The length of time which it is possible to attend to the same object or
idea may be reckoned in seconds. It is impossible to hold the attention
on an object for any appreciable length of time. In order to hold the
attention the object must change. The simple experiment of trying to pay
attention to a blot of ink or the idea of bravery proves that change is
necessary if the attention is not to wander. What happens is that either
the attention goes to something else, or that you begin thinking about
the thing in question. Of course, the minute you begin thinking, new
associations, images, memories, come flocking in, and the attention
occupies itself with each in turn. All may concern the idea with which
you started out, but the very fact that these have been added to the
mental content of the instant makes the percept of ink blot or the
concept of bravery different from the bare thing with which the
attention began. If this change and fluctuation of the mental state does
not take place, the attention flits to something else. The length of
time that the attention may be engaged with a topic will depend, then,
upon the number of associations connected with it. The more one knows
about a topic, the longer he can attend to it. If it is a new topic, the
more suggestive it is in calling up past experience or in offering
incentive for experiment or application, the longer can attention stay
with it. Such a topic is usually called "interesting," but upon analysis
it seems that this means that for one of the above reasons it develops
or changes and therefore holds the attention. This duration of attention
will vary in length from a few seconds to hours. The child who is given
a problem which means almost nothing, which presents a blank wall when
he tries to attend to it, which offers no suggestions for solution, is
an illustration of the first. Attention to such a problem is impossible;
his attention must wander. The genius who, working with his favorite
subject, finds a multitude of trains of thought called up by each idea,