The 3rd Light Cavalry was the regiment directly involved in the firing parade and its consequences and one would have expected that, when the outbreak came, it would have begun among them: indeed the outbreak is often vaguely ascribed to them, with the implication that the native infantry joined in. It comes as somewhat of a surprise to find that the evidence points to the disturbance in the cavalry lines having followed, not preceded, that in the infantry lines. This seems to be indicated, somewhat obscurely, by two out of the three deponents from the cavalry regiment. More significant, perhaps, is the fact that none of the infantry deponents alleges that the infantry broke out following upon some disturbance in the cavalry lines, and if the facts had not been notoriously against it this would surely have been suggested: as it is, all infantry witnesses make their outbreak follow upon the bazar outcry and do not mention the cavalry at all. Once the cavalry had risen, they took to their horses, and the consequence is that their actions then ranged farther and moved faster than those of the infantry. This increases the difficulty of co-ordinating in point of time what was occurring in the native infantry and cavalry lines respectively.

There is no clear statement about the initial stages of the outbreak of the cavalry. It seems to have started about 5.30 p.m., even a little after, the time when the officers of the 20th N.I. were getting down to the lines.