influence on growth of ligature of the main blood supply to malignant tumours in the human subject. We feel the extent of the investigations justify the conclusions appended. CONCLUSIONS. Complete obliteration of the blood supply to a tumour means sphacelation, separation, and cure. Complete obliteration is impossible if there is infiltration of neighbouring structures such as is noticeable in the case of the mice. experimented on when muscle is infiltrated. Complete obliteration is only comparable to complete removal of the tumour, and this being so, it must be looked upon as very inferior to the usual mode of treatment by extirpation. In cases where there is a tendency to spontaneous cure, partial operations may prove successful, and the disappearance of a tumour may be even hastened. Such is the probable explanation of the success of partial operation in the case of mouse no. 2 of the batch 76 F shown in fig. 4. As far as is yet known there is no means of recognising the tendency to spontaneous healing in the living mouse, though by the consideration of the behaviour of a batch of tumours as a whole and of the previous history of the tumours* the possibility of its occurrence may be surmised. Were it possible to follow the natural fluctuations in the growth of tumours of the human subject with the same accuracy as in these experiments, it appears permissible to surmise that valuable indications might be obtained as to when surgical interference could be most advantageously undertaken.. There is no reason to suspect that the interference with these tumours as carried out in this series of experiments in any way increases their malignancy. When continuation of growth takes place it is of no greater rapidity than in the other tumours observed as controls, nor is there greater tendency to metastatic deposit. On the other hand, neither the partial absorption of a tumour nor its complete separation by sphacelation in any way influences the growth on the opposite side of the animal, of a previously established control tumour of the same histological structure. The mice in which tumour material has been Those interested in the study of the growth of cancer in mice are referred to a paper "Experimental Analysis of the Growth of Cancer," by E. F. Bashford, J. A. Murray and W. H. Bowen; Proc. Roy. Soc. B. vol. 78, 1906, and "Die Experimentelle Analyse des Carcinomwachstums," Zeitschrift für Krebsforschung, Bd. v. 1907. absorbed after operation exhibit the specific protection to subsequent inoculation as already described. Whether or not metastases may be prevented from establishing themselves remains a moot point. Whereas partial operations may result in diminution in size, continuation of growth is the invariable rule and is usually associated with early ulceration owing to interference with the vascular supply of the skin. The continuation of growth proceeds from the maintenance and compensatory increase in the size of the nutrient connections already existing, rather than from the establishment of new sources of supply. It is unnecessary to enlarge upon the difference between obliterating the blood supply in artificially propagated tumours where, as was noted at the beginning, the circumstances are ideal, and doing so in the case of a sporadic growth, not encapsulated but on the contrary infiltrating along fascial and muscular planes, and possessing a blood supply from many sources. The majority of these sources are invisible without a long and unjustifiable dissection. A TRANSPLANTABLE SQUAMOUS-CELLED CARCINOMA OF THE MOUSE.* By J. A. MURRAY, M.B., B.Sc. IN the account of spontaneous cancer, the features of a spontaneous squamous-celled carcinoma of the axilla were described on a previous page, and mention made of its transplantability. As this tumour (No. 32 of our transplantable strains), has been made use of frequently in our experiments, it is necessary to give a fuller account than would have been appropriate in the general paper, of the histology of the primary and transplanted tumours and of the course of its experimental propagation. Squamous-celled carcinoma takes a special position among the malignant new growths of vertebrates. It has apparently a wider zoological distribution than any other form of cancer, and has been observed frequently in many races of mankind in whom cancer is found with difficulty. In addition, the histological appearances of squamouscelled carcinomata are so characteristic and familiar, that this type of new growth has come to stand as a paradigm of all the epithelial new growths. This is due in part to the circumstance, that the majority of the smallest human carcinomata which have been investigated belong to this group, their superficial position rendering early recognition easy. From a consideration of all these points the advantages of a material of this type for experimental work are great and obvious. We have given a preliminary account in collaboration with Bashford and Haaland of the main points of interest in connection with this tumour, in a short paper which appeared in the 'Berliner klinische Wochenschrift' in 1907, No. 38. Conjointly with Haaland, it was also the subject of a communication to the Pathological Society of Great Britain and Ireland. The tumour was situated in the left axilla of an old female mouse. In form it was elliptical, measuring 2 cm. by 15 cm. by 1 cm. in thickness. On June 6th, 1906, the mouse having been anæsthetised with ether, the tumour was partially excised, and the wound closed. The skin covering the growth was adherent to its outer surface and was freely removed along with it, Bleeding had occurred from the surface, and, the superficial part of the tumour with the skin attached was preserved for microscopical examination. The microscopical appearances are described along with the other spontaneous tumours, but some additional points will be referred to later. The deeper parts, consisting of softer and firmer areas in which minute opaque white spots were scattered, were used for transplantation, with the exception of several slices preserved for histological examination. Transplantation was carried out by introducing small fragments of 0.01 to 0.02 gr. each, under the skin of the back in 201 young normal mice. In addition a similar fragment was inoculated into the spontaneously affected mouse herself (Mouse). During the first five days after inoculation 5 or 6 mice were killed each day, and the graft excised and preserved for examination of the processes at the site of inoculation. Four tumours developed in the 156 mice which were alive three weeks after the inoculation; they were the size of small shot, hard and movable under the skin. Three of these mice died during the next four weeks, and their tumours, being still too small for transplantation, were preserved and examined histologically. When the spontaneously affected mouse was killed five weeks after operation, the fragment inoculated was found and preserved. The tumour in the fourth mouse grew very slowly; after an interval of three months it had attained the size of a pea. During the next month, after having remained of the same size for more than a week, it diminished in size, till, when four months had elapsed since inoculation, it could be felt no longer. Three weeks later (Nov. 10th, 1906) a small nodule could be felt again in the same situation and it grew to the size of a pea in the succeeding four weeks (Dec. 12th, 1906). This tumour now grew more rapidly and had attained a diameter of 1.5 cm. by January 8th, 1907, seven months after inoculation. On this date the mouse was anæsthetised, and the tumour partially excised. Transplantation was effected into 64 mice. Seven tumours developed, and from now onwards the further transplantation was effected without difficulty, compare chart (fig. 1). The primary tumour grew again after the partial excision on June 6th, 1906. Five weeks later, when the mouse was killed, it weighed 15 gram. With part of this recurrent tumour, 53 mice were inoculated by the same method as before, but no tumours developed in the 52 mice which survived. The remainder of the material was preserved in situ with the whole mouse. The tumour which developed in the primary transplantation and was operated upon, did not grow again, and no trace could be found of the nodule left behind at the operation of January 8th, when the mouse was autopsied four weeks later. Fig. 1-showing graphically the percentage of success of a number of series of experiments commencing with the primary transplantation— illustrates the phenomenon which Bashford and Murray recorded in accustoming Jensen's Danish tumour to English mice (1904) and which Ehrlich has also described (1905) in connection with the first steps in propagation of other spontaneous tumours, and has designated an increase in virulence ("Virulenzsteigerung"). It indicates also the nature of the reservations which must be made, when the growth of transplanted carcinoma cells is compared to the cultivation in vivo of a pathogenic micro-organism in a succession of susceptible animals. The increase in percentage of successes and in rapidity of growth, can be referred with a high degree of probability to the rapid increase in adaptability of the parenchyma cells at this period, which is essential for continued transplantation. As a consequence, a progressively larger number of cells survive the injury inseparable from transplantation, and, the size attained by the tumours in equal times is therefore greater. Once this preliminary stage has been passed, the rate of growth and percentage of successes fluctuate between somewhat wide limits, in the manner we have described along with Bashford and Bowen in the "Experimental Analysis of the Growth of Cancer (v. post.). The limits to which this process of adaptation is subject, and the effects of alterations in the dose of tumour material inoculated on percentage of successes and rate of growth of transplanted tumours, are described in a later paper. دو The transplanted tumours grow with great rapidity. Tumours of one to one and a half grams, in weight are usually present after ten days in every series, even when the initial doses do not exceed 0.03 gram. The tumours grow with a great tendency to infiltrate the muscles of the thoracic and abdominal walls and fungate into the serous cavities, features which are illustrated in a later paper in this report. One of the M |