The outstanding feature in the resistant animals was their failure to supply the characteristic connective tissue and vascular scaffolding, but it was not clear to which one of two factors the absence of the reaction was to be ascribed. Thus it was conceivable that either the tissues had been altered in such a way that they no longer replied to the stimulus of the cancer cell, or that this cell itself had been robbed of the power to incite the specific stroma reaction. It seemed justifiable to explain the failure by the assumption that the malignant cell was deprived of its chemotactic properties, and eventually of its power of assimilation and growth. There must be present either in the circulating fluids or the tissues of resistant animals something able to inhibit chemotaxis; but this unknown agent could not be a very active cell poison, because the cancer cells retained their proliferative power for from seven to ten days, and because, furthermore, it was those at the periphery of the graft which persisted in their growth, although they were the very ones most exposed to the influence of a hypothetical poison. Burgess (6) described necrosis of the central part of the graft, degeneration of the stroma, and invasion of the tumour mass by new capillaries and proliferating connective tissue cells, not only in normal mice but in those of a naturally resistant breed as well. In the latter, however, after about a week of active growth, the tumour became surrounded by an inflammatory exudate which impaired its nutrition, and apparently as a part of this reaction there occurred in many of the non-susceptible mice an over-production of fibrils on the part of the more centrally located portions of the new stroma. Peripheral extension ceased, central necrosis at the same time advanced, and ultimately the whole tumour underwent destruction and absorption. The author pointed out that the increased connective tissue reaction in the graft which he had observed in refractory mice, where Russell, on the contrary, had established an absence of reaction, might have ensued because his own mice belonged to very closely related varieties capable of interbreeding freely. It is hardly necessary, however, to go so far afield for an explanation of the difference. There has been an unfortunate failure of late on the part of several authors to distinguish between mice which are immune at the moment of implantation, so that the tumour never gains a foothold in them, and those in which spontaneous absorption is preceded by a certain period of growth. The histological appearances substantiated for the latter cases by Gaylord and Clowes, Bashford and Murray, and others, vary so widely from those described by Russell in the case of grafts which have never established themselves as to indicate the existence of a distinct difference between the reactions manifested under the two conditions. Nor is further evidence of this dissimilarity lacking. Da Fano (9) found plasma cells distributed throughout the tissues of mice in which tumours were undergoing spontaneous absorption, as well as in those that had been unsuccessfully inoculated with sporadic growths. But when animals had once become resistant, a second inoculation of an immunising material was powerless to reproduce the reaction. Russell's investigations have been recently confirmed by Anitschkow (1), and their value and accuracy have, in fact, met with almost universal recognition, although an occasional doubt regarding their validity has found expression, as in the paper of Goldmann (14). The thesis which he defended, and which has been expressed perhaps more clearly in earlier papers (12, 13), was, that the presence of blood vessels within tumours represented the product of a defensive reaction on the part of the host; the subsidiary assumption was entertained that the vessels were endowed with the power of destroying tumour cells. The opinion was expressed that the method of removing grafts practised in Bashford's laboratory was suitable only for the investigation of the larger blood vessels, and that in its employment the smaller ones were necessarily destroyed. Still, fig. 18 in the Second Scientific Report and fig. 8 in Russell's article, representing fragments removed in this way, show a large number of capillaries at the margins of the grafts, most of them so fine in calibre as hardly to exceed the diameter of a red blood corpuscle, while capillaries no greater in size can be discovered without difficulty in the figures accompanying the present article. The adequacy of the method cannot, therefore, be gainsaid. Goldmann urged further against the validity of Russell's conclusions, that spontaneously receding tumours were well provided with blood vessels. There is, however, a fundamental difference between the failure of a graft to establish itself in a new host, and the regression of a tumour which has begun to grow. How great this difference must be is shown by the fact that it absolutely nullifies all efforts to bring about the disappearance of established growths by means of the inoculation of embryo skin, for example, which nevertheless is capable of evolving an almost complete immunity against implantation. The purpose of the present paper is to extend the investigation of early stages to another species, and to describe in detail the reactions taking place in and about fragments of the Flexner-Jobling adenocarcinoma of the rat, after its transplantation into normal or resistant animals. Immune rats were obtained by selecting those which had proved themselves resistant to one or two inoculations of the tumour in question, or by subjecting animals to previous treatment with 0-20.3 c.c. of an emulsion of rat embryo skin. The rats injected with this latter material were divided into three groups, two of which received no further immunising inoculation, while in the the third, preliminary treatment was repeated 21 days later with the tumour under consideration, to which they were found resistant. The introduction of the grafts destined to be excised for the study of early stages took place in all cases two or three weeks after the last immunising treatment. The technique employed throughout the investigation was identical with that which Russell has already described in full. Small fragments, selected from the healthy margin of a growing tumour, were deposited in the subcutaneous tissue of the axilla by means of a hollow needle inserted into an area in the groin previously epilated to prevent the introduction of hairs with its consequent confusion of the histological picture. Grafts introduced in this manner were removed with sharp curved scissors at varying intervals after inoculation, and in the extirpation as much of the neighbouring tissue as it was possible to include was maintained in undisturbed connection with the implanted fragment. Six series in all were employed, and several grafts, usually three from normal, and an equal number from resistant rats, were procured in every one of the series for each of the stages investigated. The specimens. thus obtained were fixed in Borrel's fluid and cut in serial paraffin sections, which were stained in Heidenhain's iron hæmatoxylin. It would be hard to emphasize sufficiently the imperative necessity of serial sections in work of this sort, for only by their aid is it possible to resolve an assortment of complicated phenomena into anything approaching an orderly sequence of events. From the large number of sections available, there were chosen for purposes of illustration those affording the best general idea of the conditions extant at the particular period that was to be described; hence, all the appearances discussed for any one period may not be found in the corresponding illustration, for the text attempts a description of the state of the average graft. E The tumour utilised for these experiments was an adeno-carcinoma which was discovered in 1906 by Flexner and Jobling in the left seminal vesicle of an adult white rat. An extensive description of the histology of the primary growth, and that of the transplanted tumours belonging to the earlier generations, has been given by the authors (10) just mentioned. Through the kindness of Drs. Flexner and Jobling, the tumour was received at the laboratory of the Imperial Cancer Research Fund, where it has now been in cultivation for about four and a half years. While the yield of tumours attendant upon the inoculation of this growth into English rats varies enormously from one generation to another, it may be said that ten days after implantation from 85 to 100 per cent. of the animals have developed growths of from 0.2 to 0.5 gm. in weight. Many of these nodules, however, begin to recede soon afterward, and 50 per cent. of progressively growing tumours is but seldom realized, while in individual series all the growths may disappear spontaneously. As with the strain retained in America, so in the one grown here, a diagnosis was not readily to be made in the first few generations, although after a short time the growth proved clearly enough to be an adeno-carcinoma. The parenchyma, except at the growing edge, arranges itself in acini, and there is a frank tendency toward the assumption of a papilliferous character, which, however, is always curtailed by the onset of necrosis. The stroma, which in common with other fibrous connective tissue of the rat is significant for its collagenous nature, is well vascularized, and appears fairly cellular by reason of the large number of fibroblasts which are distributed through it. The histology of the adult growth is reproduced in fig. 1. The various phases of the connective tissue elements of the rat have been so fully and so accurately described by Maximow (18) that any further description would be superfluous. Gratts removed from normal rats twenty-four hours after inoculation exhibit changes of which those set forth in fig. 2 may be regarded as typical. The fragment is infiltrated with polymorphonuclear leucocytes, and in the parenchymal cells toward its centre degenerative changes have already commenced. These elements are shrunken, their protoplasm is filled with tiny granules stained a deep brown or even black by the osmic acid in the fixative, their nuclei have become smaller and irregular, and active proliferation is in progress only among those at the margin. The collagen fibrils of the stroma have fused into thick glassy bundles, and the great majority of the connective |