Maurer,[54] Gibson,[55] François Frank,[56] and Edgren[57] deal with this problem. The curve of intraventricular pressure cannot be studied in man for obvious reasons, and only in some cases has an attempt been made to compare the sphygmographic curve with the curve of intraventricular pressure obtained from animals. One of the most interesting attempts in this line will be mentioned later. To the second class belong all those investigations, by which experimental evidence in favor of one or the other hypothesis has been collected. The experiments which belong to this class are in so far more decisive as the conditions of the experiments are better known and, therefore, easier to interpret. Von Kries proved the existence of the dicrotic in the femoral artery of an animal after having replaced the heart by a bag filled with liquid.[58] Grashey[59] and Hoorweg[60] have demonstrated the existence of secondary waves in models, on which peripheral reflection was impossible. To the same type of experiments belong Marey's[61] and Grashey's registration of the waves in elastic tubes, and Mach's[62] tracings from a mechanical model on which the resulting movement of two simple components could be registered. Without giving any physiological theory Mach showed how curves similar to the pulse curves can be obtained by the registration of a movement, the mechanical conditions of which are known. As the results of these investigations, we may state the following facts as arguments against any hypothesis of the peripheral origin of the dicrotic elevation. (1) Automatic registration of the pulse wave shows that the dicrotic appears sooner in the regions nearer to the heart than in regions which are more distant. The opposite would be the case if the dicrotic elevation were due to a wave travelling from the periphery to the heart. (2) The dicrotic appears at the same time after the primary wave in a dwarf as in a tall man. This would be impossible if the wave had to travel so much farther. (3) Inhalation of amyl nitrite makes the dicrotic almost disappear. The adherents of the theory of the peripheral origin of the dicrotic wave explain this fact by supposing that this drug dilates the arterioles and makes little reflection possible. Their opponents say that the action of the heart and the resistance of the system are so enfeebled that the backward flow is slight and gives rise only to a small wave. (4) If an artery is opened and the blood allowed to spurt on a revolving drum of white paper a curve is obtained which shows the dicrotic elevation (the hemautographic curve of Landois). The resistance of the periphery is totally lacking in this case and the dicrotic elevation could not appear if it were due to a wave reflected at the periphery.