difficulty with which the cells of the hæmorrhagic tumours establish themselves. The following facts can be stated with certainty : 1. By the inoculation of large doses of tumour tissue immunity is induced against other tumours. 2. By the inoculation of small doses, consisting of unground tumour tissues, numerous hæmorrhagic tumours, 37 out of 48, have been transferred in London. Hertwig and Poll, who have successfully transplanted a hæmorrhagic tumour, also use only particles of tumour the size of a millet-seed or of a peppercorn. Perhaps the unfavourable effect of larger doses continues to exist to a certain extent during continued transplantation, if in a less degree than for spontaneous tumours. For, one often notices that in a series of transplantations of the same tumour with large and small doses respectively, the small doses yield the higher percentage of successful inoculations. Retrogression after transitory growth can often be observed with tumours arising from larger doses. In my own experiments, I have seldom exceeded 01 to 0.15 ccm., in order not to obtain series too limited in number: perhaps, the susceptibility to the size of dose which exists for spontaneous tumours would manifest itself also for the daughter tumours obtained in the course of continued propagation, by further raising the amount of the dose. Generally speaking, tumours appear to diminish in their susceptibility to dosage during the course of continued transplantation. How important the influence of dosage is in the case of re-inoculations also, will be demonstrated later in more exact detail. Some statements may be made here of the macroscopic and microscopic features of hæmorrhagic tumours. Apolant's statements with regard to hæmorrhagic spontaneous tumours have been confirmed in essential points and extended to the propagated tumours. The site of these spontaneous tumours is the same as that of the other mammary tumours, viz., mostly on the flanks and the rump, occasionally towards the back. The tumours are more or less movable on the subjacent tissues and raise the skin with which they become adherent when of any size. Their hæmorrhagic nature can be recognised usually by inspection because of their bluish colour. The skin is often greatly stretched through the presence of distension cysts which may be light or dark red in colour. When such cysts rupture reddish serum or a bloody fluid is discharged, an entrance is afforded to bacterial infections, and local ulceration or general septic conditions may follow. The inoculated tumours behave on the whole in a similar way. At first they are moveable on the subjacent tissues and under the skin. On further growth they become fixed and the skin is thinned. Very frequently quite similar hæmorrhagic and serous cysts develop which are accompanied by the same dangers from infection and ulceration. When we cut through such a tumour, we find it of a very soft consistence for the most part, and a chain of hæmorrhagic and serous cysts running through it and varying both in number and size; the smallest appear as pin points, the largest about the size of a cherry and containing either fluid or glutinous grumous blood. The tumour tissue appears pinkish-white and very soft in those parts where hæmorrhages are scanty. Should metastases of the lungs be apparent macroscopically, their aspect varies very much. The smallest look reddish-white, the largest, which in some circumstances can comprise a whole lung, are likewise uniformly hæmorrhagic. Naked-eye metastases in other organs have so far not been discovered. Microscopically, the tumours in question belong to the groups which Apolant has described as adenoma-cysticum ædematosum sive hæmorrhagicum, and cystocarcinoma hæmorrhagicum. The former, which, as Apolant states, frequently contain distended glandular spaces (Apolant's cystadenoma simplex) can be shown to be derived from the adenoma simplex, by studying the varying degrees of the change presented in the parenchyma and stroma. Distension of the lumina of the alveoli of the parenchyma often occurs through the accumulation of thick opaque liquid or of serous secretion; the epithelium lining the lumina is thereby flattened, and neighbouring cavities often become confluent, thus forming large cysts. Apolant distinguished several processes in the scanty stroma. Degenerations of the stroma occur. Lymphatic obstruction with resulting oedema is most frequent, then the alveoli are not only separated from each other but the histological picture is also confused, the alveoli being distorted and elongated through the pressure. The pressure frequently causes atrophy and destruction of the alveoli. Cysts likewise occurring as the result of the process just described, may look at a first glance very much like secretion cysts, although in reality they are completely different in nature. The non-epithelial origin of such cysts can be demonstrated before the histological picture is too extensively blurred through modifications, by means of the vessel which traverses the "pseudocyst" and the thin connective tissue lamella which intervenes between the fluid and the epithelial columns. Changes in the blood-vessels determine more than anything else the histological pictures presented by the tumours we are considering. The thin-walled capillaries which run in the delicate stroma are frequently dilated, becoming, particularly in the oedematous parts, huge blood-sinuses, or, by reason of the limited opportunity for expansion, exhibiting aneurysmal dilatations. Such dilatations of the bloodvessels proceed, here and there, to such an extent that the picture of a cavernous angioma is reproduced, and, in such cases the septa between the sinuses are formed by the compressed adenomatous tissue. All these changes predispose to rupture of vessels and to hæmorrhages, their extent varying greatly in individual cases. A distinct rent may be seen frequently in the capillary wall of those vessels which traverse the oedematous spaces, and through it the blood streams into the pseudocysts till eventually it quite fills them, and ultimately contributes to their further distension. The blood then escapes from these pseudocysts and courses still farther between the epithelial columns and may even penetrate into the genuine epithelial cysts. Finally, large areas of the normal tumour-structure are completely riddled with hæmorrhage. The constituent parts of the tumour thus cut off and enclosed within a hæmorrhagic area, may completely degenerate as the result of the pressure of blood and the removal of sustenance. The contents of the cysts present all transitions from blood-stained serous liquid to fluid or coagulated blood as the result of the mixing of blood with oedematous fluid and epithelial secretion. According to Apolant a smooth-walled cyst may finally develop from what at the outset had the structure of an adenoma, the contents of such a cyst being a thick chocolate-brown mass consisting of epithelial debris mingled with blood, while only quite insignificant remains of the original adenomatous structure remain attached to the connective tissue of the walls of the cyst. According to Apolant, the cystocarcinoma hæmorrhagicum is derived from an adenomatous tumour which has been modified by hæmorrhage in the way described, and in which atypical epithelial growth has occurred concomitantly with the hæmorrhages. Then compact epithelial complexes or garland-like columns of cells intervene between the hæmorrhagic areas described above, and all transitions exist between the solid and the adenomatous structure. Degenerations and hæmorrhages arise in the manner already indicated for the pure adenoma, and as the end result cysts filled with blood may form in like manner. Apolant repeatedly asserts that no sharp distinctions exist between the adenomata and the carcinomata. He found typical carcinomatous epithelial cords in the capsule of hæmorrhagic cysts, the inner wall of which bore remains of adenomatous tissue. Apolant has, as a matter of fact, "comparatively seldom come across a mouse tumour bigger than a cherry that did not in some place or other, but as a rule over wide areas, exhibit a carcinomatous structure." After my investigations relating to the 35 hæmorrhagic tumours, with which transplantation was attempted in the Imperial Cancer Research Laboratory, I can fully endorse Apolant's histological descriptions and therefore I can omit detailed descriptions of the separate tumours. Nevertheless I wish to emphasise that in my judgment a distinct division between hæmorrhagic adenomata and carcinomata appears impossible, especially after the microscopical examination of hundreds. of tumours obtained in transplantation experiments. Useful as Apolant's systematic classification is, we must bear in mind the fact that it is a purely histological division, and insist that it deals with tumours identical biologically, and only slightly varying in their microscopical appearances. In the case of the tumours I examined, like Apolant, I found with great constancy histological pictures which were difficult to reconcile with their purely adenomatous nature. The biological behaviour of the tumours during transplantation, to be mentioned later on, and their formation of metastases, as well as their powers of recurrence, are all evidence that they form one homogeneous group. The general features of the sporadic hæmorrhagic tumours at my disposal are shown in the microphotographs of tumour 19 (figs. 3 & 4). In many sporadic hæmorrhagic tumours the adenomatous parts preponderate, but, in quite as many, large solid epithelial alveoli are the outstanding feature, and then such solid alveoli often exhibit a curious arrangement of lumina, viz. a sieve-like puncturing of the epithelial mass, owing to the presence of little secretion spaces and the grouping of the cells around them (fig. 5). Portions with an acinous structure are connected at numerous places in the preparation with compact epithelial masses (figs. 6 & 7). At a first glance one might suppose, in agreement with Apolant's presentation of the facts, that we were concerned in such cases with the conversion of what histologically is an adenomatous formation into an alveolar or carcinomatous. But closer examination of the histological appearances shows that their evolution is exactly the reverse. The acinous portions. arise from the alveolar, and correspond to a differentiation (maturation) of the tissue of the tumour. This process may be observed |