Page 104, 
72 ON SPHYGMOGRAPHY. 
terminated before the loss of blood from the proximal aorta into the 
heart is made good by any retardation of the onward current, or a 
reverse stream in the more distant vessels originates; consequently 
this restoration of equilibrium has to be in great measure effected after 
the valves have closed, encroaching on the diastole to form the second 
cardio-arterial interval of the cardio-sphygmograph trace, which does 
not cease until all the complications attending the closure of the semi- 
lunar valves have come to an end, whereupon the again augmenting 
centrifugal current, or in very low-tensioned pulses, such as that 
figured in Fig. 1, the diminishing centripetal one, originates the 
secondary rise in the sphygmoscope trace. 
In both the primary and secondary rises of the pulse-beat (Fig. 1), 
it is found on inspection that the summits of the pulsations (8) are 
delayed upon, or are not reached so soon as the centrifugal current 
maxima (@); and that such should be the case is essential, in a cireu- 
lation maintained by an intermittent motor organ, which, like the 
when precisely stated, the problem is a. manifest contradiction. In p. 22 the diaspa- 
sis is discussed, and in the summary of the results arrived at, in p. 27, the following 
are given as the relations of the lengths of the various elements of the pulse-beat to 
one another. 
1. The systole together with the diaspasis, or in other words, the first cardiac 
interval varies as the square root of the whole revolution. 
2. The systole varies as the square root of the diastole. 
3. The diaspasis is constant. 
The incompatibility of these three statements it is not difficult to prove, for if 
the first and second are true, the third cannot be so, and, as a fact, the diaspasis 
is not constant. There is every reason to believe that for a given pulse-rate the 
diaspasis does not vary ; and, on subtracting its length, as obtained by measurement 
in quick cardiograph tracings, from the first part of slow ones, I found that the 
remainder, the systolic interval, varied very nearly as the square root of the diastolic. 
On repeating the operation on the pulse of intermediate rapidity a similar result was 
obtained, and the error being extremely small, I attributed it to my not having 
extracted the necessary square roots to a sufficient number of decimal places, and 
thus felt justified in making the generalization given above. Since the mistake has 
been pointed out to me I have repeated the arithmetical computations more care- 
fully, and find that what I had first supposed were errors on my part, are constant 
variations, which prove that, the other statement standing as above, the diaspasis is 
slightly longer in slow pulses, occupying approximately 0°002 of a minute in a pulse 
of 61, and 0°0018 of a minute in a pulse of 152 in a minute. This fact, therefore, 
leads to the conclusion that the rapidity of the fall of blood-tension has an influence 
on the length of the diaspasis, lengthening it slightly when the tension-fall is 
retarded, probably because the previous systole is then more powerful and gradual. 
It is to be noted that the second cardio-arterial interval of cardio-sphygmograph 
tracings is almost of the same length as the diaspasis, and varies in the same or in a 
very similar manner, which may be the cause of the somewhat undecided nature of 
the notch in the sphygmosystole of slow pulses. From these remarks it is necessary 
to substitute for statement 3, as given above, the following: 3. The diaspasis varies, 
being slightly longer in slow pulses. 
