A second point of great interest is the absolute constancy and fixation of its terminal portion at the point of junction with the jejunum, more correctly termed second ascending or fourth portion. Mr. Treves says that this fourth portion is never less than an inch, and is practically constant. It extends along the side of the left crus of the diaphragm opposite the second lumbar vertebra, and is there firmly fixed to the front of the aorta and crus of the diaphragm by a strong fibro-muscular band, slinging it up and absolutely retaining it in position. This band has been termed the "musculus suspensorius duodeni," but is chiefly composed of white fibrous tissue, and is more of the native of a ligament than a muscle. This ligament is always present, and its position is never altered. The curve of the duodenum may descend as far as the iliac fossa, but the terminal portion is always maintained by this band in its normal position. Another point of great constancy is the position of the pancreas and its relation to the curve of the duodenum. The duodenum always curves round the head of the pancreas and is, as it were, moulded on it and retained in position by it. In birds the duodenum always forms a long loop embracing the pancreas. Further, the ducts of the liver and pancreas always open into the center Of the duodenum, either separately or by a common opening. [Illustration] Now, the absolute constancy of the curve of the duodenum, the complete fixation of its fourth portion, the position of the pancreas, and the place of entry of the ducts of the pancreas and liver, are all component parts of a siphon trap, whereby gases generated below the duodenum are prevented from passing upward. A reference to the accompanying diagrams will make this quite clear. A is a diagram of a siphon trap copied from Parkes' hygiene. B is a very diagrammatic outline of the stomach and duodenum, _a_ is intended to mark the position of the fibrous band, or musculus suspensorius duodeni; and _b_ the position of entry of the ducts of the liver and pancreas. The duodenum, then, is a siphon trap, and a most efficient one. Now, the efficiency of a siphon trap depends not only on its shape, but what is absolutely essential is that the curve must be kept constantly full of fluid, without which it ceases to be a trap, and would allow gases to ascend freely. The position of the place of entry of the ducts of the pancreas and liver assures that this _sine qua non_ shall be present. The discharge of the secretions of the pancreas and liver, although more active during and after feeding, is practically constant, and so insures in an admirable manner that the curve on which the efficiency of the trap depends shall be constantly kept full not only with fluid, but, as I would suggest, antiseptic fluid. There is no other trap in the intestinal canal, but the peculiar position of the colon would no doubt have more or less effect in preventing gases ascending through