'Actively acquired tolerance' introduces into immunology the concept of a specific inhibition of response. Tolerance of a tissue homograft comes about when an animal has been confronted in foetal life with cells taken from its future donor, or from some other member of the donor's inbred strain (section 3 \cdot 1). It depends (a) upon an embryo's inability to respond to antigens by becoming immunized, and (b) upon its continued inability to do so in later life. Methods for inducing tolerance in mice (section section 3 \cdot 2, 4 \cdot 1, 9), rats and rabbits (section 3 \cdot 4), and birds (section section 3 \cdot 3, 5, 7) are described in full. In normal development, response to an antigenic stimulus by becoming tolerant gives way to response by becoming sensitized or immune. The transition from the one modality of response to the other occupies a 'null period' during which the exposure of animals to an antigenic stimulus has no appreciable effect. Most but not all mice and birds at birth or hatching have already entered this transitional period (section section 4 \cdot 1, 5). Tolerance is antigenically specific. An animal injected in foetal life with cells from a donor A becomes completely tolerant of homografts transplanted in later life from a donor B if, and only if, B contains no antigens that are not also present in A. (In practice, this condition is most easily fulfilled when A and B are members of the same highly inbred strain.) The reaction of a tolerant animal against a homograft from an unrelated donor is not perceptibly impaired. Tolerance does not, however, discriminate between the tissues of a single individual; the inoculation of foetal or newborn mice with leucocytes or with the cells of a mammary carcinoma may confer tolerance of later homografts of skin (section section 4 \cdot 2, 9). Tolerance of a homograft is neither caused by nor accompanied by an antigenic adaptation of the grafted cells. An animal that is tolerant of a homograft in one part of its body is tolerant in another; tolerance is systemic, and a tolerated graft does not build up a privileged position within its own lymphatic territory (section 4 \cdot 3). Every degree of tolerance is possible, from that which allows a homograft to live only a few days beyond its normal median expectation of survival to that in which it is permanently accepted by and incorporated into its host. An inhibition of response which is partial may nevertheless be permanent, for the weakening of the 'secondary response' in partially tolerant animals is proportional to the weakening of the first (section section 3 \cdot 2, 4 \cdot 4). The stimulus which confers tolerance must be fully antigenic, i.e. must be one which would have caused an older animal to have become immune. Cells such as erythrocytes which have no power to elicit transplantation immunity are incapable of causing tolerance of tissue homografts; all treatments which abolish the power of cells to confer tolerance upon embryos will also abolish their power to make older animals immune (section 5). Immunological reactivity can be promptly and permanently restored to a tolerant animal by inoculating it with cells taken from the regional lymph nodes of actively immunized members of its inbred strain. It may also be restored, more slowly, by the inoculation of normal unimmunized lymph node cells. A tolerant mouse thus retains in full the power to give effect to an immunity of adoptive ('passive') origin; a tolerated homograft continues to be a source of antigenic stimuli, and its susceptibility to a reaction directed against it remains unchanged. Tolerance represents a central failure of the mechanism of the immunological reaction, and is not caused by an intercession at a peripheral level (section 6). The relationship between twinning, fertility, tolerance and red-cell chimerism is analyzed. Like dizygotic twin cattle, twin chicks that arise from double-yolked eggs are synchorial, are red-cell chimeras, and are tolerant of grafts of each other's skin. Tolerance and infertility are not causally connected. The tolerance produced in chicks by artificial synchorial parabiosis from the 10th day of embryonic life until hatching is accompanied by a true persistent red-cell chimerism. The disappearance of chimerism in partially tolerant chickens does not reveal a return to normal reactivity, for their secondary response to red cells reintroduced by cross-injection is profoundly impaired (section 7). Some measure of tolerance of skin heterografts may be achieved by the synchorial union of embryonic ducks and chicks (section 8). Tolerance may be produced by, and in respect of, tumour homografts, and by tumour homografts in respect of skin. A degree of immunity which does not suffice to hold in check the growth of a tumour may destroy a normal homograft completely; the growth of a tumour homograft is therefore a less exacting measure of tolerance than the survival time of a homograft of skin (section 9). A naturally acquired tolerance of maternal homografts is believed to occur, very rarely, in guineapigs, presumably by the accidental incorporation into a foetus of maternal cells. No such natural tolerance has been observed in mice or rabbits (section 10). Phenomena cognate with tolerance are considered. The partial inhibition of transplantation immunity which is caused by injecting adult animals with variously modified antigenic matter differs fundamentally from tolerance in mode of origin, for the substances which enhance the growth of homografts after administration to older animals have no power to confer tolerance upon embryos, and the substances which cause embryos to become tolerant merely cause adults to become immune (section 11 \cdot 1). It is argued that the iso-antigens responsible for transplantation immunity should be sharply distinguished from those specialized end-products of differentiation which are iso-antigenic because they are potentially auto-antigenic, and which are potentially auto-antigenic because the antibody-forming system has no opportunity in normal development to become tolerant of their action (section 11 \cdot 2). The phenomenon of tolerance is considered for its bearing upon the relationship between mammalian mother and foetus; upon the different stages of development at which immunity to different antigens may arise; upon the antigenic and genetic composition of the different tissues of a single individual; and upon the fate of iso-antigens in normal life (section 11 \cdot 3).