On account of the rapid spontaneous formation of fibrin and serum when the plasma is removed from the body and allowed to die, the exact chemical condition of the liquor sanguinis during life cannot be investigated, the separation occurring before the simplest chemical method can be carried out.

We have no reason to suppose that fibrin exists normally in the blood, but it would appear that this substance is only formed at the moment of coagulation, its appearance being one of the most obvious of many changes which take place at the time of the death of blood plasma.

The chemical changes comprehended under the term coagulation, occurring when plasma is deprived of its means of vitality, and ending in the production of fibrin and serum, are naturally of the first importance in studying the chemical relationships of living plasma. They can best be followed out in the coagulation of plasma when separated from the corpuscles, for (although the stages in the coagulation of blood are the same, the appearance of an insoluble albumin - fibrin - being the one essential in either case), the corpuscles complicate the process and modify the appearance of the clot.

Not only is the fibrin not present as such in the living plasma, but it requires for its production the presence of other substances which either do not exist in the living plasma, or are there so chemically associated as not to bring about the change which occurs when the plasma dies.

The reasons for believing this are the following: Fluids which sometimes collect by a slow process in the serous cavities of the body, e.g., hydrocele fluid, pleural effusion, etc., if kept quite clean do not generally undergo spontaneous coagulation. If to one of these some serum or recently washed blood clot be added, coagulation takes place just as in plasma (Buchanan). That is to say, we have here two fluids, neither of which coagulates when left to itself, but which do if mixed together. From each of these fluids a substance can be precipitated by passing a stream of carbon dioxide (C02) through the fluids. Both precipitates readily redissolve in weak saline solutions.

The solution prepared from the hydrocele fluid causes blood serum to coagulate; that prepared from the blood serum causes the hydrocele fluid to coagulate; and when mixed together the mixture of the two solutions coagulates; while the serum and hydrocele fluid from which the substances have been removed no longer have the power of exciting coagulation in each other or in like fluids. Here, then, are two materials; one, obtained in considerable quantity from serum after coagulation, is called paraglobulin (Schmidt) or serum globulin (Hammarsten); the other, occurring in serous fluids, is named fibrinogen. Both these substances are present in the dying plasma of the blood prior to coagulation. They can be obtained both together from the plasma if the plasma be treated with sodium chloride to saturation after either of the precautions already mentioned - viz., the application of cold, or the addition of neutral salt - has been taken to prevent the formation of fibrin. This precipitates a substance which readily dissolves if water be added to weaken the salt solution, and after some time the solution undergoes spontaneous coagulation, while the plasma from which it has been made has lost that power. This plasmin (Denis) no doubt is made up of different globulins, chiefly serum globulin, and fibrinogen, and contains in itself ail the necessary "factors" of fibrin formation, but is not at all identical with fibrin, since it readily dissolves in weak saline solutions, like the class of proteids called globulins, while fibrin is quite insoluble in such solutions.

In plasma removed from its normal relationships, both serum globulin and fibrinogen exist; but the former in far greater quantity than the latter, since the serum, after the blood clot is formed, contains no more fibrinogen, while the serum globulin makes up nearly half the remaining solids.

In preparing fibrinogen and serum globulin Schmidt found that the more carefully he operated, the weaker and more uncertain their action as fibrin factors became; and, finally, he made solutions which, when added together, did not produce coagulation, but which, when added to less pure solutions, gave good, firm clots. From this he suspected that a third agent which acted as a ferment was necessary to put into operation the fibrin-producing properties of the other two factors. He further succeeded in preparing the third agent, to which he gave the name of fibrin ferment. By treating serum with strong alcohol the proteids are precipitated; the ferment is carried down with them, and extracted with water. This extract, added to the mixture of the pure fibrin factors, which previously did not clot, caused rapid coagulation, but not when added to either of them singly.

This material is influenced by those circumstances which affect the activity of ferments in general: it has a minimum, o° C, optimum, 380 C, and maximum, 8o° C, temperature of activity, with various gradations of rapidity of action between each, and is destroyed if heated above 80 C. An active solution having the properties of the ferment can be extracted from whipped fibrin preserved in alcohol by an 8 per cent, solution of common salt (Gamgee).

Hammarsten thinks that the serum globulin is not indispensable to the formation of fibrin, because (1) a solution of fibrinogen may be made to coagulate without its presence; (2) the fibrinoplastic property of serum globulin is shared by casein and calcic chloride; (3) and is absent from pure serum globulin, (4) such as is present in hydrocele fluid which does not coagulate on the addition of fibrin ferment.

The source of fibrin is still a question of much difficulty, and will be further discussed with the question of blood coagulation within and without the vessels, after the morphological elements have been described.