leucocytopenia and the lymphorrhoea occur together; but he does not find the apoplasmia of the blood noted by Heidenhain. In my experiments throughout the leucocytopenic phase of the blood the apoplasmia of it was obvious and steadily progressive. I have not yet systematically examined the flow of lymph from the duct in my form of experiment, but I have noted that the flow is sometimes increased, and that the lymph may contain hæmoglobin in solution. The absence of obvious increase, indeed the estimated decrease, of the ratio of hyaline leucocytes to chromocytes, as well as of hyaline leucocytes to granular leucocytes, indicates that any increased addition of hyaline hæmic leucocytes that may occur, vid the thoracic duct, is out-balanced during the leucocytopenic phase by increased conversion of hyaline leucocytes into granular. There is, I take it, little doubt that many of the granular leucocytes are developed from the hyaline form (Kölliker, Virchow, &c.); and an increased rate of development si probable at the beginning of many forms of leucocytosis (Römer's* "formativer Reiz.")

In my experiments I have not found the leucocytopenia bear any very constant relation to the fever as judged by body temperature and respiratory rate. It may bear a closer relation to the process of apoplasmia; it seems to be slight when the apoplasmia is only slight.

Second.-The Leucocytotic Phase.

The time of onset of leucocytosis varied. It became obvious in something later than three-quarters of an hour after the establishment of the local lesion. In the ligation experiments it was particularly late. I did not detect any very constant relation between it and body-temperature or respiratory rate.

Its duration varied. It may be prolonged for several days and without much abatement.

The anatomical details of inflammatory leucocytosis have been recently reviewed and studied by Rieder. I will merely point out that in my observations as in his the increase in the total hæmic leucocytes has been accompanied by upset of the normal numerical ratio of granular to hyaline (in his observations polynuclear to mononuclear) in favour of the granular leucocytes. Rieder saw the proportion rise sometimes to 20: 1. I have seen it rise from 6.2 : 1 to 19.4 1.

In this feature again there is a resemblance between this form of leucocytosis and that ensuing upon injection of albumoses, bacterial cultures, &c., into the circulation. Hankin and Kanthack‡ have

*Virchow's Archiv,' 1892.
+ Op. cit.

Op. cit.

pointed out that the leucocytosis after bacterial injections is chiefly due to increase of the granular leucocytes; Löwit* has since shown the same thing for the leucocytosis following injections of albumose, nuclein, &c. In one of his experiments the polynuclear were to mono. nuclear leucocytes as 87 to 13. V. Limbeck,† and Everard, Demoor, and Massart‡ note the same feature in the leucocytosis resulting from subcutaneous injections of bacterial cultures.

As to interpreting the meaning of the leucocytosis, Römer§ has asserted that the increase of leucocytes is due to rapid multiplication of the leucocytes in the blood, especially in the blood of the veins. Kanthack|| was, however, unable to confirm Römer's statement that the venous blood showed greater leucocytosis than the arterial; and Löwit¶ has recently pointed out well-founded objections to Römer's observations. Löwit considers the leucocytosis due to increased supply of young leucocytes to the blood, these developing into the polynuclear form. Löwit supposes that the excessive production of leucocytes following their diminution (in his view their dissolution) is due to chemical stimulation of leucocyte-forming organs (lymph-glands, &c.) by substances shed into the blood plasma at the time of disintegration of the hæmic leucocytes. He considers this explanation applicable to all forms of leucocytosis, and from it argues the probability that "inflammatory leucocytosis" will be found to be preceded by a diminution of hæmic leucocytes. My experiments bear him out in that point; but, as to basing inflammatory leucocytosis on a previous dissolution of hæmic leucocytes my observations lend no help, and are capable of interpretation in other ways.

I think in my experiments the degree of leucocytopenia has not always been similarly proportioned to the succeeding leucocytosis. Since reproductive division of leucocytes, inclusive of the granular, polynuclear, or adult form, has now been shown to occur in the blood, the finely granular leucocytes may, therefore, increase in number by reproduction within the circulation.

Third.-Behaviour of the coarsely-granular Leucocyte.

A striking and I believe hitherto unrecorded feature of the change in the leucocytic elements of the blood relates to Wharton Jones's "coarsely-granular cell."

This leucocyte, like the other granular leucocytes, suffers numerical * Op. cit.

+ Op. cit. Op. cit.

§ Op. cit.

Acute leucocytosis produced by bacterial products," 'Brit. Med. Journ.,' June 18, 1892, p. 1301.

Op. cit.

reduction in the leucocytopenic phase, indeed it would appear to undergo even greater numerical decrease than the ordinary granular leucocyte with fine granules. But when the leucocytopenic phase passes off and the total number of granular leucocytes in the blood becomes greatly increased, their increase is due entirely to the "finely granular" cell, and there is no accompanying increase of the coarsely granular" cell. On the contrary the number of the latter becomes still fewer, not merely in comparison with the rest of the polymorphic, but also in proportion to the number of chromocytes and, more striking still, absolutely as measured per unit volume of the steadily concentrating blood.

66

The diminution proceeds to such a degree that usually after the seventh hour from the time of establishment of the local lesion this species of leucocyte has been in my experiments only with some diffi. culty demonstrable in the blood. It disappears from the small samples used in the Thoma-Zeiss counters, and often from the larger samples employed in fresh and dried films spread on § in. circular cover-slips. If the blood be examined by drawing 3 c.c. into a test-tube, oxalating, centrifuging, and making films from the separated layer of leucocytes, examples can usually be then found without very prolonged search, but at the ninth hour and later I have several times failed to find any, even by prolonged search, after separation by the centrifuge.

It seems, therefore, that in consequence of severe local inflammation the "coarsely-granular" hæmic leucocyte of Wharton Jones may practically disappear from the general circulation, although at the same time his "finely granular " hæmic leucocyte may be enormously increased in number.

As to the significance of this disappearance, two possibilities arise at once for consideration.

1. The cell may have become caught and so to say hidden in some part or parts of the vascular system, or have wandered out of the blood circulation altogether.

2. The cell may have become so altered in appearance that it is not recognisable by the criteria I adopt for distinguishing it. Or its resistance to the procedures employed in looking for it may have become so lowered that it is destroyed in the process of search.

To take the last hypothesis first. I consider the coarsely granular hæmic leucocyte of the mammalian blood I have examined, to be normally one of the most resistant and easily preserved cells in the body. It can be frozen and thawed again, irrigated with 33 per cent. alcohol, with saturated ammonium molybdate solution, with strong aqueous solution of pyrogallic acid, with ammonium sulphide solution, with 0.3 per cent. acetic acid; it can be heated after partial drying to 120° C., and higher, and yet its granulation not be lost or characters distinctive of them destroyed, or the cell itself altered

beyond recognition. Now when the cell is discoverable only in scanty number in the blood, those individuals still present offer the same strikingly resistant quality as do the similar cells obtainable from normal blood. Yet in order on the above view to explain its disappearance from the blood, we have to suppose that in the simple process of the shedding of a drop of blood direct upon a clean coverglass, and the spreading of the drop into a film under the action of capillarity when the cover-glass is dropped upon the glass slide, the cell forthwith disappears to leave behind no recognised trace. This seems improbable, the more so as the search for the cell was in my case often commenced less than twenty seconds after it left the circulation.

The supposition that the appearance of the cell is altered, and that for that reason the cell, although still circulating, escapes recognition, merits more consideration.

Ehrlich was, I believe, the first to suggest that the coarsely granular leucocyte is a unicellular gland. From this suggestion one passes by a step, which is easy in the light of Heidenhain's and Langley's discoveries of phases of granularity in secreting cells, to the supposition that the granules of the coarsely granular leucocyte may disappear from it at certain periods of its activity. Recently Kanthack and Hardy* have most importantly extended some suggestive work by Hankin, and have proved that under bacterial irritation the coarsely granular leucocyte of the frog does actually change its appearance, and that the granules disappear from it more or less completely, while the surrounding microbes suffer damage.

Although bacteria were certainly never in the blood in my experi ments it is impossible to suppose that this coarsely granular leucocyte discharges itself of granules under irritation of a bacterial kind only. The local inflammation in my experiments it is conceivable adds substances to the blood which may irritate these cells in the same way. If so it might be expected that in the frequent and prolonged examinations of these cells in living films, in hanging drops, &c, some good and indubitable evidence of the phenomenon should have met me. In the blood, say at the fifth hour, when the numerical disappearance of the cell is already advanced, there might have been expected, unless the disappearance of the granules is almost momentary, a certain number of cells in which the removal of the granules is incomplete. I have never observed any such appearance. In this connection I have especially borne in mind the occurrence of the subvarieties of the cell already mentioned. It is obvious that these subvarieties may be merely phases of development or degradation of charging or discharging of the cell; one of them certainly seems an

*Roy. Soc. Proc.,' December, 1892.

early phase of development. But I have never satisfied myself that the individuals of a subvariety became more numerous or less numerous in proportion to the typical cell. All of these forms appeared to be removed together.

Nor do I believe that the cell without its granules would have in most cases escaped recognition and passed muster with the other leucocytes. As stated above, the granules are far from being the only distinctive feature of the cell.

It is quite certain that, when the cells are being or have been reduced to a minimum in the blood, the remaining individuals are usually of perfectly normal granular appearance, active and unimpaired in their amoeboid action. I have, in the stage just previous to the disappearance of the cell from ordinary films, made a dozen such films and left them, at various temperatures, protected from evaporation, and then examined them at different intervals of time to see if they did lose their granulation.

Many hæmic leucocytes, under certain circumstances, live a long time after removal from the body (frog's leucocytes nine weeks in vitro,* dog's leucocytes three weeks, kept cool, in vitro†). It might therefore have been expected that, if blood containing leucocytes which had freed themselves of their secretion granules were removed and kept alive in vitro for a time, the granules might be reformed. I placed samples of blood in which I was unable to find any leucocytes containing coarse granules in vitro under appropriate conditions, and examined it at repeated intervals; but I did not succeed in obtaining any reappearance of the coarsely granular cell. A great number of the leucocytes do, and before very long, become granular with fine and medium sized granules, and these granules are highly refractive and not basophil (i.e., not micrococci). I have not satisfied myself that they are fatty. The mode of their appearance seems to be as

follows:

A great number of the finely granular leucocytes gradually become, as above mentioned, vacuolated. The vacuoles contain fluid, and are for the most part small and spherical, but some are larger and oval in shape. In most of the small vacuoles a single, fine, highly refracting particle, the size of a small micrococcus, is to be found, dancing under Brownian movement. In many of the vacuoles no particle at all is discoverable. In the larger vacuoles are obvious ingesta, but the bright particles in the small vacuoles do not seem to be ingesta because the plasma in which the cells lie is generally quite free from particles. The cells seem to exhibit the same vacuolation whether they have remained spherical and inactive as regards amoeboid move* 'Zahn; cf. V. Kahlden, 'Ber. u. d. Verhandlungen der Path. Anat.,' Sect. a. d. 10 Intern. Medic. Congress, Berlin, 1890.

+ Sherrington, 'Intern. Congress of Physiol.,' Liège, 1892.