of red blood-globules, which are usually well preserved, but in old exudations may be more or less decolorized.

Aside from injuries, hemorrhagic exudations are most common in connection with tuberculosis and new growths, so that their microscopic examination is of much diagnostic and prognostic value.

For the examination of tubercle bacilli see page 91. In this form of exudation it is rarely possible to demonstrate them. On the other hand, it is not infrequently possible to make the diagnosis of a malignant growth, especially of cancer, from the examination of the sediment.

No cell is significant of cancer or other neoplasm, but the occurrence of numerous cells which vary greatly in form is suspicious. The cells from new growths are often unusually large, up to 120μ, frequently contain one or more vacuoles, and usually lie in clumps. Large drops of fat are also considered suspicious.

A positive diagnosis can only be made by obtaining bits of tissue which show the structure of the new growth, such as the atypical alveolar arrangement of the cells in cancer.

Purulent exudations appear more or less thick and yellow, and deposit a corresponding layer of pus. Microscopically, they present no peculiarities other than the organisms to which they are due. Among the etiological factors actinomyces must always be thought of in puzzling cases.

Putrid exudations occur in the pleural and peritoneal cavities in consequence of gangrenous masses breaking into them and from stomach or intestinal ulcerations, from new growths, occasionally from no clear cause. The fluid resulting from the perforation of a gastric ulcer may show yeastcells and sarcinæ, and give an acid reaction.

Lumbar Puncture.-The diagnostic value of lumbar puncture has been sufficiently demonstrated. Not only is it possible to diagnosticate inflammation of the meninges, but the character and cause of the inflammation may usually be demonstrated if the examination of the fluid is properly performed. In a number of cases of general infection in which there was no inflammation of the meninges a diag

nosis has been made by means of cultures taken from the cerebro-spinal fluid. Finally, a number of cases of hemorrhage into the brain and spinal canal have been diagnosticated by lumbar puncture.

The operation and the subsequent examination of the fluid should be as carefully performed as any other bacteriological investigation in order to obtain accurate results. The back of the patient and the operator's hands should be made sterile. The needle should be boiled for ten minutes. The patient should lie on the right side, with the knees drawn up, and with the uppermost shoulder so depressed as to present the spinal column to the operator. This position permits the operator to thrust the needle directly forward rather than from the side. An antitoxin needle 4 cm. in length, with a diameter of 1 mm., is well adapted for infants and young children. A longer needle is necessary for adults and children over ten years of age.

Aspiration of the fluid is not necessary, but some operators prefer to attach a hypodermic syringe to the needle to afford a better grasp for the hand. In this case the syringe would have to be detached to allow the fluid to flow. The additional manipulation, and possibly the defective sterilization of the syringe, might impair the subsequent bacteriological examination.

The puncture is generally made between the third and the fourth lumbar vertebræ; sometimes between the second and third. The thumb of the left hand is pressed between the spinous processes, and the point of the needle is entered about 1 cm. to the right of the median line and on a level with the thumb-nail, and directed slightly upward and inward toward the median line. Care must be exercised to prevent the point of the needle from passing to the left of the median line and striking on the bone. At a depth of 3 or 4 cm. in children and 7 or 8 cm. in adults the needle enters the subarachnoid space, and the fluid flows usually by drops. If the point of the needle meets with a bony obstruction, it is advisable to withdraw the needle somewhat, and to' thrust again, directing the point of the needle toward the

median line, rather than to make lateral movements, with the danger of breaking the needle or causing a hemorrhage. The smallest quantity of blood obscures the macroscopic appearance of the fluid by rendering it cloudy. The fluid is allowed to drop into an absolutely clean test-tube which previously has been sterilized by dry heat to 150° C. and stoppered with cotton. The fluid should be allowed to drop into the tube without running down the sides. From 5 to 15 c.c. of fluid is a sufficient quantity for examination.

Cultures on blood-serum should be made at once from the fluid. The test-tube is then held toward the light and gently shaken; by this means the slightest cloudiness can be detected.

In meningitis there is always an exudation of cells which makes the fluid more or less cloudy. The degree of cloudiness is to some extent proportionate to the amount and character of the exudation. In tubercular meningitis the amount of cellular exudation is sometimes so slight that the fluid appears clear unless examined carefully. A guinea-pig should be inoculated with some of the fluid as a control-experiment. The inoculation should be made a short time after withdrawing the fluid before the formation of fibrin occurs, because the web-like coagulum of fibrin holds the tubercle bacilli in its meshes. After standing for several hours more or less fibrin forms in the fluid, and contains the cells in its meshes. The supernatant fluid is left clear. Cover-glass preparations should be made from this sediment, dried by passing through the flame of an alcohol lamp or Bunsen burner, and stained with methylene-blue, which stains the nuclei of the cells and any pathogenic bacteria which may be present.

In tubercular meningitis the predominant cell is the lymphoid cell. In purulent meningitis the polynuclear leucocyte predominates. Cover-glass preparations may also be stained for tubercle bacilli by appropriate methods of staining.

It is often necessary to examine twenty or more preparations, however, before finding the bacilli. The albumin should be quantitated. Normally, the cerebro-spinal fluid contains from to of 1 per cent. or less. In menin

gitis the amount is increased from to of 1 per cent., or often more. Percentages of albumin down to T of I per cent. can be estimated by the ferrocyanide-of-potassium-and acetic-acid test and a centrifugal machine. Take 3 c.c. of a 20 per cent. solution of ferrocyanide of potassium, 1 c.c. of acetic acid, and 10 c.c. of the fluid in which the albumin is to be quantitated. The mixture of reagents and fluid is poured into a conically-shaped graduated glass vessel designed for the purpose and centrifugalized. Sugar is rarely present and has no diagnostic value.

Ovarian and Parovarian Cysts.-The simple cysts of the ovary due to distention of Graafian follicles or to cystic change of corpora lutea, and the parovarian cysts contain a thin, clear, serous fluid of low specific gravity.

The contents of the multilocular and papillary adeno-cystomata of the ovary are usually tenacious and mucous, of very varying specific gravity, from 1005-1050, but usually between 1020 and 1024. The fluid generally contains much albumin and is rich in metalbumin, which is precipitated by alcohol, but not by acetic acid, nitric acid, or boiling, so that it can readily be distinguished from mucin. Before making the test the albumin must be removed.

The cyst-contents are usually yellowish, but sometimes. may be dark-red or chocolate-colored. Microscopically, red and white blood-globules, occasionally blood-pigment and cholesterin crystals, often fat-granules and large vacuolated cells, are found in the cyst fluid. Bizzozero considers cylindrical epithelial cells, ciliated and beaker cells, and colloid concretions especially important from a diagnostic point

of view.

Pancreatic Cyst or Fistula.-The fluid obtained from a permanent fistula or large cyst of the pancreas contains much less solids than the normal pancreatic juice, and the trypsin ferment may be present in very small amount or possibly be entirely wanting. The fluid is colorless, alkaline in reaction, and has a specific gravity of about 1011. It is characterized by three distinct properties on which its recognition depends-namely:

1. It splits up fat into fatty acids and glycerin. Mix together equal parts of neutral olive oil and the alkaline fluid. Test with litmus-paper. Place the mixture in the incubator at 37° C., and test from time to time. If the fluid is pancreatic, an acid reaction will be obtained in twelve to eighteen minutes.

2. It transforms starch into sugar. Place in the incubator equal parts of a 1 per cent. aqueous solution of starch and of the fluid to be tested. In ten to twenty minutes test for sugar with Fehling's solution.

3. It digests fibrin in an alkaline solution (trypsin ferment). Place some fibrin in the alkaline fluid and set it in the incubator. In one-half to one hour examine for peptones by the biuret test. Add caustic potash or soda and a few drops of a dilute solution of sulphate of copper. If peptones are present, a beautiful reddish-violet color will be produced.

Dropsy of the Gall-bladder.-Puncture is generally not advisable. The fluid is usually colorless and mucoid or serous in character. All trace of biliary constituents may have disappeared. According to Lenhartz, numerous colon bacilli are usually present.

Hydronephrosis and Renal Cysts.-The fluid is almost always clear as water, rarely reddish or yellow. Specific gravity always under 1020 (usually between 1010 and 1015). Urea and uric acid are generally present, but may be absent. (Small amounts of urea are sometimes present in ovarian cysts.) Albumin is slight in amount. Microscopically, almost nothing is found.

Echinococcus Cysts.-The fluid is perfectly clear, free from albumin, and contains a little succinic acid and much chlorid of sodium. The specific gravity varies between 1008 and 1013.

Microscopically, often no traces of morphological elements can be found. Occasionally, however, hemosiderin granules or chlolesterin crystals occur, or the characteristic structures from which a positive diagnosis can be made— namely, scolices, hooklets, or pieces of cyst-membrane.