Omitting complications, when first passed, the urine is always clear. After standing a few hours there falls to the bottom of the vessel a muddy sediment (cloudy in its upper layers) of red or grayish red color. Spectroscopically, this urine shows the lines of oxyhemoglobin, as Corre first demonstrated, though often, too, especially after standing some time, those of methemoglobin, and frequently of urobilin (Louvet). According to Grocco, in mild attacks urobilin alone is present.*

Berthier claims that the hemoglobin cannot be removed from the urine by dialysis, and assumes that it is not exactly in a dissolved condition, but is connected with a very fine red blood corpuscular debris.

The reaction of the urine is feebly acid, neutral, or even slightly alkaline. The specific gravity varies considerably. According to Berenger-Feraud, it fluctuates between 1014 and 10.35. It appears to depend, therefore, more on the concentration of the urine as a whole than on the amount of coloring matter, which, in spite of its impressive color, considered quantitatively, is not very great. At the close of the paroxysm a large amount of clear urine of low specific gravity (108 or even under) is excreted.

During the attack the daily amount of urine is likewise very variable (it is, as a rule, increased), though it may be usually decreased, and often to a few cubic centimeters. Complete anuria for one or several days is not uncommon. Berenger-Feraud describes a very rare case that passed, in one day during the attack, 4.5 liters. A similar observation was made by Corre.

The urine always contains albumin, and often in larger amounts than would correspond to the quantity of hemoglobin. Denozeilles

* For the sake of the historic interest, we may mention that the first describers of this disease attributed the color of the urine sometimes to biliary coloring matter (Berenger-Feraud, Daulle), as well as to the blood (Pellarin, Barthelemy-Benoit). This explains the different names applied to the condition at various times. The demonstration of blood coloring matter spectroscopically was made by Corre, Venturini, and Karamitsas. That it was dissolved blood coloring matter, or, in other words, a hemoglobinuria and not a hematuria, was shown by Corre, Grennet, Louvet, and Karamitsas. (Compare Corre's work, "Traite des fievres bilieuses," etc.) determined the amount in several cases, and found it between 0.3 to 1.6 per cent.

Ordinarily the albuminuria continues a few days after the attack, when it gradually decreases. In some cases its continuance indicates the development of nephritis.

In addition to serum albumin, peptone and propeptone have been found. The phosphates are decreased (Grocco). Bilirubin is generally absent. As long as the urine is very dark, Gmelin's test is unsatisfactory. When the amount of hemoglobin diminishes, the test is occasionally positive. Calmette claims he never succeeded in finding it.

Microscopically, none at all or only a few red blood corpuscles are observed. The sediment consists of isolated leukocytes, hyaline and sometimes granular casts, to which often amorphous yellowish granules are attached. These granules are also found free in large numbers. [They are probably the remains of the destroyed red cells.-Ed.] Epithelium from the kidney and the urinary passages may be present in larger or smaller amounts.

Icterus is an essential symptom of blackwater fever. It is sometimes present in a slight degree during the prodromal fever paroxysm. Ordinarily it appears first with the hemoglobinuria, becomes especially evident during the hot stage, and later increases rapidly in intensity. It often continues a few days after the attack. The intensity of the icterus varies between a slight suggestion of yellow on the sclera to the most intense discoloration of the skin.

As to the course of blackwater fever, we may mention, first, that the paroxysm of hemoglobinuria itself is usually ushered in by prodromes. These consist in paroxysms of fever, which are repeated two or three days and which are often characterized by being introduced with a more or less pronounced chill. These prodromes, therefore, are differentiated from those of the ordinary tropical malarial fevers in that, as mentioned before, the latter often lack the cold stage. .

Still we must add that here, too, the prodromes may occur without the chill. In this case there are malaise, pain in the limbs, and slight fever. It is rare to see the blackwater attack set in without any prodromes (Berenger-Feraud).

We have first one to three prodromal fever paroxysms, frequently very slight, and, apart from the chill, presenting no special symptoms, then all at once the peculiar attack and the remainder of the symptoms occur.

According to the severity of the symptoms, the majority of writers distinguish mild and severe cases. Berenger-Feraud describes four grades of severity-namely, first, mild or intermittent; second, moderately severe or remittent; third, severe or subcontinued; fourth, especially severe (fulminating) forms. This outline shows, too, the connection between the type of fever and the severity of the condition, though in this regard there are exceptions.

In the mild cases the paroxysm begins with a chill of varying severity. The patient complains of headache, constriction in the region of the liver and the stomach, and very often of lumbar pain. There is nausea, with an inclination to vomit. Vomiting soon occurs and becomes frequent, consisting first of the remains of food, later of pure green bile; yet it may be entirely wanting. The patients suffer from violent thirst, but on account of the constant nausea, dare not alleviate it.

The epigastrium and right hypochondrium are painful. Sometimes the whole abdomen is spontaneously painful and sensitive to pressure. The liver is somewhat enlarged. The spleen usually shows a large tumor (the result of earlier infection) and is always enlarged to percussion.

Constipation is common, though there is sometimes an inclination to diarrhea.