nomenon.

In the organic world, the combination of these elements for the construction of the alkali is a common pheFrom the experiments of Milne-Edwards, it is known that a slight loss of nitrogen occurs in respiration. What is the destination of this? There is ammonia in the air. Is this a product from the animal world, or a gift to it? In the metamorphosis of the nitrogenous tissues in the living body, is an ammonia a normal and necessary product? These are mere suggestions, and as such are left with the reader.

The reason why the evolution of an ammonia from blood has not been previously shown is easily explained. It had not been looked for in blood-halitus. Notwithstanding Scudamore and Polli's laboured researches, it had not been absolutely proved that coagulation depends on the elimination of volatile matter at all. Lastly, the proportion of volatile alkali necessary for holding blood fluid, is so exceedingly minute, as compared with the mass of blood, that its presence might easily escape observation. Ammonia, indeed, could not be expected to be found in quantity in the serum or clot of healthy blood, its elimination occurring in the act of coagulation. According to the time at which the analysis of blood commenced after abstraction, and according to other accidental conditions, such as the healthy or modified condition of the blood operated on, a little of the alkali might or might not be fixed and retained. Hence some chemists have found, in their inquiries, evidence of chloride of ammonium; others have not. The explanation of the two facts is easily understood.

[In some extreme conditions of disease, however, where the alkali entered into combinations which fixed it for the chemist's hand, or where the proportion of

ammonia was raised, as in Dr. Blair's examples of the blood in yellow fever (to which reference will be made in the Appendix), the evidence of its presence was obvious enough.

To ascertain with precision the absolute quantity of ammonia required to hold fibrin in solution in the circulating blood, is a point surrounded with grave difficulties, not only on account of the smallness of the quantity itself, but from the fact that the physical conditions of the healthy circulation cannot be imitated to perfection out of the animal body, and that the proportion is obviously not the same in every kind of animal, nor yet in the same animal at all times. These facts are of immense moment. In sheep's blood, coagulation occurs in a period of time at least two-thirds less than in that of the ox, while ammonia must be added in larger quantities to sustain the fluidity of the blood of the ox than of the sheep. The inference from these facts is, that in ox's blood the amount of the volatile solvent is greater in proportion to the fibrin; while in the sheep, the amount is smaller in proportion to the fibrin. When, again, an animal is exhausted by fatigue, the coagulation of blood is slow, and the solvent evolved is better secured, as I have shown by experiment. In like manner, in exhausted states, the evolution of ammonia by the breath is more free and decided. These results are possibly due to a greater development of ammonia, as a result of metamorphosis of the tissues.

At first sight, the amount of ammonia required to hold blood fluid seems incredibly small when considered in relation to the mass of blood. But it is wrong to consider this relationship at all; for the point to be settled is, what proportion must the alkali bear to the

fibrin in any given quantity of blood? If, then, we take 2.2 grains of fibrin as the proportion normal to a thousand grains of blood, we have to discover how much of alkali would be required to hold this small quantity of fibrin in solution, in a serum already slightly alkaline, but not sufficiently so for such solution. Moreover, we have to consider this problem in connexion with the admirable physical arrangements of the circulation, by which it is insured, not only that the mass of fibrin thrown into the general circulatory channels is added particle by particle, but that, in the rapid circulation, the fibrin is held in solution for the brief space of time only, during which it remains in the blood before being given up to the tissues.

In health, there is unquestionably a definite relationship between the amounts of fibrin and of its solvent; which relationship is susceptible of but little variation without the development of general disorder. If the proportion of alkaline solvent is beneath the natural standard, the fibrin is imperfectly held in solution, and its deposition is imminent; while, if the proportion of alkali exceed the normal standard, not only is the fibrin held more decidedly fluid, but the bloodcorpuscles also become subject to dissolution-an event which seems inevitable, if the quantity of alkali exceed one part in a thousand of blood.

Returning to the absolute question, what proportion of ammonia is required in the normal circulation to hold the fibrin in solution, there are two modes of investigation-the one directed to the detection of the exact amount of ammonia evolved from a given quantity of recently drawn blood-the other directed to the determination, as far as possible, of the smallest amount

of the alkali required to sustain the fluidity of drawn blood under conditions approaching, as nearly as possible, to the conditions of the circulation itself. Both these investigations are, however, attended with so many difficulties, that the results to which they lead can be considered as approximative only. The experiments given at pp. 274-6, while valuable in their way, must be accepted simply as qualitative; since a great loss of volatile matter occurred in transferring the blood from the animal operated on to the containing vessel. To render these more complete, I made several attempts to receive the blood at once in large quantities from the vessels of animals, and to wash over the gases evolved into an acid solution, driving the atmospheric air first through sulphuric acid, to remove all fallacies in reference to the traces of ammonia in the atmosphere. But, owing to the delicacy of the fact to be determined, and to the modifying influences detailed above, the results have been too varied to admit of close argument, though sufficiently valuable to lead to a longer series of similar observations.

By the synthetical process, some kind of approximation to the answer of the problem of quantity may be cautiously approached; for, although it does not show the minimum of alkali required for the solution of the fibrin in the normal circulation, it proves that a certain quantity there present is sufficient for all required purposes; and that an excess, over and above a certain proportion, would be detrimental. Thus Experiment CCCXXVII shows that one-fourth of a grain of ammonia to five hundred grains of ox's blood, or, in other words, one grain of the alkali to two thousand of blood, are sufficient, even after exposure to the air, to hold the blood long fluid. Experiment cccxxx shows that one

grain of the alkali, in three thousand grains of sheep's blood, was sufficient to retain fluidity for twelve minutes; this blood being at rest and exposed to the air at 60° Fahr.

But the proportion of alkali is better shewn in Experiment CCCXLII, where a fourth of a grain of the alkali in vapour, diffused through 87.5 cubic inches of air, was sufficient, thus diluted, to retain the fluidity of five hundred grains of blood, so long as it gently passed through it. In this experiment, better than in any preceding, the relative positions of the ammonia and blood in the system were imitated. The blood, from the first fluid, was steadily supplied with the diluted portion of ammonia gas, which constantly passed in current with the air through the blood, at the rate of 0.12 cubic inch of the combined gas per second. With such thorough current of air through the blood, it is doubtful if any accumulation of ammonia would occur; and we must, therefore, leave the cause of the sustained fluidity to the equal and constant diffusion of the alkaline gas.

Another experiment further proves this point.

EXPERIMENT CCCXCIX. Blood Retained Fluid by Ammonia and Motion. I received two thousand grains of blood from an ox, added to it one fourth of a grain of ammonia in a drachm of water, and instantly transferred the whole to an apparatus similar to the one described at page 224. When the blood was fairly in its closed circuit, it was kept in steady circular motion for an hour. It was then withdrawn, entirely fluid; but, on exposure to the air, quickly underwent coagulation. There was no separation of fibrin within the bags.

To sum up, and as an approximation towards the fact, I opine, from all my experiments, that a propor