1 cm. in diameter in the left groin. This was extirpated and the wound healed promptly. A second growth appeared in the left axilla. This also was extirpated and proved on microscopical examination to have the same structure. The mouse had to be killed a fortnight later, and at the autopsy two masses were found on the right side of the chest. One of these appeared to be from its form and position an enlarged anterior mediastinal gland. The second had the shape of the upper lobe of the right lung, and had apparently replaced that structure entirely.

19

The other localised example of this condition occurred in a mouse (1) with spontaneous hæmorrhagic-carcinoma of the left axillary mamma. The growth which now concerns us filled the whole upper part of the left side of the thorax, and the pulmonary artery and vein Fig. 9.

Fig. 10.

FIG. 9.-Transverse section of a mouse with spontaneous carcinoma mamma (1) and mediastinal lymphadenomatous growths. X 3.

a. Hypertrophied normal mamma (lactation). b. Cystic mamma anterior to mammary carcinoma. c. Mass of growth, lymphadenoma replacing left lung, lower lobe. X 3.

FIG. 10.-Transverse section of same mouse through growth at level of diaphragm. a. Normal mamma. b. Mammary carcinoma.. c. Posterior end of lymphadenoma in left lung. d. Metastasis of mammary carcinoma in right lung, upper lobe.

could be traced into it from the heart. The rest of the lung tissues were completely destroyed, only delicate isolated elastic fibres widely separated from each other, being distinguishable. Extension had also occurred on the right side of the chest along the peri-bronchial lymph spaces. Histologically the tumour consisted of closely packed masses of small polygonal cells with large nuclei (fig. 8). With the exception of this case artificial transmission was attempted with these

growths, both by subcutaneous transplantation, intraperitioneal and intravenous inoculation. In none of the animals was any effect produced.

These four cases came from four separate breeders at long intervals. A development of the disease has not been observed in the animals kept in the laboratory.

MALIGNANT NEW GROWTHS OF THE MAMMARY REGION.

The references in the literature to the spontaneous mammary carcinomata of the mouse are closely bound up with the development of experimental cancer research by the successful transplantation of these tumours. The year 1903 marks the modern development of experimental cancer research as exemplified by the papers of Loeb, Jensen, Borrel, Bashford, which rapidly followed one another. Already with the first contribution by Morau in 1894, in which the first successful implantation of mouse tumours was recorded, the structure and classification of the new growths with which he worked received a careful description accompanied by several excellent figures. Five years had then elapse since Hanau in 1889 had first directed attention to this method of study by his successful transplantation of a squamous-celled carcinoma of the vulva from rat to rat, and had thereby indicated the direction in which further success was to be expected. Nearly ten years elapsed since Jensen in 1903-04 carried the subject a long step forward by his masterly papers on the process of transplantation, studied on the alveolar carcinoma, now generally known by his name and propagated in laboratories all over the world. Borrel simultaneously published an account of successful transplantation of adeno-carcinomata of the mouse. He described, with beautiful figures the structure of the primary and transplanted tumours, and the embolic metastases which they formed in the lungs. In addition he recorded squamous-cell carcinoma of the jaw with lymph gland metastases, and several cases of malignant lymphoma. Bashford's earliest communications followed. immediately on those of Borrel and were in time succeeded by those of Michaelis, Ehrlich and Apolant, Clowes and Gaylord. In the interval between Morau and Jensen, pathological-anatomical descriptions of mouse tumours were published by Livingood who recorded primary adenomata of the lung, sebaceous adenomata, and three subcutaneous adeno-carcinomata of the mamma, and by Eberth and Spude who recorded tumours in three mice of the same parentage. These tumours they regarded as endotheliomata, led astray as Apolant was able to show

later by the diffuse distribution of the mamma in the mouse, the acini of which they figured as lymphatic vessels.

Michaelis distinguished three types in the thirteen spontaneous tumours which he had examined in 1905: alveolar carcinoma, adenocarcinoma, and malignant adenoma. In 1905 also, Haaland published an account of thirty spontaneous tumours obtained by Borrel and himself in the Pasteur Institute. He described the tumours as adenomata or adeno-carcinomata, and pointed out that alveolar areas frequently occur even in tumours which have a distinct acinous structure. He also described Borrel's and several other cases of squamous-celled carcinoma of the jaw, primary adenomata of the lung, and malignant lymphomata as well as an epitheliomatous tumour of peculiar structure. In addition he gave a very full description of the lung metastases found in animals with spontaneous and transplanted tumours both of the Paris strains and of Jensen's.

In the Second Scientific Report of the Imperial Cancer Research Fund with Bashford and Cramer, we recorded alveolar and adenocarcinomata as well as a tumour of the cystic and hæmorrhagic type which could be transplanted successfully.

In 1906 Apolant published a monograph on 276 tumours occurring . in 221 mice at Ehrlich's Institute. All his tumours were found in adult or aged female mice and corresponded in their distribution to the mamma. He pointed out the great similarity they present in histological structure to the thyroid tumours in man, and gave a classification in which he distinguished two main groups, the adenomata and the carcinomata. In the adenomata he distinguished adenoma simplex, cystadenoma, cystadenoma-papilliferum, and cystadenoma œdematosum seu hæmorrhagicum, in which secondary changes in parenchyma or stroma alter the general characters of the growths. The carcinomata he divided into carcinoma simplex alveolare, cystocarcinoma hæmorrhagicum, carcinoma papillare, and fissure-forming carcinoma. Apolant lays stress on the combination of all these types in one and the same tumour and in single tumours of the same animal when multiple, a condition present in 12 per cent. of the animals. He considers the transformation of the adenomata into carcinomata as of frequent occurrence, and interprets in this sense most of the cases in which both histological types occur in the same tumour. He also holds that direct development of alveolar or other carcinoma occurs, without an intermediate adenomatous stage. There can be no doubt that Apolant recognised in this possibility of transformation of one type of growth

into another the key to the great variability and complicated histology of the mouse tumours. In the sequel the details of such transformation will again be discussed on the basis of personal observations. The papers which have appeared since Apolant's monograph, recognise as established the carcinomatous character of the tumours of the mammary region, and are principally concerned with the phenomena and results of artificial propagation. They will be referred to later so far as they concern the subjects of subsequent chapters along with the corresponding new observations.

Before proceeding to an account of the adeno-carcinomata and alveolar carcinomata of definitely glandular origin, it will be convenient to review briefly, tumours of the mammary region of different histological type, which have occurred in our material.

ANGIOMA OR ANGIO-SARCOMA.

The tumour, of a dark blue colour, was situated in the left inguinal region of an adult male mouse. It was almost entirely formed by a cyst containing blood clot and fluid blood. Microscopical examination (fig. 11) showed no evidence of epithelial elements, but only irregular spaces lined by flattened pavement-like epithelium and filled with blood. In some of the larger spaces were recent thrombi. Small islands of large epithelioid cells lay between adjacent blood-spaces. The whole structure of the growth indicated its probable development from the endothelium of blood-vessels. A small nodule present in the right lung had the structure of an adenoma of the lung. Both tumours were transplanted, the inguinal growth into 82 mice, the lung nodule into 8 only. In both series no tumours developed.

CHONDRO-OSTEO-SARCOMA OF MAMMA.

Up to the present, tumours of the mouse in which cartilage or cartilage and bone are present have been described by Ehrlich and by Haaland. In Ehrlich's case the primary tumour was intraperitoneal, and was easily transplantable. Ehrlich regarded an origin from retained foetal structures or from a congenital foundation as probable. The transplanted tumours consist almost entirely of hyaline cartilage growing as rounded masses into which blood-vessels penetrate. In older tumours these vessels are enormously dilated and hæmorrhages

Microphoto by R. Muir. FIG. 11.-Mouse. Angioma or angio-sarcoma of mamma. The growth consists of irregular spaces, partly filled with blood and lined by endothelium. In places solid cell-masses are formed. X

90
1

FIG. 12.-Mouse. Transplanted chondroma, from a preparation presented by Professor Ehrlich. Note the arrangement of the cartilage cells in columns, with delicate septa of cartilage matrix between.

40

1

UNIV

OF