Three permanent magnet arrays were aboard each Viking lander: a strong array fixed on a photometric reference test chart (RTC) on top of the landers, and two other arrays, one strong and one weak, incorporated in the backhoe of the surface sampler. The RTC magnets on both landers have attracted magnetic particles from the dust cloud caused by the retro-rockets on landing and by dust raised in connection with surface sample acquisition and delivery. A considerable amount of magnetic particles has been attracted to both backhoe magnets from the surface material as a result of sample acquisition. We judge that the loose Martian surface material contains from 1 to 7% highly magnetic mineral. Preliminary spectrophotometric analysis shows the material adhering to both weak and strong magnets to be identical and indistinguishable from the normal Martian surface material exposed in trenches. The highly magnetic minearal could be present as discrete grains of iron, magnetite, or pyrrhotite, each with a red ferric iron oxide coating, or as particles of maghemite, &ggr;Fe2O3, with or without other red oxide. Alternatively, the magnetic mineral could be uniformly distributed as a subsidiary component of composite clay or other silicate mineral particles which constitute all or part of the surface material sampled. Mechanical mixtures of the above alternatives are also possible. Although none of the possibilities can be unambiguously excluded, the simplest explanation which fits the available data is that the red pigment in or on all surface particles consists, in part at least, of &ggr;Fe2O3 and that it is the principal source of the magnetic susceptibility of the surface material.