The short-lasting 'burst' activity was recorded from decerebrate, unanesthetized cats and dogs, by means of a single, bipolar needle electrode, originally aimed at recording of 'slow spikes' of internal anal sphincters. The recording of the 'burst' was more accessible in cats than in dogs, although the same macro-electrode, of which diameter was 0.3 mm, has been used throughout. There was no such species difference of accessibility for recording of 'slow spikes' of internal sphincters. Histological examinations indicated that the difference of accessibility may be due to relative thinness of muscular tissue in the cat. Some of the bursts were seen simultaneously on the background of continuous 'slow spikes', however, the most of the bursts appeared in reciprocal manner with those 'slow spikes', namely, the bursts were elicited by rectal distension, while the 'slow spikes' of internal sphincters were wholly inhibited by the same distension. The participation of bursts to the anal canal pressure was little in fact, though the longitudinal length of anal canal was invariably shortened, whenever the repetitive bursts occurred either spontaneously or in response. It is concluded that the 'burst' activity is generated by longitudinal muscles of anal sphincters. The observation suggests further that cell-to-cell electrical communication does not operate between longitudinal and circular muscles in this portion, instead, that neural mechanism regulates dominantly for co-ordination of those muscles encircling the anal canal.