[Illustration: Fig. 30.--Figures are to be read from left to right. The top line records the points where sounds were heard, the figures above the short vertical lines refer to tones (see text). Mx. B. P., maximum blood-pressure. M. B. P., minimum blood-pressure. P. B., pressure bulb recorder. It was impossible to lower and raise this bulb by hand without obtaining the great irregular oscillations of the attached lever above the mercury manometer. B. L., base line.] Erlanger showed some years ago, that with his instrument, the point at which diastolic pressure should be read was at the instant when the maximum oscillation of the lever suddenly became smaller. While checking up the graphic with the auscultatory method using Erlanger's instrument, it was noticed that the disappearance of all sound did not correspond with the sudden diminution of the oscillation of the lever connected with the brachial artery. A series of records were carefully made on patients. It was seen that during the period of the third tone phase the oscillations of the lever on the drum reached a maximum (Fig. 31) and remained at approximately the same height for some millimeters while the pressure was gradually falling. At a point at which the third tone, clear and distinct, became dull, there was an appreciable decrease in the height of the pulse wave. From this point to the disappearance of all sound there was a gradual diminution of the size of the pulse waves. [Illustration: Fig. 31.--Fast drum. Sudden decrease in size of pulse wave at 4, marking the change from clear sharp tone to dull tone.] [Illustration: Fig. 32.--Slow drum. Sudden decrease in amplitude at 4.] For normal pressures the difference between the fourth (dull) tone and the fifth (disappearance of all tone) phase, amounted to 4 to 10 mm. Occasionally the difference was so little, the change from sharp third tone through fourth dull tone to disappearance of all sound was so abrupt, that one could take the disappearance of all sound as the diastolic pressure, with an error of not more than 2 to 4 mm. This is within the limits of normal error and practically may be used by those who have difficulty in noting the change from third to fourth phase. For high pressures, however, the difference between fourth and fifth phases was never less than 8 mm., and was found as much as 16 mm. The diastolic, therefore, should always be taken at the fourth phase if possible. It was found that with the dial instrument the greatest fling of the lever corresponded to the third phase and the sudden lessened amplitude of the oscillation was at the fourth phase and was coincident with the change of tone from sharp to dull. Thus the diastolic pressure may be read off on the dial scale by watching the fling of the hand and with some practice one might acquire considerable accuracy. It is better, simpler, and, for most observers, more accurate to use the stethoscope and hear the change of sound.