Here again it is self-evident that with regard to the course of response, we must first consider the character of the single stimulus of the series, and this must be done from all those standpoints already here discussed. However, a new factor is met with here, that is, the frequency of the single stimuli of the series, or that which has the same meaning, the duration of the intervals between them. This is a feature upon which the result of stimulation depends in a very high degree. But here, too, however, it is not a case of the absolute frequency of the single stimulus, but simply of the relative frequency in regard to the rapidity of reaction of the particular living system. I should like to remark here that it is of greatest importance whether the interval between the two single stimuli of the series is sufficiently long or not to allow the living system time to completely recover from the effect of the _preceding_ stimulus. In the cases, for instance, where we have recovery, we have the same rhythm of stimulation as that of response. When recovery _does not_ occur, interferences of the response are developed, which are of great physiological importance, with the analysis of which we shall later on find occasion to occupy ourselves in detail. The physiological example for these stimuli is the rhythmical discharge of impulses of the nerve centers; the physical method, which is most widely used for experiments, is the faradic current. It is apparent that the question of frequency must again be combined with all those factors previously discussed in connection with the _single_ stimulus. In consequence another complication arises and with this another point must be taken into consideration, namely, the fact that the duration of the single stimulus in a series undergoes alteration by increasing frequency beyond a certain limit. Beyond this limit the duration of the single stimulus must become less and less. As the result of the fact that stimulation is, as we have seen, dependent on the duration of stimulus, it is evident that, depending upon the rapidity of response of the living system, sooner or later the rhythmical stimulation must become ineffectual. Nevertheless, this effect of shortening the duration of the single stimulus can be compensated by a corresponding increase of its intensity. In this connection _Nernst_[40] showed a very simple relation for induction currents of higher frequency of interruption, which furnishes a law according to which such a compensation takes place. In conjunction with _Barratt_ he found, namely, that the intensity must increase proportionately to the square root of the number of single stimuli if the threshold value of the stimulus is to be maintained, that is, I : √m = const., in which _I_ is the intensity of the current and _m_ the frequency of interruptions. The limits of the validity of this law cannot at present be conclusively established.