We used, therefore, the strip of intestinal muscle as an indicator. Later Hoskins[42] modified our procedure by taking, instead of the strip, a short segment of the rabbit intestine. The segment is not subjected to danger of injury during its preparation, and when fresh it is almost incredibly sensitive. It may be noticeably inhibited by adrenin, 1 part in 200 millions! The strip, or the intestinal segment, was suspended between minute wire pincers (_serres fines_) in a cylindrical chamber 8 millimeters in diameter and 5 centimeters deep. By a thread attached to the lower serre fine the preparation was drawn into the chamber, and was held firmly; by the upper one it was attached to the short end of a writing lever (see Fig. 2). When not exposed to blood, the strip was immersed in a normal solution of the blood salts (Ringer’s). The blood or the salt solution could be quickly withdrawn from or introduced into the chamber, without disturbing the muscle, by means of a fine pipette passed down along the inner surface. The chamber and its contents, the stock of Ringer’s solution, and the samples of “quiet” and “excited” blood were all surrounded by a large volume of water kept approximately at body temperature (37° C.). Through the blood or the salt solution in the chamber oxygen was passed in a slow but steady stream of bubbles. Under these circumstances the strip will live for hours, and will contract and relax in a beautifully regular rhythm, which may be recorded graphically by the writing lever. [Illustration: Figure 2.--Diagram of the arrangements for recording contractions of the intestinal muscle.] The first effect of surrounding the muscle with blood, whether “quiet” or “excited,” was to send it into a strong contraction which might persist, sometimes with slight oscillations, for a minute or two (see Figs. 4 and 5). After the initial shortening, the strip, if in quiet blood soon began to contract and relax rhythmically and with each relaxation to lengthen more, until a fairly even base line appeared in the written record. At this stage the addition of fresh “quiet” blood usually had no effect, even though the strip were washed once with Ringer’s solution before the second portion of the blood was added. For comparison of the effects of “quiet” and “excited” blood on the contracting strip, the two samples were each added to the muscle immediately after the Ringer’s solution had been removed, or they were applied to the muscle alternately and the differences in effect then noted. The results obtained by these methods are next to be presented. REFERENCES [Footnote 29: Jacobi: Archiv für experimentelle Pathologie und Pharmakologie, 1891, xxix, p. 185.] [Footnote 30: Biedl: Archiv für die gesammte Physiologie, 1897, lxvii, pp. 456, 481.] [Footnote 31: Dreyer: American Journal of Physiology, 1898-99, ii, p. 219.] [Footnote 32: Tscheboksaroff: Archiv für die gesammte Physiologie, 1910, cxxxvii, p. 103.]