turned "Watts" also into fák. It is possible, as M. Taine suggests, that to her there is some shade of difference in the sounds which escapes adult ears. At twenty months twenty-five days she said vats or váts. "Walk" has its proper sense as a mode of motion, opposed to riding, in perambulator for herself, or in carriage for others. She is much interested in watching callers going away, and says to them dyi dyi or zhi zhi (gee-gee). . . wák, as if to ask how they mean to go; or, perhaps, merely to show her knowledge. Sometimes she begins to say tá tá to a visitor, not that she is tired of his or her presence, but that she wants the amusement of seeing the departure.
She has learned to repeat no no after she has been told not to do something, as an act of assent to the prohibition, and she seems to take pleasure in saying no no to the cat.
Twenty months. Dash or dásh dust. Tá'sh or tá'sh, learned, I think, from "touch," one day repeated several times without assignable meaning, and then dropped. Tásh, however, is adopted for [mus]tache. N. B.—Final sibilants are more under command than initial. Final g now produced: geg fizgig (toy so called).
At this time a sudden advance was made to dissyllables. Several words were produced with success on or about the same day: "Fanny, honey, money" (these two learned from the rhyme of "Sing a song of sixpence"), very distinct. "Money," however, seems to be confused with "moon:" when told to say moon she says money. Others are attempted with more or less success: as fá-wá, flower; la-ta, letter; ha-pi, happy (taught her as opposite of "poor," but I doubt if she sees the meaning. She has taken up ha-pi to stand for "empty," which we tried to teach her, and in that sense uses it without prompting). Bá-ta, butter. The child's own name, Alice, is given as A-si, or perhaps A-śi (later á-si). As to sound, she is now acquiring the English long sound of i (ai). R is still impracticable, and attempts to form it sometimes give d (but this was very transient, and l soon became the common substitute): compare the converse Bengalese treatment of Sanskrit d, which, I believe, is in Bengal regularly pronounced as r. "P'ram," for perambulator, becomes thlam: the th, with an extra aspiration, almost ΧΘ. A few weeks later this was simplified into khlam. There seems to be a difficulty about initial vowels: "egg" becomes lleg (or perhaps yleg would be nearer), which I can only write symbolically: the sound marked as ll or yl is something like the Spanish ll with an aspiration. A few days later the initial sound was more sibilant and less vocal, say (symbolically) zhy.
Early in March (at twenty months) we noted the first attempt at sustained conversation. The child was looking, or pretending to look, for a lost object on the floor. We told her she would get her hands dirty. On this she exclaimed, in a tone of dissenting interrogation, "Dirty!" (da-ti), and then, after looking at her hands, holding them
