unimportant when _ Pb was known only to 10 percent accuracy. Damour and Taylor carried out a careful estimate of this e ect using data on the location and proper motion of the pulsar, combined with the best information available on galactic rotation, and found _ P GAL b apos; ?(1:7 0:5) 10?14 : (50) Subtracting this from the observed _ Pb ( Table5 ) gives the residual _ P OBS b = ?(2:408 0:010[OBS] 0:005[GAL]) 10?12 ; (51) which agrees with the prediction, within the errors. In other words, _ P GR b _ P OBS b = 1:0023 0:0041[OBS] 0:0021[GAL] : (52) The parameters r and s are not separately measurable with interesting accuracy for PSR 1913+16 because the orbit apos;s 47 inclination does not lead to a substantial Shapiro delay.... ..."

"... In PAGE 2: ...ig. 1. Scatter plot of brightness in VHE gamma rays and rotational energy loss of pulsars. Table2 . Observations of the Crab, PSR B 1706?44, Vela and Geminga Group (reference) Flux (cm?2 s?1) E( ) th ============ Crab ============ | unpulsed | Whipple (1993) 8.... ..."

"... In PAGE 21: ... Of these, about half are neutron star-white dwarf binaries in which the mass determination depends on the validity of the Pb{m2 relation and the isotropy of the binary orbits with respect to the line-of-sight, as discussed in the previous section. All other mass constraints are listed in Table3 and shown in Figure 5. Although we defer a full discussion of the underlying neutron star mass distribution and it implications for neutron star formation and evolution until after analysis of the X-ray binary systems (Paper II), we note here a few points of particular interest about the radio pulsar binaries.... ..."