Résumé
We reconsider the rejuvenation mechanism as proposed by Shull, Fesen, & Saken (1989). These authors suggest that an active pulsar can catch up with, and rejuvenate the shell of the associated supernova remnant. The morphology of the SNRs G5.4-1.2 and CTB80 seem to confirm this rejuvenation mechanism. The spindown energy is deposited by the pulsar as a relativistic pulsar wind, and has a sufficient power to explain the observed radio emission observed in these remnants. Shull et al. (1989) did ıt not explain the observed lengthscales of the rejuvenated parts of the SNR shell. therefore one needs to consider the diffusive transport of the injected electrons by the pulsar wind. We propose to apply a diffusion mechanism as introduced by Jokipii (1987), which makes a distinction between diffusion along the magnetic field lines and perpendicular to the magnetic field lines, parameterised by the gyro factor$\eta$ . We show that one has to assume a high value for the gyro factor,$\eta\simeq 10^3-10^4$ , i.e. diffusion of the electrons along the magnetic field line is much faster then perpendicular to the magnetic field line, in order for the rejuvenation mechanism to work on the observed lengthscales.