It is much to be regretted that dry preparations only at the beginning and at the end of the five year period of observation were at our disposal. It appears from the paper of Riequer as if in this case the lymphocytosis had established itself one month after the operation, and had lasted for a very long time, just as Kurloff has found in some animal experiments. Just as little as a polynuclear increase is abnormal, is an increase of the lymphocytes remarkable; and in this case the lymphocytic increase was recognisable after the end of the fifth year. The eosinophil cells oscillate at this period about the upper normal limit. From all that we know, it is probable that their number in the meantime had undergone an intercurrent increase.

The cases are more frequent in which a splenectomy has been undertaken on account of disease of the spleen. Amongst these, the clearest results are à priori to be expected from splenic cysts, since the part of the spleen not affected by the cyst formation often shews quite a normal structure, and therefore is physiologically active. On the other hand, the excision of chronic splenic tumours may be—for the blood condition—of no importance inasmuch as the function of the spleen may have previously long been eliminated by pathological changes.

Amongst these cases, we must in the first place mention the well-known and carefully investigated case of B. Credé. In a man 44 years of age the spleen was extirpated on account of a large splenic cyst. Within two months of the operation there developed a thoroughly leukæmic condition of the blood, exclusively brought about by the increase of the lymphocytes, as is seen from the results of Credé and the table contained in his paper. It is further remarkable that four weeks after the operation a painful doughy swelling of the whole thyroid appeared, which remained, with variations, for nearly four months. With the general recovery of the patient this shrank to a small remnant. We notice further that this very interesting swelling of the thyroid, which doubtless stands in the closest connection with the splenectomy, is nevertheless no constant accompaniment of this operation, as for instance in the case of v. Beck, where it was not present.

The most recent work on extirpation of the spleen for tumours is from Hartmann and Vasquez. As the result of their researches the authors arrive at the following conclusions:

1. A slight post-operative increase of the red blood corpuscles and a true acute hyperleucocytosis occur and pass rapidly away.

2. The hæmoglobin equivalent of the corpuscles sinks at first but recovers its original value by degrees.

3. 4-8 weeks afterwards a lymphocytosis of varying duration is established.

4. Later, after many months, a moderate eosinophilia occurs.

We have ourselves been able to investigate three conclusive cases.

Cases of splenectomy of this kind are transitional to the chronic diseases of the spleen. The latter present great difficulties, for one never knows how far in the most chronic diseases the other organs are damaged or influenced by the general illness.

An increase of the lymphocytes, so long as an affection of the lymphatic glands may be excluded, should be referred to functional exclusion of the spleen.

On the other hand, an increase of eosinophil cells associated with a chronic tumour of the spleen, is analogous to Kurloff's secondary reaction of the bone-marrow. Such cases are frequently found in the literature. For instance Müller and Rieder bring forward three cases of splenic tumour caused by congenital syphilis, cirrhosis of the liver, neoplasm in the cranial cavity, and in which the numbers of the eosinophils amounted to 12.3%, 7.0%, 6.5% respectively. In three cases of acute splenic tumour in typhoid fever the figure 0.31% with a maximum of 0.82%, was found. These authors have already raised the question "whether the increase of the eosinophil cells is connected with the splenic tumour or the bone-marrow? Perhaps the functional activity of the latter is vicariously raised to meet the more or less complete exclusion of the spleen from the formation of the blood; since Ehrlich has distinctly asserted that the probable place of formation of the eosinophil cells is the bone-marrow."

From what has been brought forward no doubt can now remain that the question has been decided quite in Ehrlich's favour.

But what then are the physiological functions of the spleen, since that organ is unnecessary for the persistence of life? Doubtless its chief duty is the taking up of the greater part of the decaying fragments of red and white blood corpuscles in the blood-stream, so that this valuable material is not quite lost for the organism. Thus Ponfick has found that after destruction of the red corpuscles the spleen takes up a portion of their "shadows," and for this reason calls the splenic tumour a spodogenous splenic tumour (σποδος, ruins). Ehrlich has made a corresponding observation for the products of dissolution of the white blood corpuscles, and has proved that the splenic tumour which occurs in many infectious diseases and in phosphorus poisoning is to a large extent caused by the parenchyma of the spleen taking up the remains of the neutrophil protoplasm.

The question of the relation of the spleen to the fresh formation of red blood corpuscles is a problem of comparative anatomy. Observations on this point made on one kind of animal can certainly not claim validity for other kinds. In lower vertebrates, as in fishes, frogs, tortoises, and also in birds, the blood-forming activity of the spleen is pronounced and of great importance. In mammalia on the other hand, in some cases this function cannot be demonstrated, and in others only to a very small degree. In the spleen of normal mice nucleated red blood corpuscles are seen in relatively large numbers; in the rabbit they are less numerous and often only to be found with difficulty. In the dog they only make their appearance after anæmia from loss of blood, normally they are absent. In the human spleen nucleated red blood corpuscles are not to be found normally or in cases of severe anæmia, but exclusively in leukæmic diseases. U. Gabbi in his recently published work on the hæmolytic function of the spleen, also emphasises the difference between the various animal species. In guinea-pigs he found that the spleen acts largely as a scavenger of the red blood corpuscles; in rabbits very slightly. Consequently after removal of the spleen in guinea-pigs the number of red blood corpuscles rose 377,000 in the cubic millimetre, and the amount of hæmoglobin 8.2%. After splenectomy in rabbits the increase in these values is absent.

Shortly summarising our analysis of the facts before us, we must say that the importance of the spleen for the production of the white blood corpuscles can in no respect be considerable, and that if these cells really are produced by it, they must be free from granulations. The spleen therefore stands functionally in closer connection with the lymphatic gland system than with the bone-marrow. The spleen has not the least connection with ordinary leucocytosis[14].

(β) The Lymphatic Glands.

As it is impossible experimentally to prevent the lymphatic glands as a whole from contributing to the formation of the blood, we are dependent almost entirely on clinical and anatomical researches for an elucidation of their function.

Since Virchow's definition of the lymphocyte it has been admitted that the lymphocytes of the blood, both the small and larger kinds, are identical with those of the lymphatic glands and the rest of the lymphatic system. This is proved by the complete agreement in general morphological character, in staining properties of the protoplasm and nucleus, and from the absence of granulation.

Abundant clinical experience testifies that the lymphocytes of the blood really do arise from the lymphatic system. Ehrlich had previously observed that when extensive portions of the lymphatic glandular system are put out of action by new growths and similar causes, the number of the lymphocytes may be considerably diminished. These observations have since that time been confirmed by various authors. For example, Reinbach describes several cases of malignant tumour, particularly sarcoma, in which the percentage of lymphocytes, which normally amounts to about 25%, was very considerably lowered; in one case of lymphosarcoma of the neck they only made up 0.6% of the total number. These conditions are quite easily and naturally explained by the exclusion of the lymphatic glands. It is difficult for the advocates of the view that the lymphocytes are the early stages of all white blood corpuscles to reconcile it with these facts. According to their scheme the low number of lymphocytes is to be explained in such cases by their unusually rapid transformation to the polynuclear elements—the old forms; or to appropriate the expression of Uskoff, by a too rapid ageing of the lymphocytes.

Further evidence for the origin of the lymphocytes of the blood from the lymphatic glands is to be obtained from those cases in which we find an increase of the lymphocytes in the blood. These "lymphocytoses" occur, in comparison with other leucocytoses, relatively seldom. Under certain conditions in which a hyperplasia of the lymphatic glandular apparatus makes its appearance, we often see at first an increase of the lymphocytes in the blood. Ehrlich and Karewski in some unpublished work have investigated together a large number of typical cases of lymphoma malignum, and were able constantly to observe a lymphocytosis, which in some cases was of high degree and bore almost a leukæmic character.

The lymphocytosis following extirpation of the spleen also belongs to this category, since a vicarious enlargement of the lymph glands is always to be observed in these cases.

On investigating the conditions under which in healthy individuals an increased number of lymphocytes enter the blood-stream, we have in the first place to notice the digestive canal, whose wall contains a thick layer of lymphatic tissue. According to the results of Rieder the proportion of the lymphocytes to polynuclears is practically normal in the leucocytosis of digestion, indeed the lymphocytes are rather in excess. The eosinophils on the other hand shew a marked relative reduction in this condition. The leucocytosis of digestion consequently differs essentially from the other kinds, in which the neutrophil elements are chiefly increased. The simultaneous increase of lymphocytes and polynuclears is doubtless brought about by a super-position of a raised income of lymphocytes, and an ordinary leucocytosis caused by the assimilated products of metabolism.

The influence of the digestive tract is still more evident in certain diseases, more particularly in intestinal diseases of infants. A considerable increase of the lymphocytes in the blood-stream is here to be observed. Thus Weiss found an important increase of the white blood corpuscles in simple catarrh of the stomach and intestines, which presented the main features of a lymphocytosis.

Whooping-cough, according to the recent observations of Meunier, also belongs to the small number of diseases which are accompanied by a pronounced lymphæmia. In the convulsive period of this disease both the polynuclear cells and the lymphocytes are increased, the latter in preponderating amount. The former cells are increased to twice, the lymph cells to four times their normal amount. Doubtless in these cases also the lymphocytosis is due to the stimulation and swelling of the tracheobronchial glands.

An increase of the lymphocytes from chemical stimuli is exceedingly rare, though, as is well known, a large number of substances (bacterial products, proteins, nucleins, organic extracts, and so forth) can call forth a polynuclear leucocytosis. In quite isolated cases, an increase of the lymphocytes in the blood in consequence of the injection of tuberculin into tuberculous individuals has been seen. (E. Grawitz.) From the rarity of these cases it can scarcely be doubted that here a tuberculous disease of the glands also plays a part, so that the increased immigration of lymphocytes is brought about not by a chemical property of the tuberculin but by the extensive specific reaction of the diseased glands.

Only one single substance has so far been mentioned in the literature as capable in itself of producing a lymphocytosis. Waldstein asserts that he has produced by injection of pilocarpine, a lymphæmia which undergoes a progressive increase with a rise in number of the injections.

The origin also of lymphocytosis is therefore sharply marked off from that of the ordinary leucocytosis, which consists in an increase of the neutrophil elements. Whilst the latter is admittedly the expression of chemiotactic action, and arises by action at a distance of soluble substances on the bone-marrow, lymphocytosis is due to a local stimulation of certain glandular areas. Thus in the leucocytosis of digestion, of intestinal diseases of children, we refer it to the excitation of the lymphatic apparatus of the intestine, in tuberculin lymphæmia we recognise mainly a reaction of the diseased lymph glands. Hence we conclude that a lymphocytosis appears when a raised lymph circulation occurs in a more or less extended area of lymphatic glands, and when, in consequence of the increased flow, more elements are mechanically washed out of the lymph glands. The pilocarpine lymphocytosis does not contradict this view, for pilocarpine causes extraordinary though transient variations in the distribution of water, whereby the inflow into the blood of fluid containing lymph cells is increased. We therefore regard lymphocytosis as the result of a mechanical process; whilst leucocytosis is the expression of an active chemiotactic reaction of the polynuclear elements.

This view finds its best support in the fact that the polynuclear leucocytes possess lively amœboid movement, which is completely wanting in the lymphocytes.

Corresponding to the absence of contractility in the lymphocytes it is also observed that in inflammatory processes in contradistinction to the polynuclear neutro-and oxyphils, the lymphocytes are not able to pass through the vessel wall. A very interesting experiment on this point was described by Neumann years ago. Neumann produced suppuration in a patient with lymphatic leukæmia, in whom the blood contained only a very small number of polynuclear cells. Investigation of the pus shewed that it consisted exclusively of polynuclear cells, and that not a single lymphocyte had come into the exudation, although this kind of cell was present so abundantly in the blood.

Histological examination of all fresh inflammatory processes, in which mainly polynuclear elements are found, leads to accordant results. It is well known that small-celled infiltration occurs in the later stage of inflammation, apparently consisting of lymph cells; nevertheless this does not in the least prove that these lymphocytes have emigrated here from the blood vessels. This is not the place to enter into the very extensive controversy on this point. We are content to refer to the most recent very thorough paper of Ribbert. Ribbert regards these foci of small-celled infiltration as the analogues of the lymphatic nodules, and explains their origin by an increase in size of the foci of lymphatic tissue, normally present, though in a condition but little developed.

It consequently follows from clinical and morphological researches, as well as from the observations on inflammatory processes, that the lymphocytes are in no way connected with the polynuclear leucocytes. We shall reach the same result in another way in the following section.

(γ) The Bone-marrow.

The spleen and lymphatic glands were at first regarded as the sole places of formation of the blood corpuscles. The almost simultaneous researches of Neumann and Bizzozero first attracted general attention to the importance of the bone-marrow. These authors showed that the early stages of the red blood corpuscles are produced there; a discovery which was quickly and generally recognised, and which soon became pathologically useful through the observations of Cohnheim and others. In this connection the observation was of great value, that after severe loss of blood the fatty marrow of the larger hollow bones again changes to red marrow, as it is evidence of the increased demands on the regenerative function of the bone-marrow.

We are unaware of a second place of formation of the red blood corpuscles in man. In other mammalia however, as we have above mentioned (see page 99), the spleen may also take a small share in the production of erythrocytes. The type which the normal blood formation follows in adults, and the deviations therefrom shewn in pernicious anæmia, have been described in the chapter on the red blood corpuscles. Ehrlich's view that the blood formation in pernicious anæmia belongs to a different type, which is analogous to the embyronic, was also described there.

In this section we have therefore to deal chiefly with the white blood corpuscles and their connection with the bone-marrow. In man as in a large number of animals (for example the monkey, guinea-pig, rabbit, pigeon and so forth) the bone-marrow exhibits the peculiarity that the cells it produces bear a specific granulation, in sharp contrast to the lymphatic glandular system, which contains elements free from granules, in the whole animal series.

The granulated cells of the bone-marrow fall into two groups.

The first group of the cells with "special granules" is very important since it constitutes a characteristic for certain species of animals. According to the class of animal they shew different tinctorial and morphological properties. Man and monkey for example have neutrophil granulation; guinea-pig and rabbit the pseudo-eosinophil granulation described by Kurloff; in birds we find two specific granulations present side by side, which both are oxyphil, and of which one is imbedded in the protoplasm in crystalline form, the other in the form of granules.

The kinds of special granulations so far investigated have the common property, that they stain in acid and neutral dyes respectively; they shew a much smaller affinity for the basic dyes. The fact that they greatly exceed the other elements of the bone-marrow in all classes of animals, is evidence of the importance of these granules.

The second group of bone-marrow cells contains granules which we find in the whole vertebrate series from the frog to man, and which therefore are not characteristic for any one species of animal. They are, (1) the eosinophil cells, (2) the basophil mast cells.

The bone-marrow forms which are free from granules consist mostly of mononuclear cells of different type. They are not nearly so numerous, or so important as the granulated kind, more especially as the first and predominant group.

Amongst the granule-free forms the giant cells deserve special mention, for they are an almost constant constituent of the bone-marrow of the mammalian class. According to the recent researches of Pugliese the giant cells are considerably increased after extirpation of the spleen in the hedgehog; an organ of quite extraordinary size in this animal and doubtless therefore possessing important hæmatopoietic functions.

Pugliese asserts that in the hedgehog after splenectomy the nucleated giant cells pass into leucocytes by amitotic nuclear division. Unfortunately in his preliminary communication there are no notes of the granules of the bone-marrow cells.

On examining a stained dry preparation of the bone-marrow of the guinea-pig, rabbit, man, etc. it is seen that the characteristic finely granular cells are present in all stages of development, from the mononuclear through the transitional to the polynuclear (polymorphously nucleated) forms, which we meet with in the circulating blood. A glance at a preparation of this kind shews that the bone-marrow is clearly the factory where typical polynuclear cells are continuously formed from the granule-containing mononuclears.

Here also the same process of ripening can be seen in the polynuclear eosinophil leucocytes.

Ehrlich has been able by special differential staining to bring forward proof that the constitution of the granulation changes during the metamorphosis of the mononuclear to the polynuclear cells. In the young granules there is prominent a basophil portion that becomes less and less marked as the cell grows older. The pseudo-eosinophil granules of the mononuclear cells, of the guinea-pig for example, stain bluish-red in eosine-methylene blue after long fixing in superheated steam: in the transitional stages this admixture is gradually lost, and finally completely vanishes in the granules of the polynuclear leucocytes which stain pure red. Analogous observations may be made in the eosinophil cells of man and animals, and in the neutrophils of man. Hence it is even possible to decide whether an isolated granule belonged to an old or to a young cell.

It is still impossible to judge with certainty the rate at which the ripening of the mononuclear to the polynuclear cells proceeds, or further to decide if the ripening of the granules always runs parallel in point of time with that of the whole cell. On the grounds of our observations we would suppose that in general the two processes run their course side by side, but that in special cases the morphological ripening of the cell may proceed more rapidly than that of the granules. It is particularly easy to observe this point in eosinophil cells. Ehrlich had already mentioned in his first paper (1878) that side by side with the typical eosinophil granules isolated granules are often found which shew a deviation in tinctorial properties: for instance, they stain more of a black colour in eosine-aurantia-nigrosin; in eosine-methylene-blue, bluish-red to pure blue. Ehrlich had already described these as young elements in his first paper. The same differences are found more sharply marked in leukæmia even in the circulating blood, in the neutrophil as well as in the eosinophil group. Ehrlich has repeatedly found in leukæmic blood polynuclear eosinophil cells, whose granules must almost exclusively be regarded as young forms[15].

Ehrlich regarded these as typical examples of a relative acceleration of the morphological ripening of the cells, as compared with the development of the granules.

In normal blood we find only the ripe forms of the specific granulated cells of the bone-marrow. The mononuclear and transitional forms of the neutrophil group, do not under normal circumstances pass over into the blood-stream.

Ehrlich regarded the mononuclear neutrophil granulated cells as characteristic for the bone-marrow, since they are found exclusively in the bone-marrow, never in the spleen or lymph glands, and for this reason named them "myelocytes," κατ' εξοχην[16]. When myelocytes, no matter of what size, appear in considerable numbers in the blood of an adult, a leukæmia of myelogenic nature is nearly always present. (For the very rare exceptions to this rule, which it may be added can never be confused with leukæmia, see pages 77, 78.)

Exactly similar conditions hold good for the eosinophil cells, in as much as the singly nucleated forms, which one may call eosinophil myelocytes, occur, almost exclusively, in leukæmic blood. These forms, which were first recognised by H. F. Müller, are however of less importance, for in myelogenic leukæmia the chief part of the foreign admixture of the blood is made up of Ehrlich's myelocytes.

Very important conclusions on the interesting question of leucocytosis can be drawn from these observations. Bearing in mind that polynuclear neutrophil cells are developed and stored up only in the bone-marrow, that in ordinary leucocytosis only the polynuclear forms are increased in the blood-stream, it is evident that leucocytosis is purely a function of the bone-marrow, as Ehrlich has always insisted with all distinctness. It is only on this assumption that the frequently sudden appearance of leucocytosis, as has so often been observed in morbid and experimental conditions, can be satisfactorily explained. In these cases the space of time, amounting often only to minutes, is far too short for a new formation of leucocytes to be conceivable; there must be places in which these cells are already completely formed, and able thence to emigrate on any suitable stimulus. This place is single, and is the bone-marrow alone. Here all mononuclear forms gradually ripen to the polynuclear contractile cells, which obey each chemiotactic stimulus by emigration, and which thus bring about sudden leucocytosis.

The bone-marrow thus fulfils, amongst others, the extremely important function of a protective organ, by which definite injurious influences which affect the organism may be quickly and energetically combated. Just as in a fire-station ample means of assistance is continuously in readiness immediately to answer an alarm from any quarter.

We wish to insist once more, that the large mononuclear leucocytes and the transitional forms of the normal blood are not concerned in the increase in ordinary leucocytosis; in leucocytosis of high degree their relative number may indeed be lowered, in consequence of the exclusive increase of the polynuclear cells. It appears then that these elements do not react to chemiotactic stimuli, and that possibly they reach the blood by entirely different ways than the polynuclears do.

No definite statement can as yet be made as to the places of formation of the non-granulated large mononuclear leucocytes.

Kurloff has demonstrated, that in the guinea-pig these cells are present both in the bone-marrow and in the spleen, but that after extirpation of the spleen the absolute number does not change. The bone-marrow then in the guinea-pig can also preserve the balance of the large mononuclear, non-granular cells in the blood.

The numbers we found in our blood investigations in man after splenectomy were also normal. We may then doubtless assume that the large mononuclear granuleless cells of human blood also arise for the most part from the bone-marrow. In this tissue they are to be picked out in the medley of the different kinds of cells only with the utmost difficulty, owing to their small number and their but little characteristic properties. Consequently an exact investigation of their origin could probably only be successful if it were possible experimentally to produce a disease in which these forms in particular underwent important increase. This advance is not quite hopeless, since in man at least an absolute increase of the large mononuclear cells is observed in the post-febrile stage of measles.

On the grounds merely of microscopical investigations we conclude that the bone-marrow is by far the most important of the blood-forming organs, for its function is the exclusive production of red blood discs as well as of the chief group of the white corpuscles, the polynuclear neutrophil.

The physiological, experimental investigation of the functions of the bone-marrow offers insurmountable difficulties. An exclusion of the whole bone-marrow or of larger portions only is an impossible operation. Nor can we ascribe any value to the researches which endeavour to obtain a result by comparative enumerations of the arterial and venous blood of a bone-marrow area. J. P. Roietzky working under Uskoff's direction has recently made counts of this kind in the dog, from the nutrient artery of the tibia and the corresponding vein. He found that the number of white corpuscles of the vein is slightly greater, that on the other hand the absolute number of "young corpuscles" (Uskoff), i.e. of the lymphocytes, has been considerably diminished, whilst the number of "ripe" corpuscles, which for the most part correspond to our polynuclear, is considerably increased. He gives the following table:

The argument based on figures such as these assumes that the function of the bone-marrow is continuous; an assumption which Uskoff indeed seems to make.

But if the bone-marrow is constantly absorbing the lymphocytes to such an extent, it is quite incomprehensible how the normal condition of the blood can be preserved, bearing in mind the extent of the bone-marrow and the rate of the circulation. All evidence indeed tends to shew that on the contrary the bone-marrow performs its functions discontinuously, inasmuch as elements continually grow to maturity in the bone-marrow, as we have above explained, but they only emigrate at certain times as the result of chemical stimuli. It is obvious à priori from this consideration how inconclusive must be the results of experiments such as these of Roietzky[17].

Far more important for the elucidation of the function of the bone-marrow are clinical observations on cases in which considerable portions of the bone-marrow are replaced by tissue of another kind. We may best divide the observations on this point into two groups: 1. malignant tumours of the bone-marrow, 2. the so-called acute leukæmia.

There are unfortunately very few available observations as yet upon the first group. Still rarer are the cases in which as is necessary the whole bone-marrow has been subjected to an exhaustive examination, which alone affords adequate evidence of the extent of the defect.

Amongst the changes of the bone-marrow arising from tumours one may distinguish two groups, according to the nature of the condition of the blood. The first type is exemplified by a case of Nothnagel published in his work on lymphadenia ossium. Here during life the blood shewed, in the main, the features of a simple severe anæmia; but in addition isolated normoblasts, small marrow cells, and moderate leucocytosis. The autopsy, at which the whole skeletal system was subjected systematically to an exact examination, shewed a complete atrophy of the bone-marrow, and replacement of the same by the tumour masses. In this case then the condition of the blood in vivo is satisfactorily explained by the absence of function of bone-marrow. Nothnagel conjectured that the formation of the scanty nucleated red blood corpuscles occurred vicariously in the spleen, that of the leucocytes in the lymph glands.

In the second series to which the cases of Israel and Leyden, as well as the recently published one of J. Epstein from Neusser's wards, belong, the blood shews, besides the usual anæmic changes, other anomalies which are peculiar partly to pernicious anæmia, partly to myelogenic leukæmia. In Epstein's case of metastatic carcinoma of the bone-marrow, there was found a considerable anæmia, with numerous nucleated red blood corpuscles both of the normo- and megaloblastic type; their nuclei presented the strangest shapes, due not merely to typical nuclear division, but also to nuclear degeneration. The white blood corpuscles were much increased, their proportion to the red was 1/25 to 1/40; the increase concerned in the main the large mononuclear forms, which bore for the most part neutrophil granulation, and were therefore to be called myelocytes. In all the specimens, only two eosinophil cells were found[18].

The explanation of a blood picture of this kind, apart from the purely anæmic changes, is by no means easy, as Epstein rightly observes. The appearance of myelocytes is most readily explained by a direct stimulation of the remaining bone-marrow by the surrounding masses of tumour. In this, the mechanical factor is less concerned than the chemical metabolic products of the tumour masses; which at first act on the adjacent tissue in specially strong concentration, and also in a negatively chemiotactic manner on the wandering cells. This view receives support from the careful work of Reinbach on the behaviour of the leucocytes in malignant tumours. Out of 40 cases examined, in only one, of lymphosarcoma complicated with tuberculosis, were myelocytes found in the blood, amounting to about 0.5-1.0% of the white blood corpuscles. The autopsy shewed isolated yellowish white foci of growth in the bone-marrow, reaching the size of a sixpenny piece. Bearing in mind that in none of the other 39 cases were myelocytes demonstrated, one does not hesitate to explain their presence in the blood in this single case by the metastases in the bone-marrow. The small extent of the latter is likewise the cause of the small percentage of the myelocytes.

In explaining the presence of the megaloblasts in the blood of Epstein's patient we must keep before us what we have said elsewhere on this kind of cell. They are not present in the normal bone-marrow; they arise on the contrary, according to our view, when a specific morbid agent acts upon the bone-marrow, as we must assume is the case in the pernicious forms of anæmia. In the cases of anæmia from tumours, in which we find megaloblasts in large numbers in the blood, we must likewise assume that chemical stimuli proceed from the tumours, leading to the formation of megaloblasts in the bone-marrow.

The presence of megaloblasts in the bone-marrow does not itself cause their appearance in the blood, for in pernicious anæmia the bone-marrow may be filled with megaloblasts, and yet only very scanty examples are to be found in the blood. Whether the emigration of the megaloblasts from the bone-marrow into the blood-stream is in general to be referred to chemical stimuli, as it is in the particular case of Epstein's, or to mechanical causes, cannot at present be decided.

The bone-marrow may be replaced by typical lymphatic tissue, as well as by the substance of malignant tumours. The former occurs constantly in lymphatic leukæmia according to the well-known results of Neumann, which have since been generally confirmed. In these cases extensive tracts of bone-marrow are replaced not by masses of malignant growth but by an indifferent tissue, so to speak, a tissue which is unable to exercise the above-described stimulating influence upon the remaining bone-marrow. It is owing to this circumstance that we are able to observe in the cases of lymphatic degeneration of the bone-marrow the phenomena due to its exclusion, in their most uncomplicated form[19].

The most convincing results are obtained from cases of acute (lymphatic) leukæmia, the pretty frequent occurrence of which was first noticed by Epstein, and which has lately been very thoroughly studied by A. Fränkel. For the purpose in question, acute leukæmia is specially suited, since the abnormal growth of the lymphatic tissue takes place very rapidly, and for this reason brings about a quick and uncomplicated exclusion of the bone-marrow tissue; as it were, experimentally. Under its influence the neutrophil elements of the bone-marrow vanish rapidly, and in many cases so completely that it needs some trouble to find a single myelocyte, as for example in a case of Ehrlich's. The polynuclear leucocytes are produced in the bone-marrow, consequently where the bone-marrow is destroyed, as in this case, it is clear that their numbers must be absolutely very much diminished in the blood.

Dock has also arrived at similar results, as we see from a preliminary report; and he similarly explains the absence of neutrophil cells in lymphatic leukæmia by the replacement of the myeloid by lymphatic tissue.

Thus lymphatic leukæmia affords a striking proof that the lymphocytes are cells of a peculiar kind, and which are quite independent of the polynuclear cells. It is therefore exceedingly surprising that Fränkel, after accurately examining and analysing eight cases of acute lymphatic leukæmia, believes he has found in them imperative reasons for the assumption that the lymphocytes are transformed to polynuclear cells. This can only be explained by the confusion which Uskoff's doctrine of "young cells" has brought about.

We define lymphocytosis as an increase of the lymphocytes of the blood; Fränkel like Uskoff regards it as the emigration of the young forms of the white blood corpuscles into the blood. He concludes logically from the diminution of the polynuclear cells in this form of disease "that the conditions of the transformation of the young forms have undergone a disturbance." But if one assumes that the lymphocytes are young forms, and the polynuclears their older stages, it is much nearer to the facts to speak, not of a disturbance in lymphatic leukæmia, but of an absolute hinderance to the ripening process. It is easy to conceive any particular stimulus or injury bringing about an acceleration of the normal process, that is, a premature old age, but it is equally difficult to represent clearly to oneself conditions which retard or completely prevent the normal ageing of the elements. The discovery of such conditions would be really epoch-making, both for general biology, and for therapeutics. The only escape from this dilemma would be the assumption of a very premature death of the lymphocytes, for which however not the smallest evidence is to be found, even in Fränkel's monograph. Fränkel distinguishes the acute from the chronic forms of leukæmia by the fact, "that in the former the newly formed elements emigrate from their places of formation into the blood-stream with extraordinary rapidity. Hence there is not time for further local metamorphosis. In chronic leukæmia the emigration takes place very probably much more slowly." This distinction is contradicted by the facts; for there are chronic forms of lymphatic leukæmia whose microscopic picture is identical with that of acute leukæmia. And hence the starting-point of all Fränkel's deductions is rendered insecure.