Every such chronological view of the sequence of sounds is uncertain, because we can not observe the child uninterruptedly, and hence the first appearance of a new sound easily escapes notice. The above synopsis has a chronological value only so far as this, that it announces, concerning every single sound, that such sound was heard in its purity by me at least as early as the given month. The sound may, however, have been uttered considerably earlier without my hearing it. I know from personal experience that in other children many sounds appear much earlier; in my child, e. g., _ngae_ was observed too late, and I have no doubt that the first utterance of _f_ and _w_ was unobserved, although I was on the lookout for them. When it is maintained, on the contrary, that _m_ is not heard from a normal child until the tenth month, then the _am_ and _moe_ which appear universally in the first half-year have escaped notice. Earlier tabular views of this sort, which have even served as a foundation for instruction of deaf-mutes in speaking, do not rest exclusively on observation. Besides, in this matter, even two children hardly agree. According to my observations, I am compelled in spite of this disagreement to lay down the proposition as valid for all healthy children, that the greatly _preponderating majority of the sounds the child makes use of after learning verbal language, and many other sounds besides these, are correctly formed by him within the first eight months_, not intentionally, but just as much at random as any other utterance of sound not to be used later in speech, not appearing in any civilized language. I will only mention as an example the labio-lingual explosive sound, in which the tip of the tongue comes between the lips and, with an expiration, bursting from its confinement is drawn back swiftly (with or without tone). All children seem to like to form this sound, a sound between _p_, _b_, and _t_, _d_; but it exists in few languages.