This paper discusses Chapman's suggestion that a stream of corpuscles, emitted from the Sun by a flare, should be spectroscopically detectable before and during the magnetic storm which it is thought to cause. After an investigation of the mode of expansion and the state of ionization of such a stream, it is estimated that the particles in it may cause abnormal absorption near the H and K lines of calcium which is detectable soon after the end of the flare. Should this absorption not be observable, however, the failure will not necessarily disprove the corpuscular theory of magnetic storms, as the depth of absorption may lie below the observational limit.