AND THE FORCES WHICH VARY IT. 43 
pulse-quickening effects of a hot bath, whether air or water. Nume- 
rous experiments by the author prove that the effect of copious blood- 
letting is not to modify the pulse-rate at all, thus showing that the 
law given by Marey respecting pulse frequency is not correctly based. 
The above points, namely, the law of Poiseuille, the dependence of the 
capacity of the arteries and ventricles on the pressure of the blood, 
the dependence of the pulse-rate on the peripheral resistance and its 
non-dependence on the blood pressure, can ull be correlated by only one 
theory, namely, that the heart always re-commences to beat when the 
tension or pressure in the arteries has fallen at invariable proportions, 
which also assumes that the capacity of the heart and arteries varies 
directly as the pressure. The facts that the arteries are generally 
empty after death, and that the cavity of the heart is sometimes found 
to be obliterated on rigor mortis, show that absence of pressure and 
capacity go together. 
This theory explains the known peculiarities in pulse-rate attending 
change in position, by showing that while standing all the pressure of 
the body weight is borne by non-compressible rigid tissues, and so the 
circulation is normal, but while lying, the soft parts are compressed 
and resistance introduced into the circulation, reducing the rapidity 
of tension-fall, and therefore the frequency of the pulse; an interme- 
diate condition attends the sitting posture. The pulse quickens during 
inspiration, and becomes slower during expiration; for during the 
former act the reducing pressure in the chest lowers the aortic blood 
pressure, and makes the tension-fall more rapid, while in expiration 
the reverse occurs. 
This theory also is the only one which throws light on the cardio- 
graph law published by the author (see “Journal of Anatomy and 
Physiology,” 1870-71), which may be thus stated:—For any given 
pulse-rate the first part of the heart’s revolution has a constant length, 
but it varies as the square root of the length of the complete pulsa- 
tion. The pulse-rate not depending on the blood pressure, and the 
length of the first cardiac interval not varying when the rate is con- 
stant, its length also does not depend on the blood pressure. The first 
cardiac interval may be divided into the systolé and the interval 
* between that and the closure of the aortic valve (the diaspasis) ; 
these combined not varying as the blood pressure, it is almost certain 
that separately they do uot do so either ; so it may be said that neither 
the length of the systolé nor of the diaspasis depends on the blood 
pressure. But the fall of tension between the pulse-beats being but 
small, and the diaspasis length not depending on the blood pressure, 
there is no reason why it should vary in length with different pulse- 
rates; and assuming this in connection with the measured diaspasis 
length in a particular case (‘00183 of a minute), it can be deducted 
