1. The nature of a tumor can not be predicted from the metabolism because too much overlapping of metabolic rates exists between the pathological groups.

2. There is no evidence metabolically one way or another as to whether malignant lymphomata of any type should be classed as neoplastic or as infectious processes.

3. The degree of cell differentiation can in most cases be foretold by the percentage difference between the aerobic and the anaerobic glycolysis. The greater the differentiation the greater the percentage difference. Sarcomata in general constitute an exception to this rule.

4. The degree of malignancy in carcinoma, but not in other tumors, can, with certain exceptions, be predicted from the height of the value U.

5. Human sarcomata appear to have a metabolism far more closely comparable to that of benign tumors than to that of carcinomata. They do not behave as malignant tumors under the Warburg classification. Their energy requirements are not of the same order as those of carcinoma.

6. One can not from the value U or from the glycolytic rates predict whether or not a tissue should be classed as neoplastic.

7. Warburg's findings for carcinomata are confirmed and amplified.