In 1997 October daily monitoring observations of the Crab pulsar at 327 MHz and 610 MHz with an 85ft telescope in Green Bank, WV showed a jump in the dispersion measure by 0.1 cm$^{-3}$ pc. Pulses were seen simultaneously at both old and new dispersions over the course of several days. During the dispersion jump the pulsed flux diminished by an order of magnitude. In the months before this event the average pulse profiles contained faint ``ghost'' pulse components offset in phase from the regular main pulse and interpulse components by a nearly frequency independent time delay that quadratically diminished to zero over a month. After the dispersion event there was an order of magnitude increase in the level of scattering, as measured by pulse broadening at 327 MHz. There was also a curious shift in the rotational phase, a slowing down, at both frequencies at the time of the dispersion jump which we associate with intrinsic timing noise. All of the observed phenomena except this slowing down can be explained by the variable perturbing optics of a triangular plasma prism that is located in the filamentary interface between the synchrotron nebula and the supernova ejecta and which crosses the line of sight over a period of months. The required density, scale length and velocity are reasonable given previous observations and analysis of these filaments. Our study thus provides a probe of the plasma column on scales of 30 microarcsecond to 3 milliarcsecond ($10^{12-14}$ cm) which complements scales accessible to optical emission line studies with HST resolution ($10^{16-18}$ cm). In combination both observations provide a detailed look at a sample of the interface region that can be matched statistically to the results of numerical simulations.