_Window--red_ is an individual reaction; _red--flowers_ is found in the frequency tables; therefore, _red_ is classed as an association to preceding reaction. In cases in which neither the reaction in question nor the preceding reaction happens to be one of our stimulus words, but a relationship between them may be judged to exist without considerable doubt, the reaction in question is also classed here. Example: _priest--father_ _ocean--mother_ _Ocean--mother_ is an individual reaction; neither the word _father_ nor the word _mother_ is among our stimulus words; but the association between the words _father_ and _mother_ may be judged to exist without considerable doubt; therefore, in this case _mother_ is classed as an association to preceding reaction. In such cases as this personal equation must necessarily come into play; comparative uniformity of judgment may, however, be attained by systematically excluding any reaction the relationship of which to the preceding reaction is subject to any considerable doubt and by placing any such reaction in the unclassified group. *Repetition of Previous Stimulus.*--Here we place any reaction which is a repetition of any previous stimulus from amongst the ten next preceding, at the same time placing *repetition of preceding stimulus* under a separate heading. *Neologisms.*--Here we place the newly coined words, so commonly given by the insane, excepting such as possess a sound relationship to the stimulus word, for which, as already stated, a special place in the classification has been provided. Neologisms might be divided into three types, as follows: (1) those which arise from ignorance of language (_comfort--uncomfort, short--diminiature_); (2) distortions of actual words, apparently of pathological origin and not due to ignorance (_hungry--foodation, thief--dissteal_); and (3) those which seem to be without any meaning whatever (_scack, gehimper, hanrow, dicut_). It is, however, impossible to draw clear-cut distinctions between these types, and for this reason we have made no provision in our classification for such division. *Unclassified Reactions.*--This group is important, in the first place, because it is numerically a large one, and in the second place, because it contains certain fairly definite types of reactions which are placed here for the sole reason that we have not been able to find strictly objective criteria for their differentiation from other types. It has already been stated that the frequency tables, even together with the appendix, fail to exhaust all normal possibilities of association, so that a certain small number of perfectly normal reactions must fall into the unclassified group. We submit the following examples: _music--listen_ _smooth--suave_ _sour--curdled_ _earth--mound_