The purpose of this paper is to ascertain whether the impulses producing bigeminal pulse from benzol insufflation, etc., traverse the spinal cord, to determine the route through the cord and to the heart. To exclude every known pathway from the brain to the heart except the spinal cord route, the vagi and depressors were sectioned and the inferior cervical sympathetic ganglia removed in 6 rabbits. One period of benzol insufflation was sufficient to produce a bigeminal pulse in 4 rabbits, while 2 required an additional period. In 4 rabbits the spinal cord was transected in the lower cervical region; bigeminal pulse from insufflations was blocked. Of 3 rabbits in which the spinal cord was cut at the level of the 3rd thoracic vertebra, 1 gave a bigeminal pulse from insufflation; 2 did not, nor did 2 others in which the cord was cut at the 4th thoracic. In 3 rabbits in which the cord was cut at the levels of the 6th, 7th and 8th thoracic vertebrae respectively, bigeminal arhythmia could be elicited by insufflation but was usually delayed. Both stellate (1st thoracic sympathetic) ganglia were completely removed in 9 rabbits; in no instance was bigeminal arhythmia obtained by insufflation or other means. In 2 rabbits in which the left stellate was extirpated bigeminal pulse followed the 4th and 10th benzol insufflations respectively; in 2 rabbits in which the left stellate was removed and the right sympathetic trunk cut immediately below the stellate, 7 and 10 benzol insufflations, respectively, were required to elicit bigeminal arhythmia. In 3 out of 5 rabbits in which the right stellate had been removed (1 also had the left sympathetic cord cut below the stellate) a bigeminal pulse was readily produced from insufflations; the other 2 gave no arhythmia. Sectioning the last cervical and the 1st and 2nd thoracic spinal roots in 2 rabbits and the 1st 3 thoracic roots in 1 rabbit prevented bigeminal pulse from benzol insufflation. Transection of both lateral columns in some rabbits, the dorsal and ventral columns and the median 1/2 of the lateral column in others, and the lateral and dorsal columns in still others blocked bigeminal arhythmia from insufflation, transection at the 6th cervical of the dorsal and ventral columns in some rabbits and the right lateral column and the outer 1/2 of the left lateral column in 1 rabbit did not. It is concluded that such bigeminal arhythmia is caused by impulses which descend the median 1/2 of the lateral columns of the spinal cord and pass to the left ventricle by way of the upper thoracic roots and the stellate ganglia. This is obviously an action not usually attributed to the sympathetics.