impossible to imagine that the differences brought out by our experiments on the fluctuations in energy of growth are due entirely to variable susceptibility of the animals. The simultaneous appearance of numerous spontaneous absorptions, and associated therewith the rapid diminution in the percentage of success in parallel experiments,* can only be referred to fluctuations in the energy of growth of the parenchyma cells. Any other explanation, and especially an explanation by assuming increased resistance of the animals, introduces an arbitrary assumption which further complicates the conceptions.

The resistance of the animals can be altered, and this change has been described as an active immunity. We shall continue to speak of resistance or refractoriness. The absorption of tumours which have grown for a time undoubtedly calls forth this alteration in a high degree. The absorption of inoculated material without obvious tumour formation has similar consequences. Increased resistance results when blood injected subcutaneously is absorbed. The refractory condition induced in this way is not so perfect as that following the absorption of tumour material. It is, however, perfectly definite, and removes any cogency which might attach to the argument that the induction of a refractory condition by absorption of growing tumour or living tumour cells indicates the existence of a hypothetical cancer virus. The evidence we have obtained that the blood of other animals is unable to induce such an alteration in mice, points to the same conclusion. The absorption of the tumours of animals of alien species when inoculated into mice is likewise devoid of effect.

All these actions are curiously limited to the production of an insusceptibility to subsequent transplantation. We have not been able to induce such an alteration in animals with growing inoculated tumours that they should become unsuitable for growth to continue once it has started. It would appear that the negative phase in growth of the tumour cells plays the principal role when spontaneous absorption occurs, the fall in the energy of growth and the resistance of the mouse together contributing to the final absorption. Once the cellular graft has been vascularised it is in a much stronger position with reference to conditions which may be unfavourable to it than is a graft in the days immediately following transplantation. This is much more

vulnerable.

*Roy. Soc. Proc., B, vol. 78, 1906. Graphic record on p. 296, and protocol, pp. 299 and 301.

The biological reactions described in this paper are effective by means of the body fluids. They are analogous to those that are now well known as specific hemolytic, cytolytic, and precipitin reactions. As pointed out in Second Scientific Report, 1905, pp. 32-33, they are of far greater delicacy, and have only been revealed by using living cells as indicators.

The experiments recorded in this paper give further evidence of the great delicacy and the very narrow range of variations in the soil furnished by the living animal, within which growth can proceed or is inhibited. Growth has only been hindered by the intervention of mouse tumour or normal mouse tissue, to an infinitesimal degree by the normal tissues of the rat, and not at all by those of less nearly-related animals. We therefore feel justified in concluding that these reactions are parallel to those others whereby blood relationship has been established (Uhlenhuth).

When we attempt to formulate an explanation of the nature and mechanism of the changes leading to inhibition of growth, it must be clearly kept in mind that the evidence is almost entirely in the direction of showing an alteration in animals whereby they are unsuitable for or inimical to the establishment of grafts. There is no satisfactory evidence of an induced action on growing vascularised tumours, but only one against newly-introduced grafts. Spontaneous absorption, depending largely on alterations from the side of the parenchyma, is sufficiently frequent to enforce caution in assuming an artificial alteration in the resistance of animals already carrying growing tumours. Complete absorption is necessary to induce absolute refractoriness to subsequent inoculation. Hence the action of various procedures in inhibiting the growth of grafts is not necessarily or even probably due to a direct action on the parenchyma cells.

The connective tissue reaction, which a cancerous graft elicits, is taken advantage of by the cancer cells and acquires characters specific for each sporadic tumour. It is immaterial whether this reaction is protective on the host's part, and only not effective because the slight differences between the tumour cells and normal mouse tissues do not sufficiently stimulate the protective process. The necessity which exists on the part of the cancer cells for this reaction if they are to continue to grow, renders it possible that a refractory condition may

* Once the tumour cells are dead they excite an energetic phagocytosis. Cf. figs. 43 48, 'Second Scientific Report,' Part 2, 1905.

indeed not have anything to do with an action against the cancer cells, but, on the contrary, be due to an alteration of the connective tissue of the host which hinders it from supplying the necessary connective tissue reaction. The possibility must also be entertained that the alteration is directed against the chemotactic influence which may be presumed to be exerted by the cancer cells on the connective tissue of the host. Our own experiments have not yet given satisfactory direct evidence of an anti-toxic or anti-cellular action, but we are far from denying the possibility that such an action may be obtained.

The phagocytosis of formed cellular elements plays an important róle in inducing resistance; serum is impotent to produce resistance, blood corpuscles do so. The energetic phagocytosis which accompanies the spontaneous absorption of transplanted tumours, and which occurs in absorption after exposure to radium, speaks strongly for the conclusion that the processes are the same in kind when blood or tumour cells, being absorbed, produce resistance. But we are as yet unable to determine the extent to which agencies directed against the tumour cells themselves may assist in determining their early death in protected animals. Other experiments still in progress may be expected to clear up the relative importance of the parts played by the hypothetical inhibition of the specific stroma reaction, or of an equally hypothetical direct lethal action on the tumour cells.

THE NATURE OF RESISTANCE TO THE

INOCULATION OF CANCER.

By B. R. G. RUSSELL, M.B., Ch.B. Aberd.,

GEORGINA M ROBERT CANCER RESEARCH FELLOW, UNIVERSITY OF ABERDEEN.

In view of the fact that the serum of mice rendered resistant to the inoculation of cancer, had not been proved to exhibit a direct toxic action on cancer-cells in the test-tube, nor to possess curative properties when injected into mice bearing tumours, Bashford, Murray and Cramer attempted to elucidate the phenomena of resistance by experiments and direct observations "in vivo." They compared the results of exposing the cancer cells for varying intervals to the action of the tissues and fluids in living animals, in normal and resistant mice respectively. They observed that tumour tissues speedily died when implanted into immune mice. In continuation of these observations the following series of experiments has been made to ascertain if it were possible to find any constant differences between what occurred in cancer grafts implanted into normal and into resistant mice, which could be of service in throwing light upon the manner in which this resistance is produced. The method employed for this purpose has been the systematic examination of the processes at the site of implantation of cancerous grafts into normal and immune animals respectively. This consists in the inoculation of small fragments of tumour which are removed after varying periods, 6, 12, 24 hours, etc., and which are examined histologically after appropriate fixation and staining.

It was through the employment of this method of procedure, that Jensen was able to demonstrate conclusively that the tumour, which developed from inoculation, was produced entirely through an active proliferation of the epithelial cells introduced, and not through any

1 Centralblatt f. Bact. Bd. xxxiv. 1903.

transformation of the cells of the new host.

Bashford and Murray 1, who worked with Jensen's tumour and with numerous other sporadic tumours of the mouse, found by themselves, were able to confirm Jensen's results, and, always employing the method of "early stages," were able to extend our knowledge of the changes which occurred after transplantation of a tumour, to the other constituent of the new growth, namely the stroma. According to them, there took place during the first two and three days after the transplantation an extensive necrosis involving the stroma of the graft, which was later followed by a reaction on the part of the tissues of the new host, and which led to the revascularisation of, and the provision of a new stroma for the graft.

The systematic examination of what occurs at the site of implantation by this method involves a great deal of labour and patience on the part of the investigator before an amount of material can be collected sufficient to base conclusions on. The tumours which are to be studied must be propagated with great care so as to ensure a constant and adequate supply of tumour tissue free from extraneous pathogenic organisms. The growth of the tumours obtained by propagation must be carefully charted from week to week, in order that the past history of any particular tumour may indicate what result may be anticipated on transplanting it. In this way tumours can be selected as likely to give high or low percentages of daughter-tumours, as desired. Where a number of the inoculated mice are sacrificed in the days immediately following implantation, it is also important that the tumours developing in those left alive should be charted with equal care, in order that observations made on "early stages" can be controlled by the later consequences of inoculation; this is particularly essential when modifications are being sought for as the result of introducing a new experimental factor. For the purpose of this paper upwards of 4500 inoculations have been made and the results carefully followed in the manner described.

The actual technique employed in carrying out the inoculations and the removal of the implanted tissue for examination likewise requires the exercise of great care and patience during its frequent repetition, if comparable results are to be obtained. A detailed description appears desirable, since the technique itself is somewhat difficult until it has been sufficiently practised.

As regards the material to be employed, choose as rapidly growing a

1 Verhandlung des Komités für Krebsforschung Oktober 1903. British Medical Journal, Dec. 12, 1903. Roy. Soc. Proc. Jan. 1904. First Scientific Report Imperial Cancer Research Fund, 1904. Second Scientific Report, ditto, 1905, etc.