The arterioles become contracted by the action of the involuntary muscular fibre contained in their walls; they dilate partly by their own elasticity and partly by the pressure of fluid within them.

The capillaries also appear to have the power of contraction. Both arterioles and capillaries are induced to contract by the effect upon them of the nerves which pass to them from vasomotor centres. The blood-vessels may also dilate actively from irritation of vaso-inhibitory nerves. The exact mode of action of these nerves is not ascertained; they are generally looked upon as entirely separate from vaso-motor, but it seems not improbable that here also the difference between vaso-motor and vaso-inhibitory nerves is a mere question of relation, and some nerves produce contraction and dilatation according to the point where they are stimulated. Thus Dastre and Morat have found that the cervical sympathetic, which produces contraction of the vessels in the rabbit's ear when irritated between the ear and the first thoracic ganglion, causes dilatation instead of constriction when it is irritated at a point below the ganglion, in which case the stimulus has to pass through the ganglion before it reaches the ear.

In considering the action of drugs on the vessels, we have, therefore, to examine 1. Their direct effect upon a. The contractile walls of the vessels themselves with their a, muscular fibres, b, motor ganglia;

1 John Hunter's works, edited by Palmer, 1837, vol. iii. p. 91.

b. Nerve-fibres a, vaso-motor, b, vasodilating; c. Nerve-centres a, vaso-motor, b, vaso-dilating.

2. Their reflex effect on the nerve-centres just mentioned.

There are two modes of estimating the contraction of the arterioles : 1st, by direct observation and measurement under the microscope; 2nd, by ascertaining the quantity of blood or other fluid which will pass through them in a given time.

Each of these methods may be used in several ways, according as we wish to ascertain the action of a drug - 1st, on the contractile walls of the vessels alone; 2nd, on the walls together with the vascular nerves but without the nerve-centres; and 3rd, on the vessels in connection with the nerve-centres.

The method of direct observation of the arterioles may be practised in either frogs or mammals.

The part of the frog usually selected is the web, the mesentery, the mylo-hyoid muscle, the tongue, or the lung. The parts usually observed in mammals are the wing of the bat and the ear of the rabbit.1

In observing the effect of various conditions on the lung, it is necessary to inflate it. This is easily done by means of a small cannula with a bulging end which is tied into the larynx. Over the other end is slipped a small piece of india-rubber tubing, and by clamping this after the lung has been inflated, the escape of air is prevented.

An apparatus for this purpose is described by Holmgren.2 The accompanying engraving (Fig. 92) shows one which I used in 1870 for the purpose of investigating the action of heat and cold upon the lung.3

By means of the india-rubber ball I directed upon the lung; a stream of air which was previously passed either through hot water or through iced water. The pulmonary capillaries, when treated in this way, contract under the influence of cold by one-third of their diameter. McKendrick, Coats, and Newman, in an investigation on the action of anaesthetics on the pulmonary circulation, found that chloroform, ethidene, and ether, all stop' the pulmonary circulation, the action of chloroform being greatest and that of ether least.4

In observing the effects of drugs on the vessels alone, it is necessary to destroy the influence of the nerve-centres over them1 For observing the vessels of the rabbit's ear one of Brucke's lenses is very convenient. It resembles a telescope in its construction, but has a very short focus2 Ludwig's Festgabe.

3 British Medical Journal, Feb. 13, 1875, p. 204.

4 Ibid. Dec. 18, 1880.

This is usually done in a frog by destroying the brain and spinal cord. In the rabbit's ear it is done by dividing as far as possible all the nerves going to one ear, then injecting the drug into the general circulation and comparing its effect upon the two ears.

Fig. 92.   Apparatus for ascertaining the effect of heat and cold on the vessels of the frog's lungs. A, a piece of cork to which the frog is fastened, is laid on B, the stage of a microscope, and attached by an india rubber strap, c.

Fig. 92. - Apparatus for ascertaining the effect of heat and cold on the vessels of the frog's lungs. A, a piece of cork to which the frog is fastened, is laid on B, the stage of a microscope, and attached by an india-rubber strap, c. D is a small ring of cork covered with a thin circle of glass. X is the inflated frog's lung. f is a tube by which a current of air can be directed on the frog's lung. It is held in position by a piece of wire, G, which can be bent to any position. I is a flask containing ice and water. h, a flask containing hot water. K is a three-way stopcock, by which a current of air may be sent from the spray-producer, l and m, through either I or H at will, and thus cold or hot air may be applied alternately to the lung.

It is evident, however, that such experiments are not free from fallacy, because in them the circulation is dependent on the condition of the heart as well as that of the vessels; and both of these may be affected by the drug.

A better plan, therefore, is to obviate this fallacy by keeping up the circulation artificially, either in the body of the frog, or in the ear of the rabbit.

A method of maintaining artificial circulation in the rabbit's ear while the calibre of the vessels is being measured was invented by Ludwig, and described by me in the British Medical Journal, 1871.

In the frog artificial circulation is kept up by putting a cannula into the aorta, and another into the vena cava or abdominal vein after destruction of the brain and spinal cord. The aortic cannula is connected with two funnels or bottles, such as are used for artificial circulation through the intestine (p. 382). These contain either a saline solution or a mixture of saline solution with defibrinated blood. To one of them the drug is added. The circulation can be rendered quicker or slower at will, by increasing the pressure under which the fluid flows into the aorta. A suitable part of the frog is then put under the microscope, and the vessels measured while unpoisoned blood flows through them. The poisoned blood is then allowed to circulate under exactly the same conditions of pressure and the vessels are measured again. By this method of observation Gaskell ascertained that very dilute alkalies cause great contraction of the vessels, so as sometimes almost entirely to occlude them and arrest any flow of blood through them. Dilute acids counteract this effect and cause the vessels again to dilate.

Cash and I have observed that, in addition to this action, dilute acids have a tendency to increase the exudation of fluid from the vessels and produce oedema of surrounding tissues.

In many experiments which have been made on the action of drugs on the blood-vessels by direct microscopic measurement of their size, before and after the application of the drug, no account has been taken of the effect which the application of the drug may produce by its local irritating action on the nerves or tissues of the part to which it is applied, and by its reflex action through the nerves, quite independently of any special action which it may have on the vessels. Thus, irritation by the application of alcohol, either alone or as a solvent in tinctures, or by a strong saline solution, has an effect similar to that of simple irritation by pressure or scratching, and usually causes temporary contraction, followed by dilatation of the capillaries. This contraction may be more or less prolonged, according to the strength of the irritant which is applied. Unless these conditions are taken into account, observations on the effect of drugs applied locally to the web, mesentery, or tongue, are very unsatisfactory and generally worthless.

Perhaps a somewhat better result may be obtained by injecting the drug into the lymph-sac of a frog, and then observing the web. But here also we have the same difficulty, because the sensory nerves of the lymph-sac being irritated, reflex stimulation of the vaso-rootor centre and consequent contraction of the vessels may be induced.