We have just pointed out the importance of reserving the schools for defectives for improvable cases. But it is necessary to correct this word "improvable," because all defectives can be improved more or less. Their asserted _arrest_ of development is not complete, and the expression is equivocal. It would be better to replace the word "improvable" by the following more precise phrase: "Capable of being taught to gain, in part, their own living." Which of them are in this position? Unfortunately, we do not know. All such questions should have been solved long ago, since thousands of defectives have passed into the hospices. It would have been enough to have followed them up, to have found out what became of them, and to have drawn conclusions. But this has never been done methodically, and for the present we are reduced to conjecture. The nearest estimate we can form is that the social value of any individual case, not epileptic, is in inverse proportion to the degree of deficiency; the imbecile would seem to be more improvable than the idiot, and the feeble-minded than the imbecile. But this is simply hypothesis, and we accept it quite provisionally, until exact investigations have been made which will permit us to replace conjecture by demonstrated truth. Consequently we shall open wide the doors of the school to the feeble-minded and close them to the idiots, while as to the imbeciles, we shall have to find out whether the proper place for them is the school or the hospice. It will be necessary to find out in what measure, and at the price of what effort, an imbecile can be instructed to the point, say of being able to read. There are two other indications which may help us. Cases of _acquired_ mental deficiency--that is to say, cases who have become defective as the result of something which affected them after birth--are usually less improvable than _congenital_ cases, or cases where the deficiency is due to some cause acting before birth. And, secondly, cases affected by epilepsy, with fits or frequent attacks of vertigo, usually undergo a progressive mental deterioration. What distinctions can we draw between the different degrees of mental deficiency? Such a question, we think, might be asked with regard to the ill-balanced as well as the defective. With respect to the former, we have no criterion at present to offer. It will be enough to pick out and send to the hospices _the most ill-balanced_, those whose presence among normal children would be a danger owing to the perversion of their instincts or the brutality of their impulses.