They had finished the walk about a quarter of an hour before I
        came there. (Because the horse’s droppings at this point
        were quite fresh; covered with flies; not dried outside by
        the sun.)

    They had been cantering up to the point where they began the
        walk, but one horse had shied violently on passing the
        invalid in the rickshaw. (Because there was a great kick up
        of gravel and divergence from its track just where the
        rickshaw track bent into the side of the road, and
        afterwards overrode the horse’s tracks.)

    NOTE.—­I might have inferred from this that the invalid was
  carrying an umbrella which frightened the horse, and was,
  therefore, a lady.  But I did not think of it at the time and had
  rather supposed from the earliness of the hour that the invalid
  was a man.  Invalid ladies don’t, as a rule, get up so early.

Deduction

    The tracks were those of a lady and gentleman out for a ride,
        followed by her dog.

    Because had the horses been only out exercising with syces they
  would have been going at a walk in single file (or possibly at a
  tearing gallop).

    They were therefore ridden by white people, one of whom was a
  lady; because, 1st, a man would not take a big, heavy dog to pound
  along after his horse (it had pounded along long after the horses
  were walking); 2nd, a man would not pull up to walk because his
  horse had shied at a rickshaw; but a lady might, especially if
  urged to do so by a man who was anxious about her safety, and that
  is why I put them down as a man and a lady.  Had they been two
  ladies, the one who had been shied with would have continued to
  canter out of bravado.  And the man, probably, either a very
  affectionate husband or no husband at all.

    NOTE.—­I admit that the above deductions hinge on very
  little—­one link might just be wrong and so break the whole chain. 
  This is often, indeed generally, the case, and corroborative
  evidence should always be sought for.

    In the present instance my deductions proved pretty correct.  I
  saw the couple later on, followed by their collie dog, riding
  along a lower road; but I could not determine their relationship
  to one another.

Note on Examples I. and II.

    Incidentally, the horse-tracks of No. 2 gave me a clue to the
  hour at which the invalid in the rickshaw had passed that way. 
  Thus:  I came on the droppings at 7.14.

    Assuming that they were actually 15 minutes old and the horses
  had walked 1/4 mile since passing the rickshaw, 19 minutes must
  have elapsed since the passing; i.e. they passed each other at
  6.55.

    On my arrival at the point where they had passed, the rickshaw
  would now be 23 minutes ahead of me, or about 11/4 mile.