The following table, somewhat abbreviated, is taken from HoppeSeyler; the numbers are parts per 1,000.

The most striking fact illustrated by this table is the greater amount of solids in chyle as compared with lymph, and the large percentage of fat. Zawilski found in dogs fed purely on a fatty diet that the chyle might contain as much as 14.6 per cent. of fat. During the active digestion of fat, the blood plasma and serum have a milky appearance produced by the presence in them of excessively minute fat globules. 1,000 parts of the dry residue of the ethereal extract of chyle contained:7

The chyle also contains a certain proportion of soaps absorbed from the alimentary canal. This statement was originally made by HoppeSeyler, and he still maintains the correctness of his earlier observations, which have been questioned by Lebedeff, Röhrig, and Zawilski. In some new experiments he has found in the serum of the horse, ox, and dog, a percentage of fatty acids from soaps varying from 0.05 to 0.12.

1 Physiol. Chemie, p. 595.

5 Analyses by Hoppe-Seyler.

2 Ludwig's Arbeiten, xi. 147.

4 Analysis by Schmidt.

5 Analyses by Rees, Phil. Trans. 1842, p. 81. The chyle was obtained from a decapitated criminal.

6 Noel Paton. This analysis is not contained in Hoppe-Seyler's table. The chyle was obtained from a patient whose thoracic duct had been ruptured by an operation for tumour in the neck (Journ. of Physiol. xi. 109).

7 Hoppe-Seyler, Physiol. Chem. p. 597.

8 Hoppe-Seyler, Zeitsch. f physiol. Chem. viii. 503.

In the chyle1 he has found 0.225 per cent. of soaps, and 0.723 of fat.

per cent.

The increased percentage of proteids in the chyle as compared with the lymph illustrates the fact that the lacteals are not merely concerned in the absorption of fatty, but probably also of albuminous food. In the stomach and intestine the proteids of the food are converted into peptones, substances that diffuse with readiness through living animal membranes ; but no peptones 2 (or proteoses, the intermediate products in the formation of peptones) are found in the chyle; during their passage through the intestinal wall, or immediately on entering the lymph or blood stream, they are reconverted into albumin and globulin. Schmidt-Mulheim 3 tied the thoracic duct in dogs, and found that proteids were still absorbed; this, however, does not prove that the lacteals are not normally concerned in the absorption of proteid; it merely shows that animals thus treated can continue to absorb proteid by the other path-the blood vessels.

The intestinal lymphatics are thus concerned in the absorption of fat and of proteids; they, however, apparently take but little part in the absorption of carbohydrate food; the amount of sugar in lymph and chyle is approximately the same as in the blood, and no definite increase occurs when animals are fed on a starchy or saccharine diet (Bernard, v. Mering). Probably, as sugar is so easily diffusible, most of it passes into the more quickly circulating blood stream and is carried off, fresh quantities of blood being then available to carry off more. The blood vessels, moreover, lie immediately beneath the epithelium, and so the sugar never reaches the more centrally situated lacteals of the villi (Heidenhain 5). By greatly increasing the amount of sugar in the food, however, some does pass into the chyle."

Quantity of Chyle.-From two cases in which the amount of chyle was measured C. Schmidt concluded that for every kilogram of body weight 0.61 kilo of chyle was formed in the 24 hours, of which 0-34 comes from the alimentary canal, and the remaining 0.27 consists of the normal lymph. Hoppe-Seyler is inclined to think that the proportion derived from the intestine is much smaller. Exudations of Chyle.-Owing to the rupture of the lacteals or of the thoracic

1 Obtained from a case of chylous ascites, i.e. escape of chyle into the peritoneal cavity.

? This is a statement made by numerous observers; I have confirmed its accuracy in an examination of chyle collected from the thoracic duct of two dogs.

5 Du Bois Reymond's Archiv, 1877, p. 549.

4 Ludwig's Arbeiten, 1877. Arch. f. Anat. u. Physiol., Physiol. Abth. 1877, p. 379.

5 Heidenhain, Pflüger's Arch. supplemental volume, 1888, p. 71.

duct, or owing to fistulous communications between these parts and other cavities, chyle may pass into the serous cavities, giving rise to chylous dropsy. For Chyluria see Urine.

THE LYMPH IN SEROUS CAVITIES DURING HEALTH

The amount of fluid in these cavities is in health very small; excess finds its way through the stomata into the lymphatic vessels. The fluid is undoubtedly lymph-that is, dilute blood plasma which has exuded from the blood vessels. In dropsical conditions this fluid is much increased, and our knowledge of its properties is almost entirely derived from a study of dropsical fluids. If excess of fluid accumulates in two of these cavities simultaneously, as, for instance, in the peritoneum and the pleura, from alterations in conditions of vascular pressure (e.g. heart disease), it is found that the composition of the two fluids differs to a certain extent. On this ground we hold that the normal lymph which moistens the various serous cavities probably differs in those different cavities in the same way. The differences are quantitative only, not qualitative. After death the pericardium, especially in some animals (e.g. the horse), often contains a considerable quantity of liquid. This accumulation is accounted for by the changes in the circulation immediately preceding death. The liquor pericardii is a liquid which is interesting historically, as so many experiments have been made with it in dealing with the investigation of the causes of the coagulation of the blood (see p. 243). The cerebro-spinal cavity is not a serous cavity, and the fluid in it differs markedly from the lymph of the serous cavities. It will be dealt with separately.

DROPSICAL FLUIDS

The modes of causation of dropsy have been already considered (p. 332). These fluids may all be tersely described as lymph in excess and more watery than usual, except in inflammatory dropsy, where the cells and the solid constituents generally are increased.

Nomenclature. There are certain names given to the various forms of dropsy occurring in different situations:

Edema is the name given to the excessive exudation of fluid into the subcutaneous tissues.

Ascites is the name given to a dropsy of the peritoneal cavity. Hydrocele is the name given to a dropsy of the tunica vaginalis, the serous membrane originally part of the peritoneum that surrounds the testicle.

Hydrothorax is the name given to a dropsy of the pleura.
Hydropericardium to that of the pericardium.

The names of diseases that are inflammatory in nature terminate in the affix itis. Thus there is pericarditis, peritonitis, pleuritis (or pleurisy), &c., and in certain stages of all these diseases there is effusion of fluid.

Reaction. This is in all cases alkaline.

Colour. This varies directly with the colour of the blood plasma of the patient, and with the concentration of the effused liquid, but there is always a certain amount of the yellowish-green lipochrome (serum-lutein, see p. 253), which can be extracted by means of alcohol.

Specific gravity. This increases pari passu with the amount of solid constituents. Reuss has examined a large number of these effusions; he calls the fluids of inflammatory dropsy, exudations, while the dropsical fluids (i.e. diluted lymph) he terms transudations, and the following are the general conclusions he draws with regard to specific gravity.

Coagulation. Non-inflammatory dropsical fluids do not coagulate spontaneously, or only with exceeding slowness. When mixed with serum, or contaminated with blood, as they are apt to be in the process of tapping, they, however, do coagulate, forming fibrin.

The cause of

The reason why they do not clot is that they contain either very few cellular elements or these may be practically absent. the coagulation of the blood, we have already seen, is the formation of the fibrin-ferment from white corpuscles and blood tablets. When, therefore, blood or serum or a solution of pure fibrin-ferment or cellglobulin, or of some other active globulin like myosinogen (see Muscle), is added to one of these dropsical fluids, the fibrinogen contained therein is converted into fibrin (see Coagulation of the Blood, p. 241). The inflammatory fluids or exudations are, however, different; they contain abundance of white corpuscles and invariably clot when shed.

1 A Reuss, Deutsches Arch. f. klin. Med. xxviii. 317. Hoffmann has made similar observations, Virchow's Archiv, lxxiii. 250.

? Such as is obtained by blistering.

Sometimes they clot within the serous cavity, the fibrin sticking to the sides of the membrane.

Constituents. These are the same in kind as those in blood plasma. a. Proteids. These are fibrinogen, serum-globulin, and serumalbumin.2

b. Extractives. This term is used in the sense explained on p. 251. In some cases cholesterin is found in marked excess. Sugar seems to be a fairly constant constituent.

c. Salts. These are alike not only in kind but in actual amount to those in the blood.

The different dropsical fluids differ from one another in their richness in organic constituents, especially in proteids; the pleural fluid is richest in these substances, then the peritoneal, and lastly the fluid of subcutaneous œdema. These facts may be illustrated by the following tables :

Composition of various dropsical fluids removed after death from a case of
albuminuria (C. Schmidt)3:—

Composition of various dropsical fluids removed simultaneously from a

case of albuminuria (Hoppe-Seyler) 1 :—

These examples illustrate sufficiently well the fact that it is the amount of proteids that varies in these different fluids, the other constituents being fairly constant.

5

Runeberg examined 77 cases of effusions of different kinds; the amount of total proteids varied from 0.06 to 2.68 per cent. ; while the

1 One often hears these strands of fibrin called lymph in the post-mortem room. 2 The serum-albumin of these fluids like that of serum can by fractional heatcoagulation be differentiated into three proteids (see p. 247).

5 Quoted by Hoppe-Seyler, Physiol. Chem. p. 602.

4 Physiol. Chemie, p. 602. Also in Arch. f. path. Anat. ix. 257.

5 Runeberg, Deutsch. Archiv f. klin. Med. xxxv. 266.