dhood which is as characteristic as the one we have represented for
the phase of development of puberty. This object selection proceeds in
such a manner that all the sexual strivings proceed in the direction of
one person in whom they wish to attain their aim. This is then the
nearest approach to the definitive formation of the sexual life after
puberty, that is possible in childhood. It differs from the latter only
in the fact that the collection of the partial impulses and their
subordination to the primacy of the genitals is very imperfectly or not
at all accomplished in childhood. The establishment of this primacy in
the service of propagation is therefore the last phase through which the
sexual organization passes.

*The Two Periods of Object Selection.*--That the object selection takes
place in two periods, or in two shifts, can be spoken of as a typical
occurrence. The first shift has its origin between the age of three and
five years, and is brought to a stop or to retrogression by the latency
period; it is characterized by the infantile nature of its sexual aims.
The second shift starts with puberty and determines the definitive
formation of the sexual life.

The fact of the double object selection which is essentially due to the
effect of the latency period, becomes most significant for the
disturbance of this terminal state. The results of the infantile object
selection reach into the later period; they are either preserved as such
or are even refreshed at the time of puberty. But due to the development
of the repression which takes place between the two phases they turn out
as unutilizable. The sexual aims have become softened and now represent
what we can designate as the _tender_ streams of the sexual life. Only
psychoanalytic investigation can demonstrate that behind this
tenderness, such as honoring and esteeming, there is concealed the old
sexual strivings of the infantile partial impulses which have now become
useless. The object selection of the pubescent period must renounce the
infantile objects and begin anew as a sensuous stream. The fact that the
two streams do not meet often enough has as a result that one of the
ideals of the sexual life, namely, the union of all desires in one
object, can not be attained.

THE SOURCES OF THE INFANTILE SEXUALITY

In our effort to follow up the origins of the sexual impulse, we have
thus far found that the sexual excitement originates (_a_) as an imitation