had shown salt retention and died. Every one seems to be agreed that the 

 chlorin and sodium are generally distributed throughout the tissues. 

 Snapper has found that when sodium sulphate is added to normal blood 

 the chlorin passes from the blood cells into the fluid and that in pneumonic 

 blood it passes in the opposite direction. He believes that the chlorid re- 

 tention is caused by a modified permeability of the tissue cells. Medi- 

 greceanu and others have found that rabbits with pneumococcal septicemia 

 do not show chlorid retention, so there does 'not seem to be anything in 

 the toxin itself which causes this condition. 



Sandelowsky has found that patients who retained their weight during 

 the febrile period and lost weight suddenly after defervescence showed a 

 diminished concentration of proteins in the blood with a return to normal 

 after the loss of weight. This indicated an hydremia in pneumonia which 

 would account for a considerable retention of chlorids in order to maintain 

 the osmotic pressure of the blood. Rowe confirmed the decrease in total 

 protein, obtaining an average of 6.2 in pneumonia as opposed to his . 

 average of 7.94 for normal controls. The globulin percentage, on the other 

 hand, was increased to 40, which was almost double the normal. 



Roehrich and Wiki have found an extraordinary rise in the sodium 

 chlorid excretion in the fourth week of convalescence. 



Blood Constituents. With the high protein metabolism of lobar pneu- 

 monia it is not surprising to find some increase in the nitrogenous con- 

 stituents of the blood. A few of the cases reported, however, seem to have 

 an actual retention by the kidneys. Tileston and Comfort, who give as 

 their normal figures for non-protein nitrogen 22.9 to 25 mg. per 100 c.c., 

 found that out of fourteen cases of pneumonia six showed a moderate 

 degree of retention, that is, more than 35 mg. non-protein nitrogen in the 

 blood, the highest figure being 50 mg. The retention reached its maximum 

 toward the crisis and disappeared early in convalescence. It was not de- 

 pendent on the absorption of exudate, for it occurred before resolution 

 took place. Most of their patients showed albumin and casts in the 

 urine, two showed impaired renal function as estimated by the phenol- 

 sulphonephthalein test, five gave normal results. 



Schwartz and McGill analyzed the blood urea in twenty cases of 

 pneumonia and found an average of 40.5 mg. per 100 c.c., with a range 

 between 12 and 104. Only one patient with blood urea over 60 recovered. 

 They record that seventeen out of their twenty patients had toxic nephritis. 

 Their normal controls showed a range of 10 to 25 mg., with an average 

 of 12.9. Whipple and Van Slyke(c) found in lobar pneumonia the non- 

 protein nitrogen of the blood between 42 and 85 mg. per 100 c.c. 



Froth ingham(&), in four cases of Type 1 pneumonia, found the phthal- 

 ein test normal and blood urea slightly increased in two cases. In six cases 

 of Type 4 pneumonia two had phthalein tests just below normal and two 

 had slight increase in blood urea. He concludes that during febrile attacks 



