The doll curve reaches its point of highest intensity between eight and nine,[3] and it is nearly ended at fifteen, although it may persist. Children can give no better reason why they stop playing with dolls than because other things are liked better, or they are too old, ashamed, love real babies, etc. The Roman girl, when ripe for marriage, hung up her childhood doll as a votive offering to Venus. Mrs. Carlyle, who was compelled to stop, made sumptuous dresses and a four-post bed, and made her doll die upon a funeral pyre like Dido, after speaking her last farewell and stabbing herself with a penknife by way of Tyrian sword. At thirteen or fourteen it is more distinctly realized that dolls are not real, because they have no inner life or feeling, yet many continue to play with them with great pleasure, in secret, till well on in the teens or twenties. Occasionally single women or married women with no children, and in rare cases even those who have children, play dolls all their lives. Gales's[4] student concluded that the girls who played with dolls up to or into pubescent years were usually those who had the fewest number, that they played with them in the most realistic manner, kept them because actually most fond of them, and were likely to be more scientific, steady, and less sentimental than those who dropped them early. But the instinct that "dollifies" new or most unfit things is gone, as also the subtle points of contact between doll play and idolatry. Before puberty dolls are more likely to be adults; after puberty they are almost always children or babies. There is no longer a struggle between doubt and reality in the doll cosmos, no more abandon to the doll illusion; but where it lingers it is a more atavistic rudiment, and just as at the height of the fever dolls are only in small part representatives of future children, the saying that the first child is the last doll is probably false. Nor are doll and child comparable to first and second dentition, and it is doubtful if children who play with dolls as children with too great abandonment are those who make the best mothers later, or if it has any value as a preliminary practise of motherhood. The number of motor activities that are both inspired and unified by this form of play and that can always be given wholesome direction is almost incredible, and has been too long neglected both by psychologists and teachers. Few purer types of the rehearsal by the individual of the history of the race can probably be found even though we can not yet analyze the many elements involved and assign to each its phyletic correlate.