The following conclusions may be drawn from an analysis of the protocols:
1. The chlorin output is very much diminished during the process of experimental pneumonia in dogs (tables I and II). The quantity of chlorin retained in the cases with well developed exudates (table I) is much larger than the exudate could account for.
2. Turpentine pleurisy in dogs (tables III and IV) also is associated with a retention of chlorin, but this is not so marked as in pneumonia. There is no direct quantitative relation between the water output and the chlorin excretion.
3. The rabbits with pneumococcal septicemia (tables V and VI) did not retain chlorin in any phase of the disease, a fact which Hutchison has also noticed.
It seems justifiable to conclude that the retention of chlorin in the organism in the course of pneumonia is conditioned by more than one factor. The formation of the exudate is one of these factors, but it alone is not capable of producing the same degree of chlorin retention as occurs in pneumonia. The additional factors are not yet determined; but it may be stated that the retention of chlorin is probably not due to the general effect of the pneumococcus on the organism or to its injury to the kidney. The principal cause of the retention seems, therefore, to be due to local changes associated with the pulmonary condition.