These results stand in marked contrast to the records of the control, which show all normal conceptions and normal offspring.

The second table shows the results of successive matings in ten of the females. The varying success of the conceptions in the same individual are striking.

Nice has quite recently recorded a similar series of experiments with alcohol on mice. Alcohol was given to the mice in their food. Nice finds that while there was a certain fatality among the offspring from alcoholic parents as compared with those from normal parents, where there was no fatality, yet nevertheless the offspring of alcoholic parents actually grew faster than those from the control. This may indicate that alcohol is not equally poisonous in its effects upon all animals, as might really be expected. The germ cells of mice. may be more or less immune to the action of alcohol. It is well known that the action of alcohol is different in its effects on individuals from different human families. Some alcoholics show chiefly nervous disorders, hallucinations, delirium, etc., while others may have no nervous symptoms but exhibit various derangements of the digestive glands, kidneys, etc., or may have a fatty degeneration of almost all organs.

Finally it may be concluded that the experimental evidence goes to show that the development of an offspring may be modified by either treating the parents so as to affect their germ cells or by subjecting the developing embryo itself to unusual or injurious conditions.

The causes of many congenital defects are therefore known. It is possible to control embryonic development to such an extent as to produce abnormal structures. May not the proposition be reversed and unfavorable environments be treated in such a manner as to render them favorable to normal development? Diseased mothers may in some cases, at least, be made fit for the function of reproduction.

The regulation of structural disease becomes then a problem of morphology and hygiene. It is most important, and must precede,

or go before, the selective mating of human beings or the eugenics movement. The most intellectual will rarely submit to direction in choosing a mate, yet every productive pair will welcome any possible means of improving the quality of their offspring.

While preventive measures are being used to protect the postnatal life of the individual, why not guard as far as possible its prenatal development?

CORNELL UNIVERSITY MEDICAL COLLEGE,

DEPARTMENT OF ANATOMY,

NEW YORK CITY.

AN AVIAN TUMOR IN ITS RELATION TO THE

TUMOR PROBLEM.

BY PEYTON ROUS, M.D.

(Read April 19, 1912.)

The behavior of malignant tumors has stimulated many hypotheses as regards their causation and yet in some of its phases has appeared to disprove all. The suggestions afforded by other normal or pathological growth processes lead in numerous and diverse directions. Today, despite an immense accumulation of data, the solution of the tumor problem waits upon fresh findings; and to foretell the line of research which will yield these findings has not seemed possible.

66

The successful transmission of neoplasms of the lower mammals, a few years ago, seemed at first to carry with it the immediate solution of the problem. But this did not prove to be the case. In order to obtain a tumor-strain for investigation, an animal with a spontaneous" tumor was required; and the transmission of the growth was soon found to involve a transplantation process, as genuine a transplantation as that of skin or other normal tissue. A new tumor in the strict sense was not thus engendered, but to a portion of the old another host was given. All efforts to separate out a cause for the neoplasm or to transmit it by other means than by graft of the living neoplastic cells were unsuccessful. The consistently negative results of such work, together with the general behavior of the transplanted neoplasms, have led many investigators to forego the idea of an extrinsic cause for malignant tumors in general and to attribute them to some inherent cell-perversion, or else to a cell-derangement precipitated by factors temporarily active. But it may be pointed out that the basis for such a conclusion so far as it rests upon experiment, rests upon work with the tumors of few species, Those of the rat and mouse have been employed almost exclusively.

The findings here to be presented were obtained in the study of a malignant tumor of the chicken, which closely resembles in its general characters the mammalian neoplasms, including those of man. That such growths exist has been generally recognized; and their status as true tumors has been established. But like the neoplasms of other birds, of reptiles, amphibians, and many mammals, they have remained almost unutilized for research. Our tumor of the fowl proved transplantable and has thus far been observed in several hundred chickens. It is a connective-tissue growth, a spindle-celled sarcoma. From its tissue there has been isolated a causal agent, ultramicroscopic in some, perhaps in all, of its forms, and undoubtedly a living organism. Though the agent gives rise to the sarcoma, and accompanies the growth, it does not take any obvious share in the disease phenomena. These are referable to the behavior of the neoplastic cells, a point now to be illustrated.

The original sarcoma arose in a young fowl of pedigreed, pureblood stock, and its transplantation was successful only in this fowl's blood-relations. A similar peculiarity has been often observed on the transplantation of normal tissues, but it has not been noted in association with the transmission of diseases caused by a parasite. After propagation in several successive hosts, the sarcoma became less precise in its demands and could be transplanted to non-related fowls of the same variety. But like certain delicate tumors of mammals it was for a long time transmissible only within the limits of this variety, and at the latest test still grew most readily in such hosts. All attempts to transmit it to animals of other species have failed.

A transplantation of neoplastic tissue is involved in the growth's transmission under ordinary circumstances, and only by special means has it been shown to be unnecessary. Ordinarily when bits of the sarcoma are placed in a new and susceptible host they survive, are vascularized, and proliferating, form a new tumor. The multiplication of the implanted cells obviously suffices to produce the neoplasm. In the histological pictures there is no indication that the elements of the host ever become incorporated as true neoplastic tissue. No tumor arises in hosts so unfavorable to the engrafted tissue that it dies.

The growth consists of spindle-shaped cells supported by a scanty, vascular framework; and the sole differentiation which these cells undergo is to an attenuated form with the production of a few intercellular fibrils. In the disposition of the cell-strands and bundles there is no suggestion of focal arrangement such as frequently indicates, in the case of the granulomata, the presence and position of an exciting cause; while at the growth's borders a cellular reaction is practically absent. The tumor elements multiply rapidly by mitosis and amitosis, and the neoplasm grows, not only by expansive enlargement but also through an active invasion and replacement of the normal structures by tumor cells. In the course of the invasion tumor cells frequently penetrate the walls of blood or lymph-vessels, and are freed in the circulation. By their transportation, lodgment and growth secondary sarcomata are caused at points distant from the primary mass. This important characteristic of tumors in general has been placed beyond doubt as regards the avian growth by means of direct experimentation. The host, which at first seems unaffected by the tumor, emaciates as the growth increases in size, and, if it escapes intercurrent processes eventually dies in coma.

The conditions which determine the success or failure of the sarcoma when transplanted to a new individual are in general referable, as are those of its behavior and dissemination, to the sarcoma cells as such. The influence of variety of the host and of blood-relationship has already been referred to. Young hosts prove most favorable, as for all transplantable tissues, normal or neoplastic. Hosts which are ill of causes that involve emaciation are relatively unfavorable, a circumstance noted in its relation to mammalian growths by other workers, and especially interesting because these hosts are more susceptible, as a rule, to the frankly infectious processes. A certain proportion of hosts, although of the proper variety for the tumor's growth, manifest a resistance such that it does not develop when implanted in them; while others in which the growth has developed and retrogressed are completely resistant, for a time at least. Similar types of resistance have already been demonstrated for the tumors of rats and mice. Furthermore they