A complex event observed in the radio pulses from the Crab pulsar in 1997 included echoes, a dispersive delay, and large changes in intensity. It is shown that these phenomena were due to refraction at the edge of a plasma cloud in the outer region of the Crab Nebula. Several similar events have been observed, although in less detail. It is suggested that the plasma cloud is in the form of filaments with diameter around 3 x 10^11m and electron density of order 10^4 cm-3

In separate series of observations of the Crab pulsar, pulse broadening due to scattering was measured at 111 MHz, and variations of dispersion due to pulse delay were measured at higher radio frequencies. In a remarkable event lasting 200 days a large increase occurred in both parameters and with similar time signatures. The increases in scattering and dispersion measure observed over the 200 days MJD 53950 - 54150 are attributable to the effects of an ionised cloud or filament crossing the line of sight. The cloud would be $10^11 - 10^12 m across, with electron density 10^3 - 10^4 cm^-3. The increased scattering might originate within the cloud itself, or the moving filament might induce turbulence in a separate higher density cloud in the line of sight.

The fourteen glitches observed during 33 years do not show the simple pattern expected from a relaxation oscillator. They may however be regarded as three major events separated by about 12 years, the third being a group of smaller glitches. There is a step increase in slowdown rate at each glitch, whose cumulative effect makes a significant contribution to the second differential nu-ddot. The braking index "n" has previously been evaluated only between glitches: the effect of the glitches is to reduce "n" from 2.51 to 2.45. This extra effect due to the glitches would be explained by an increase in dipole field at the fractional rate of 1.5x10^-5 per annum.

Gamma ray burst sources are isotropically distributed. They could be located at distances $\sim 1000$ AU. (Katz \cite{JK92}) GRB signals have many narrow peaks that are unresolved at the millisecond time resolution of existing observations. \cite{JK87} CETI could use stars as gravitational lenses for interstellar gamma ray laser beam communication. Much better time resolution of GRB signals could rule out (or confirm?) the speculative hypothesis that GRB = CETI.