A RADIO SURVEY FOR PULSAR WIND NEBULAE D. A. FRAIL

We have imaged the -elds of 35 pulsars with the at 8.4 GHz to search for evidence that the VLA1 pulsars have deposited energy into their surroundings in the form of an extended wind nebula. Candi- dates have both a wide range of transverse velocities and rotational energy losses. Of the 35 pulsars that we imaged, no pointlike or extended emission was detected toward 21 sources, to a typical rms surface brightness sensitivity of 40 kJy beam~1, for a beam size of Pointlike emission was detected toward 0A.8. 14 pulsars with peak Nux densities between 0.2 and 3 mJy. After a careful examination of each source we conclude that these detections are simply the time-averaged pulsed Nux density from the pulsars them- selves, and thus no new pulsar wind nebulae were discovered. From the surface brightness limits and an estimate of the size of the region where the pulsar wind is -rst shocked, we deduce that the fraction of the pulsarsI spindown energy going into producing a compact radio nebula must be less than 10~6 .I n contrast, this fraction is 2 orders of magnitude larger for the six pulsars which are known unam- biguously to have a radio pulsar wind nebula around them. Several explanations are given for this behavior, and we conclude that only young pulsars with high rotational energy losses produce an observable radio nebula. Subject headings: pulsars: general E radio continuum: ISM E stars: mass loss E supernova remnants