From the foregoing description of the histological changes in the leptomeninx it is quite evident that we are dealing with a chronic, stationary, healing form of tuberculous inflammation. This statement is substantiated, in the first place, by the clinical history. The only reasonable interpretation of the symptoms would establish the duration of the process as four months. The imaginable contingency that there existed first a meningeal syphilitic lesion that was dispersed by the iodide of potassium only to be followed by a tuberculous infection is so remote and unlikely that it need not be discussed. At all events the tuberculous leptomeningitis, which presented a typical distribution, began insidiously, existed at times in a latent condition, and pursued a very anomalous course, marked by a relative mildness of all the symptoms, and thus it came about that when an apparent or real improvement followed the administration of iodide of potassium able observers were induced to make an erroneous diagnosis. Death occurred as a result of an intercurrent infection. The long duration of the process is also shown, anatomically, by the thick layer of firm, translucent and gelatinous material that matted together the structures at the base, and also by the evident adhesions between the pia and the brain. The histological examination furnishes proof positive of the correctness of the conclusion in regard to the peculiar character of this process because it shows: (1) That the tuberculous proliferation is uniform in development and has reached nearly the same stage of evolution throughout the entire extent of the leptomeninx involved; it is not a process that has advanced by exacerbations and irregular extensions; the lesions are, generally speaking, of nearly the same age everywhere and must have begun at about the same time. (2) That only a very limited degree of caseous degeneration is present, pointing to an early arrest of the activity of the tubercle bacillus or to a very decided diminution or attenuation of its virulence. (3) That the subendothelial intimal proliferations of epithelioid cells, so generally found in acute tuberculous leptomeningitis,* have in this case become more or less completely changed into distinct fibrous tissue in which but very slight, if any, direct evidence of its tuberculous origin can be found. It is only by recognizing that the chronic endarteritis is most marked in correspondence with the most advanced adventitial tuberculous changes, and by finding an imperfect, much altered giant cell in one district of intimal thickening, that we were able to establish the direct kinship of the endovascular changes with those of the pia in general. (4) That acute inflammatory changes, in the form of emigration of polymorphonuclear leucocytes and of fibrinous exudation, are entirely absent in all parts of the district involved. The presence of a turbid serous fluid is of course not at all inconsistent with the view that the anatomical changes are of long duration. (5) That the granulation tissue present is, in general, undergoing fibrillation and contains a rich supply of enabryonal capillary vessels as well as of larger blood-vessels of evidently new formation. The absence of any considerable extent of polymorphonuclear leucocytic infiltration in this tissue has already been referred to. The cells in the granulation tissue correspond to the cells of embryonal or formative connective tissue. Vacuolation is rarely present. (6) That the unusually large number of giant cells present are remarkably free from evidences of necrosis and degeneration of the character ordinarily observed in tuberculous proliferations, that they do not contain in demonstrable form tubercle bacilli, and that the majority of the giant cells seem to be separating into individual cells and smaller masses often with, but sometimes also without, evidences of nuclear disintegration. The possibility that these phenomena may signify fusion instead of the sundering of cells will be discussed below. For these reasons there can be no doubt that the general claim that we are dealing with an instance of chronic, healing tuberculous meningitis must be regarded as established beyond dispute. The growth of tubercle bacilli in the glycerine-agar tubes, inoculated with the fluid from the pial meshes, and the demonstration of tubercle bacilli, though in very small numbers, between the cells of the embryonal tissue, furnish the positive evidence that we are actually dealing with a tuberculous process due to living and not to dead bacilli. The degree of virulence of the cultures of tubercle bacilli was, unfortunately perhaps, not studied. The presence of living tubercle bacilli in a tissue free from active and acute changes characteristic of tuberculosis demonstrates that, whatever the actual degree of virulence of the bacilli may have been, the tissue in which they were found was at this time relatively immune from their action. The manner in which this immunity was produced, and in which the process of healing was initiated, need not be discussed at this time any further than to again direct attention to the fact that the bacilli lost their virulency as regards the cells in this leptomeninx before these cells underwent any marked degree of degeneration. The cells of the tuberculous proliferations survived the further action of the bacilli whose original effect it was to initiate cell accumulation or proliferation; the cells also retained sufficient vitality to develop, in some instances at any rate, into formative cells according as their origin would dictate, e. g. into fibroblasts. That fibroblasts are formed only by embryonal connective tissue cells, and not by wandering cells, such as the large mononuclear leucocytes, we are well aware, is possibly still a disputable assumption, and we do not consider it pertinent to discuss the question any further in connection with this study, but would only emphasize the point that some of the cells of tuberculous proliferations may, under favorable circumstances, become formative cells, and, furthermore, that the amount of formative tissue produced may be far in excess of what is actually needed for purposes of repair only. Surely the appearances here noted indicate that the bacillus of tuberculosis has the power to stimulate fixed cells to multiply, unless one assumes that all, or almost all, the formative cells here seen are derived from wandering cells attracted by the presence of the bacillus and its products. As to the ultimate fate of the formative and other cells in this healing tuberculous tissue no final statements can be made. It must be remembered that it is only one stage in the process of healing that is dealt with. The well marked evidences of fibrillation, the quite extensive formation of new vessels, the absence of evidences of degenerative changes in the uninuclear cells, all point to the production of new fibrous tissue as sure to occur, but it seems quite probable that occasional epithelioid cells may undergo or have undergone dropsical or other forms of degeneration, although it is certainly apparent that so far as the small cells are concerned the involution of the tuberculous tissue is not occurring through disintegration. Perhaps the most interesting feature in this case is the opportunity it affords to study the changes in the giant cells of healing, non-degenerated tuberculous tissue. In the first place, the large number of giant cells is quite remarkable. The general characters of the tissue in which they are found recall the fact that giant cells are regarded as quite constant elements in chronic mild tuberculosis; often the giant cells are the only cells that contain bacilli (Koch). In this instance the giant cells do not contain bacilli that are demonstrable by the usual methods; neither do they contain bodies that can be definitely interpreted as degenerate forms of bacilli such as those found by Metchnikoff, Stchastny, Weicker, and others, in the giant cells of Spermophilus guttatus, in avian and in human tuberculosis. Metchnikoff states, however, that he knows of the occurrence of such degenerate forms only in the Spermophilus guttatus under the circumstances mentioned, and in the rabbit and guinea-pig in mammalian tuberculosis, but not in man; consequently, the manner in which the giant cells rid themselves of the bacilli undoubtedly present in their interior at some time during their existence, must as yet remain without any explanation. In the description of the histological changes the various appearances presented by the giant cells are described somewhat minutely. The essential observations made concern, in my opinion, the further fate of giant cells which are still found to persist in healing nondegenerated tuberculous tissue. It was, I believe, quite conclusively shown that the consecutive changes appear to consist in the breaking up of the nuclei, the removal of the detritus by phagocytes, and the formation of a few apparently viable uninuclear cells in the case of more degenerated, exhausted giant cells, while other, and, as it would seem, better preserved or younger giant cells, separate into a number of individual, uninuclear cells with but little or no nuclear disintegration. Objection might be raised to this interpretation of the appearances in the giant cells. While no one could very well dispute the view that part of the giant cells are undergoing retrogressive and absorptive changes with the production of some viable cells, a question might well be raised concerning the nature of the process taking place in those giant cells that have been spoken of as splitting up or dividing into uninuclear cells and smaller multinucleated masses without much evidence of nuclear disintegration. It might be claimed that the process is one of fusion of many cells to form giant cells, and not one of division of fully formed giant cells into small cells. But a broad view of the processes described speaks against fusion. In the first place we are not dealing with a stage of tuberculous proliferation (Baumgarten), or cell accumulation (Metchnikoff), in which one would look for the production of giant cells, no matter which view concerning the histogenesis of tubercle be assumed as the correct one, because it has been demonstrated that, from whichever point of view the lesions are examined, the same positive conclusion that they are in the process of healing is reached; there is, therefore, no occasion for the formation of new giant cells in such wide-spread degree throughout the district involved. It might he claimed that the cells became arrested and, as it were, fixed in the act of fusion which was taking place in the early stage of the meningitis, but it would be difficult to understand the nature of the stimulus that could hold the cells together in such a peculiar manner for such a long time. It must be remembered that bacilli or bacillary detritus could not be found among the incomplete or in the complete giant cells. In the second place the difference between the cells that are undergoing disintegration and those regarded as dividing is essentially, to a certain extent at any rate, one of degree, because in the first instance there is not much, if any, doubt but that viable smaller cells are also formed, and in the second instance some, though often very slight, evidence of nuclear fragmentation is nearly always present; it would also be correct to infer that in advanced subdivision of a giant cell much, and perhaps all, of the nuclear detritus produced might have been removed up to the last trace; finally, the two extremes of these changes in the giant cells are connected by transition stages passing by gradation from the one to the other. Hence it is justifiable to conclude, for the time being, that in healing non-degenerated tuberculous tissue, the multinucleated giant cells may in part disintegrate and undergo absorption, in part form viable small cells; that both these changes may, and usually do, affect the same cell, but that in one class of cells—presumably the older or the more exhausted—the retrogressive process is predominant, while in a second class of cells—presumably the young and vigorous—the progressive changes are the more marked. In this connection it may be pointed out that while there cannot very well be any question but that we are dealing only with dividing and not coalescing cells, yet if this conclusion should be disputed and found incorrect, then the only remaining alternative would be to infer that this tissue furnished a unique and striking example of the formation of plasmodial masses by fusion in human tuberculosis, a conclusion to which many pathologists would refuse to subscribe, if for no other reason than because it is not in accordance with the almost universally accepted teachings of Baumgarten and Weigert in regard to the mode of formation of the giant cells in tuberculosis. Believing as I do that the giant cells under consideration are in the act of division and not at all of fusion, there remain to be discussed some of the histological and other features presented by the dividing cells. Many of the giant cells, perhaps the majority, contain larger and smaller vacuoles in the protoplasm. The exact significance of this vacuolation is not always clear. When the vacuolation accompanies an evident solution of the nucleus (karyolysis), there cannot be any doubt but that we are in the presence of a distinctly retrogressive process. Vacuoles are also most numerous in the giant cells that present other evidences of degeneration, such as coarseness of the granules in the protoplasm and extensive nuclear disintegration, but they occur as well around nuclei that stain deeply, around cells that seem to be separating from the giant cell, and even about nuclei that present mitoses. The formation of vacuoles seems to be responsible, to a certain extent at any rate, for the diminution in the volume of disintegrating and dividing giant cells, as shown by the clear spaces that form about them; these spaces are too large and occur too uniformly to be attributed solely to artificial shrinking produced by the hardening in alcohol. Further undoubted evidence of retrogression in certain giant cells is the occurrence of nuclear disintegration, or karyorhexis, which sets free larger and smaller chromatin masses that are recognized in the giant cell as well as in the interior of the phagocytes usually found around such cells. Almost all the polymorphonuclear leucocytes found in this tissue are met with around giant cells with broken-up nuclei. In many nuclei of disintegrating giant cells can be noted appearances that correspond well to certain stages in the complicated karyorhexis observed in anæmic necrosis by Schmaus and Albrecht; some of the nuclei with budding processes correspond particularly well with those in certain of their drawings; the interior of giant cells of tuberculous tissue may, it would seem, present conditions favorable to the development of this series of postnecrotic nuclear change. Vacuolation, karyolysis and karyorhexis are the essential steps that lead to destruction of the whole or parts of some of the giant cells; associated with these processes there is usually observed a splitting up of the body of the giant cell into irregular fragments with as well as without nuclei; and, as described, more or less phagocytosis of the resulting remnants of various kinds is seen. But evident degenerative and necrotic processes in a giant cell may be associated with progressive changes. While some nuclei undergo vacuolation or break up, others seem to become richer in chromatin and to stain more deeply at the same time that they seem to acquire cell bodies quite distinct from the protoplasm of the giant cells: this hyperchromatosis does not, therefore, seem to be a stage in karyorhexis. A very few but undoubted karyokinetic figures were found, together with evidences of division of the cell body formed in the giant cell protoplasm. Precisely similar changes are described by Klebs in healing pulmonary tuberculosis of the guinea-pig; the nuclei of the giant cells became rich in chromatin and karyokinetic figures occurred. Krückmann among others has found occasional mitoses in giant cells around foreign bodies, as well as elsewhere, but it would seem that such mitoses have always been interpreted as indicating the probable mode of formation of the giant cells rather than of their involution. The question of mitosis in existing multinucleated cells has recently been studied by Krompecher, who concludes that the individual nuclei of such cells may undoubtedly divide by mitosis, either simultaneously or at separate times. Division by amitosis can also occur, but mitosis is the only progressive form of division, amitosis being a retrogressive, disintegrating process that must be looked upon as an evidence of degeneration of the nucleus. Ziegler states that in division of giant cells whose nuclei have multiplied by mitosis it may happen that the separating cell remains enclosed in the protoplasm of the mother cell. A singular phase in the involution of the giant cells in this pia is to be found in the existence of progressive changes side by side with nuclear necrosis and with degeneration; this finding indicates that giant cells may contain many independent elements which, though apparently fused into one large cell, may preserve their individuality so that while some nuclei die, others proliferate and perhaps feed on the remnants of their dead brethren and form new, viable small cells. The nuclei in giant cells may be looked upon as representing independent centres, capable at times of existing even though the cell protoplasm is disintegrated. Many of the giant cells separate into individual cells, unaccompanied or unassociated with much evidence of necrosis. These cells may be regarded as the more vigorous forms. Here also are observed occasional mitoses—but on the whole extremely few—and very constantly an evident increase in the amount of chromatin in the nuclei of the new cells as compared with the amount ordinarily found in the nuclei of giant cells. These deductions concerning the persistence of the vitality of some of the nuclei, even in the presence of molecular and morphological changes in the cytoplasm and in other nuclei of the giant cell that lead to disintegration, are not entirely without the support of previous observations on cells, which, although made under different conditions, are nevertheless, it would seem, applicable to cells in general. Thus the brilliant investigations of Loeb upon the effects of various unfavorable surroundings, such as absence of oxygen or reduction of the amount of water, upon the cleavage of eggs of many kinds, show that the conditions which arrest development are qualitatively alike for nucleus and protoplasm, but quantitatively less for the protoplasm; when the irritability of the protoplasm is suspended the nucleus may segment without segmentation of the protoplasm, but upon re-establishment of favorable conditions the protoplasm may divide into about as many spheres as there are nuclei preformed—the nucleus persists, preserves the irritability of the cell and stimulates the protoplasm to segmentation. From the appearances of the giant cells here described it would seem, then, that some nuclei are able to maintain their vitality longer than others in the same cell, and under certain conditions to stimulate parts of the protoplasm to segment; in other cells all the nuclei have, as a rule, preserved their irritability. The groups of cells formed by the dividing of the giant cells can be traced by studying the process at the different stages in the different parts of the tissue. They assume an oval or spindle-shaped form, becoming more and more like the formative and endothelioid cells of young connective tissue, but their ultimate fate cannot be determined because it concerns essentially only one limited period in the involution of the tissue. It may be said with reasonable certainty, however, that the new cells do not form blood-vessels, but as regards their forming lymph-vessels nothing definite can be concluded. It would not be safe to draw any definite conclusions, from the appearances described, with regard to the origin and the mode of formation of the giant cells. The resulting small cells in general resemble very much endothelial and formative cells, but some of them are, at certain stages at any rate, not unlike large mononuclear leucocytes; their final fully developed or mature condition being unknown, no positive inference can be drawn as to their pre-giant-cell origin. The evidence points to the fact that the most probable origin of the giant cells, as indicated by their form and the apparent future career of their descendants, would be the fixed mesoblastic cells of the pia. In regard to the mode of formation of the giant cells it is quite clear that it must involve some process which is not incompatible with the viability of the small cells which may spring from the giant cells. Whether this would speak more in favor of formation by fusion than by karyokinesis of a single cell without division of the cell body cannot be well determined, and as long as authors are not agreed upon the question of the production of living, procreative cells by amitosis (direct segmentation, direct and indirect fragmentation) it would not be profitable to discuss the compatibility or incompatibility of the views of those investigators who trace the origin of giant cells to amitotic division, with the progressive changes that giant cells have been shown to be capable of. The fact that giant cells in tuberculous tissue, under certain conditions, undergo progressive changes and separate into small, living cells proves that they are not, as claimed by Baumgarten, Weigert and others, necrobiotic elements that are doomed to destruction from their very inception. On the other hand it lends more strength, if that were necessary, to the teleological view urged by Metchnikoff that they are living, defensive cells (whatever their origin may be), formed for the distinct purpose, like plasmodial masses in general, of isolating and removing foreign, harmful bodies, in this case the tubercle bacillus, and, having accomplished their object without being destroyed or exhausted, or the cause of their formation being removed or neutralized in some way, they, or their nuclei, may retain enough irritability to form a larger or smaller number of living, small, uninuclear cells.