RELATION OF HEREDITY TO CANCER 199
fa *x^ (Spotted
Shotted ®P^**^^ {spotted
Between these extreme iypes of ancestry every possible intergrade has been obtained again and again^ so that if we happened by chance to pick matings of the types (a) or (e?), or others approaching them, we shonld imdonbtedly prove the inheritance of spotting to our own satisfaction. On the other hand, if matings (&) and (c) had happened to have been our experience, we should believe that no such inheritance existed. (Gen- erally speaking, a mixed population of spotted animals would form for biometric methods of analysis only confused and inconclusive material on which no conclusion of lasting value could be based.
It has^ however, long been known that spotting in mice is inherited, and I have recently been able to account for and predict the occurrence of the spotted forms on the basis of the interaction of at least three pairs of hereditary factors showing Mendelian or alternative inheritance. In making the analysis of this problem it was fortunate that in certain races in the laboratory, only one or at most two of the three types of spotting existed together. This fact made possible the recognition of certain relationships between the different types of spotted coat which would otherwise have certainly escaped notice, and without which even an incomplete explanation of the facts would have been impossible.
The case of spotting in mice has been entered into at some length because of the fact that it proves the inadequacy of purely biometric methods to detect or explain a case of herediiy even involving as few as three pairs of factors. Moreover, these spotting factors produce a series of formB recognizable in early life, and spotting, unlike cancer, is free from the effects of age, sex or any but the most radical environmental disturbances. To disprove inheritance solely by biometric methods in this simple case is impossible, and the same is certainly true in the obvi- ously less simple case of cancer.
If now we turn to a consideration of the human beings as material and of certain facts concerning the biological nature of cancer, we can recognize the handicaps under which we must work, if we attempt to investigate the course of hereditary tendencies to cancer in man.
Biologically, cancer may be considered as consisting of a mass of tissue of local origin manifesting uncontrolled and unlimited growth. The problems of its etiology are therefore essentially those of the fac- tors causing, limiting and directing cell division.
If we for a moment consider the cells of the animal body as units, we can picture the embryo of any mammal, at the gastrula stage, as consisting of essentially two types of slightly differentiated cells, ectoderm and endoderm. Each of the two types of cells may figura-
��\
�� �
