death rate by what it is willing to pay, so it can by repressive
measures regulate its desertion rate. But measures that keep the
would-be deserter in the home which constantly grows less of a home,
simply through fear of consequences if he left it, seem hardly a
desirable form of prevention from the social point of view. It would be
much better to catch the disintegrating family in whatever form of
social drag-net could be devised, and deal with it individually and
constructively along the lines which case work has laid down.

Is it possible, however, to recognize a "pre-desertion state?" And if
so, what are the danger signals? One case worker answers this question
sententiously: "Any influences which tend to destroy family solidarity
are possible signs of desertion." Another writes: "We have sometimes
found it possible to recognize a 'pre-desertion state' in the
intermittent deserter, where we know the conditions which previously led
to desertion, but I doubt whether we have very often been able to note
it in the case of first desertions. In general, I should say a growing
carelessness or a growing despondency as to his ability to care for his
family are danger signals in the man, of which it is well to keep
track."

The conditions listed in Chapter II as "contributory factors" might in
certain combinations be decided danger signals of impending desertion.
Non-support itself is, indeed, one of the most common of such signals,
though a man who has dealt with hundreds of desertion cases maintained
recently that the best and most hopeful type of deserter is the one who
supports his family adequately up to the time of leaving home.

In the following case the items that led the case worker to suspect an
approaching desertion are set down in the order stated by her. The