rather on the nature of the fever and on the history and general condition than on local signs.
Splenic enlargement may be present even when there is no malarial complication. This is rare, however, and in uncomplicated cases is seldom great. I have seen splenic tumour closely simulated by abscess in the left lobe of the liver.
Varicosity of the epigastric and hœmorrhoidal veins— one or both— is sometimes discoverable.
Œdema of the feet and ascites are rare in the earlier stages; but the former is very usual towards the termination of long-standing cases.
Local œdema over one or more intercostal spaces, or more extensive and involving the whole or part of the hepatic area, is sometimes apparent. When limited it is a useful locating symptom.
Local bulging, if attended with fluctuation, indicates the presence of pus near the surface and the pointing of the abscess. Usually this, when it occurs, is in the epigastrium; but pus may burrow and find its way down the flank, or among the muscles of the abdominal wall, and open perhaps at a point remote from the abscess cavity in the liver.
Friction, both pleuritic and peritoneal, is sometimes to be made out, and is not without its value as a localizing symptom.
Pneumonia, generally limited to the base of the right lung, and of a subacute and persistent character, indicates contiguity of the abscess to the diaphragm. It is especially common in those cases in which the abscess subsequently ruptures through the lung. This form of chronic pneumonia is a fruitful source of error in diagnosis.
Chronological relation of the hepatitis to the dysenteric attack.— This is most irregular and uncertain. In many cases of dysentery a concurrent hepatitis is manifest almost from the commencement of the attack; this hepatitis may not subside, but pass directly to abscess formation. Or the initial hepatitis and dysentery may both subside apparently, but the former may recur weeks, months, or even years afterwards, when, perhaps, the attack of dysen-
