Recent pulsar surveys have increased the number of observed double neutron stars ( DNS ) in our galaxy enough so that observable trends in their properties are starting to emerge .
In particular , it has been noted that the majority of DNS have eccentricities less than 0.3 , which are surprisingly low for binaries that survive a supernova explosion that we believe imparts a significant kick to the neutron star .
To investigate this trend , we generate many different theoretical distributions of DNS eccentricities using Monte Carlo population synthesis methods .
We determine which eccentricity distributions are most consistent with the observed sample of DNS binaries .
In agreement with Chaurasia & Bailes ( 2005 ) , assuming all double neutron stars are equally as probable to be discovered as binary pulsars , we find that highly eccentric , coalescing DNS are less likely to be observed because of their accelerated orbital evolution due to gravitational wave emission and possible early mergers .
Based on our results for coalescing DNS , we also find that models with vanishingly or moderately small kicks ( \sigma \lesssim 50 km s ^ { -1 } ) are inconsistent with the current observed sample of such DNS .
We discuss the implications of our conclusions for DNS merger rate estimates of interest to ground-based gravitational-wave interferometers .
We find that , although orbital evolution due to gravitational radiation affects the eccentricity distribution of the observed sample , the associated upwards correction factor to merger rate estimates is rather small ( typically 10-40 % ) .