344 Proceedings of the Royal Society of Edinburgh. [Sess. 
upper part of the tube. The apparatus of Riva Rocci, Stanton, Cook, 
Erlanger, and Jane way obtain readings by circular compression of one of 
the limbs, so as to obliterate one of the larger arteries and estimate the 
effects by examination of itself or one of its branches further from the 
heart ; the instruments of Marey, Hurthle, and Mosso act by embracing the 
extremities after the manner of the plethysmograph. The tonometer of 
Gartner gauges the pressure, after the removal of the compression of a 
finger, by the return of colour to the skin. It may be said at once that, 
with the exception of those sphygmomanometers which employ the method 
of circular compression of one of the larger limbs, there is none which need 
be seriously discussed. All the others are only of historic interest. The 
larger instrument of Hill and Barnard registers its results by means of an 
aneroid, and the later instrument of Oliver (20) by the use of a tube of 
spirit. Both of these also employ the method of circular compression. 
The methods of Riva Rocci, Martin, Cook, Stanton, and Janeway yield 
an accurate determination of the systolic pressure, and allow of an 
approximate estimate of diastolic pressure. The larger instrument of Hill 
and Barnard and the earlier apparatus of Oliver, in which the arterial 
pressure is gauged by the maximum excursions recorded, were intended 
to register the mean pressure within the vessel ; the theory being that 
when the pressure surrounding the vessel and that within it are equal, 
oscillations attain their maximum. It has, however, been proved by 
Howell and Brush (27) that the maximum oscillations give diastolic read- 
ings. The sphygmomanometer of Erlanger is based upon the same 
principles as that of Riva Rocci, and gives systolic and diastolic readings ; 
it is the earliest attempt to furnish graphic records of arterial pressure in 
clinical investigation. A full description of the instrument is given by 
Jane way (4). It yields tracings of the oscillations of the column of air 
between the compressing band and the column of mercury in the mano- 
meter, but it does not furnish any record of the height of the column 
of mercury by which the pressure may be estimated. It is, therefore, 
necessary to watch the manometer and note the oscillations. The return 
of the pulse at the wrist is found to coincide with the commencement 
of large oscillations, and there can be no difficulty, therefore, in the 
estimation of the systolic pressure ; while, according to Erlanger, the point 
at which the large oscillations suddenly begin to diminish is an indication 
of the diastolic pressure. My own observations with this instrument, 
which have been very numerous, lead me to concur with Erlanger in 
these views. 
The sphygmomanometer recently introduced by me (28) also takes ad- 
