In conclusion we may summarize the above biological observations in a series of theses as follows:
1. Cancer cells differ from other epithelial cells in respect to: (a) Size relations of nucleus and cell body; (b) power of indefinitely continued division.
2. Cancer cells differ from embryonic cells in absence of: (a) power of differentiation; (b) power of coördination of parts to whole; (c) power of self-regulation and limit of growth.
3. The continued development of the cancer cells is subject to the following factors: (a) the inherent potential of division of the cancer cells. (b) The natural resistance of the inoculated animals. The latter factor is usually regarded as the index of malignancy of a tumor and is based upon the percentage of takes together with the period required to kill the mice. Our experiments, however, show that the percentage of takes is independent of the time factor, and indicate the presence of a third factor which may be described as (c) the potential of "infectivity" of the cancer cells.
4. The potential of infectivity of cancer cells is characterized by more or less regular rhythms; these must be distinguished from rhythms of growth energy of the cancer cells which in all probability occur within the individual mouse. Without the division energy of the cancer cell this infectivity is inoperative, hence it follows that the cause of the infectivity lies within the cancer cell or is constantly associated with it.
5. Cancer cells differ from epithelial cells by virtue of this potential of infectivity combined with that of division energy. There is reason to believe that the latter is due to the action of stimuli and not to the liberation of a restrained growth power of embryonic tissue. There is reason to doubt that an initial and discontinued stimulus is responsible for these attributes of the cancer cells. Certain benign tumors, or vegetable galls, may be due to the action of such initial stimuli, but in them there is no infectivity. Embryonic tumors, due to embryonic cells, have a high power of differentiation combined with their division energy, but there is no infectivity. Infectivity distinguishes all cancerous growths from normal epithelium and from benign tumors or teratomata.
6. The rhythms of infectivity of cancer cells, erroneously regarded as rhythms of growth energy by Bashford, Murray and Boyen, appearing as they do in successive batches of mice which we may legitimately assume to have like powers of resistance, must have their cause in the cancer cells themselves. These cells, therefore, must be equivalent to parasites, or else parasites are contained within or associated with them.
7. Upon any other hypothesis it is difficult to conceive of cells creating a continual stimulus to their own growth energy, and it is still more difficult to explain the rhythms of infectivity.
8. Many lines of evidence point to the presence of some possible organism within the cancer cell; some organism which, acting as does Plasmodiophora brassicæ within vegetable cells, underlies the infectivity of cancer cells and provides the stimulus for their continued proliferation. Upon such an assumption the numerous cases of cage infection find their explanation.
9. The various inclusions of the cancer cell which have been described as organisms have been disproved; yet the analogy of club root and the many filter experiments show that the cause of infection may lie within the cancer cell. It is conceivable that, like the yellow fever organism, such an incitant may be in the protoplasm and beyond our powers with the microscope to locate.
10. The spirochætes which we have found in mouse cancer may have something to do with this infectivity of cancer cells. They may be useful in preparing the "soil" in new mouse hosts and making it susceptible to cell growth; or they may have intracellular stages in their life history which are too minute to be seen. The rhythms of infectivity, finally, may be an expression of the vitality of these spirochætes or of the hypothetical ultra-microscopical organisms accompanying cancer cells.