The production of the sounds by the spasmodic expiratory movements shows that the passage from the trachea into the pharynx, _viz._, the glottis or chink between the vocal cords, is partially closed. The quality of the sounds is {31} explained by the particular arrangements, at the moment of the cachinnation, of the vocal apparatus, and more particularly the shape of the resonance chamber of the mouth. Familiar though we are with them, we should find it hard to give an accurate description of the sounds of laughter. To begin with, they seem to vary considerably in the case of the same person and still more in that of different persons. Laughter has not yet lent itself to the methods of the experimental psychologist, and so has not been studied with scientific precision. By-and-by, we may hope, the phonograph will capture its sounds, and enable us to observe them at our leisure. Meanwhile, only a very rough account of them is possible. Taking the laughter of the adult male, which is perhaps more frank and better pronounced, we find the more common forms of iterated sounds to range from the broad vowel sound _aw_ (in “law”) to the sharp _a_ (in “bat”). The long _o_ sound (as in “go”), involving the rounded mouth aperture, seems to me to be far less common. The same applies to the long _ee_ and _ai_ sounds, and those which seem to be most closely allied to them. These variations appear, so far as I can judge, to go with alterations of pitch. The broader sounds, _e.g._, _aw_, seem naturally to ally themselves to the hardier deep-pitched explosion, the others to the more cackle-like utterances in the higher parts of the register. This connection shows itself, too, in the change in the vowel-quality when, as frequently happens, the laugh runs through a cadence of pitch from a higher to a lower note. These considerations will prepare us to find that the vowel-quality of the sound varies in general with sex and with age. According to Haller and Gratiolet the sounds of {32} the laughter of women and children, which correspond with their higher vocal pitch, approach in vowel-quality to the French _i_ and _e_.[16] Considerable variations from these typical forms would seem to occur now and again. In the American returns already referred to, the mode of laughing described is represented by such odd symbols as “gah! gah!” “iff! iff!” “tse! tse!” etc. These singularities, if, as it seems, they are intended to represent habitual modes of voicing mirth, are, one suspects, hardly referrible to natural differences of vocalisation, but are probably the result of the interfering agencies of nervousness and affectation which, as we know, have much to do with fixing the form of mirthful expression.