perceptible pulse wave felt beneath the palpating finger at the wrist, represents on the scale the systolic pressure. This can be seen to correspond to a sudden increase in the magnitude of the oscillation of the mercury column. The systolic pressure, thus obtained, is from 5-10 mm. lower than the real systolic pressure. The more sensitive the palpating finger, the more nearly does the systolic pressure reading approach that found by using such an instrument as Erlanger's, where the first pulse wave is magnified by the lever of the tambour. [Illustration: Fig. 23.--Method of taking blood pressure with a patient in sitting position.] [Illustration: Fig. 24.--Method of taking blood pressure with patient lying down.] The pressure is now allowed to fall, until the palpating finger feels the largest possible pulse wave, which is coincident with the greatest oscillation of the mercury. This is the diastolic pressure. Beyond this point there is no oscillation of the mercury column. The difference between the two is the pulse pressure. Thus the pulse is felt after compression at 120 on the scale, and the maximum oscillation occurs at 80. The systolic pressure is 120 mm., the diastolic is 80 mm., and the pulse pressure is 40 mm. With the "Tycos" or Faught the arm band is snugly wound around the arm, the bag next to the skin and the end tucked in, so that the whole band will not loosen when air is forced into the bag. The cuff is blown up until the pulse is no longer felt. One should raise the pressure not more than 10 mm. above the point of obliteration of the pulse. The valve is then carefully opened so that the needle gradually turns toward zero. At the first return of the pulse wave felt at the wrist, the needle is sure to give a sudden jump. This is the systolic pressure and is read off on the scale. The needle is now carefully watched until it shows the maximum oscillation. This is the diastolic pressure. The difference between the two is, as above, the pulse pressure. In taking pressure one should take the average of several, three or four. Moreover, one must not take consecutive readings too quickly and one must be sure that between every two readings all the air is out of the cuff and that the mercury or dial is at zero. _It has been repeatedly shown that in a cyanosed arm the systolic pressure is raised so that even slight cyanosis between readings must be carefully avoided._