Explanation of Plate V (figs. 1-22):

'Figures 1, 2, 3, 4, 5, 6, 7, 8, 9, 11, 14, 16, 17, 18, 19, 21, and 22 are drawn with camera lucida and Ocular 4, Zeiss Objective 1/12th, Leitz. T. L. 170. Same scale as in plate Kanthack and Hardy loc. cit.

Figures 10, 12, 13 are camera lucida drawings with Ocular 4, Objective D, Zeiss.

Figure 20 is camera lucida drawing, Ocular 2, Objective A.

Figture 15 a, b and c are sketches from Ocuilar 4, Objective 1/12th.

Figures 1, 2, 3 and 4 show oxyphile cells from mucous coat side by side with oxyphile cells from peritoneal cavity, or blood.

Fig. 1. Frog. (a) Oxyphile cell from mucous coat of intestine.

(b) Oxyphile cell from peritoneal cavity.

Fig. 2. Rabbit. (a) Oxyphile cell from small intestine, (b) coarsely granular oxyphile cell from omentum.

Fig. 3. Rat. (a) Oxyphile cell from small intestine, (b) coarsely-granular oxyphile cell from peritoneal fluid.

Fig. 4. Dog. (a) and (b) Oxyphile cells from small intestine, (c) coarselygranular oxyphile cell from blood.

Fig. 5. Rat. Basophile cells from mucous coat of small intestine.

Fig. 6. Rat. Basophile cell from peritoneal fluid.

Fig. 7. Rat. Basophile cell from small intestine with micro-organisms very abundant in lumen. Animal starved but ate its own feaces.

Fig. 8. Rat. Basophile cell from small intestine, 3 1/2 days' starvation, gut healthy.

Fig. 9. Ferret. Basophile cell from group shown in Fig. 10. Animal digesting.

Fig. 10. Ferret. Tangential section through apex of villus.

Fig. 11. Dog. Almost completely disrupted basophile cell. Small intestine. Compare with this Journal, Vol. XVII. Plate 11, Fig. 13.

Fig. 12. Dog. Optical section, longitudinal through villus of dog showing layer of basophile cells at base of epithelium. Eosine and methylene blue.

Fig. 13. Dog. Optical section, transverse through villus of dog showing the basophile layer.

Fig. 14. Dog. Hyaline cells, various sizes.

Fig. 15. (a) Living oxyphile cell in hanging drop of inflamed lymph. Frog. Trace of methylene blue added. Successive stages in staining of nuicleus shown in (b) and (c).

Fig. 16. Guinea-pig. Hyaline cell holding an ilngested oxyphile cell, and absorbed iron as droplets in its cell substance. Animal fed with Denayer' s peptonate of iron. Absolute alcohol. Potassium ferrocyanide and hydroclhloric acid. Methyl-eosine and methylene blue. Before the application of the iron test the droplets appear bright yellow.

Fig. 17. Guinea-pig. Hyaline cell holding iron.

Fig. 18. Dog. Portion of crypt of Lieberkühn about 3/4 down showing wandering of oxyphile cells into lumen. Corrosive sublimate, methyl-eosine, and methylene blue.

Fig. 19. Dog. Portion of epithelium of crypt of Lieberkühn showing degenerated oxyphile cells. Corrosive sublimate, methyl-eosine and methylene blue.

Fig. 20. Frog. Near junction of hepato-pancreatic duct with intestine. (a) focus of oxyphile cells which continues into gut wall and into peritoneal membrane, (b) pancreas.

Fig. 21. Rat. Cells at base of crypt of Lieberkühn in starving animal (3 1/2 days), (a) oxyphile cell.

Fig. 22. Rat. The same cells in full-fed animal.' (523-524)

Fig. 1 in text:

'in Amphibia (frog and newt) and Reptilia (grass snake) the splanchnic oxyphile cells agree in every particular, in the size of the cell, in the nuclear characters, and in the size, highly refringent nature, and staining reactions of the granules with those found generally in the blood, lymph, and connective tissues of the body (cf. Fig. 1 a with Fig. 1 b). They are cells of about 10 µ in diameter with an eccentrically placed nucleus which is in most cases elongated, and bent round to form a more or less complete horse-shoe.' (494)

Figs. 2-4 in text:

'In type these cells most closely approach the oxyphile cell of the coelomic group, that is the coarsely granular oxyphile cell, and in the ox, sheep, and rabbit the difference is solely one of size. The same typical crowding of granules, characterised by a high refractive index and an intense affinity for acid dyes is observed in both, but the splanchnic cells and granules are somewhat smaller (Fig. 2). Thus in the rabbit the splanchnic oxyphile cell averages 7 µ in diameter while the ccelomic oxyphile cell averages 10 µ.

In the rat too these splanchnic oxyphile cells closely resemble those of the coelomic system except in the matter of size but here the difference is greater than in the previously mentioned forms, the splanchnic oxyphile cell averaging 6 to 6.5 µ while the coelomic oxyphile cell averages 10 µ (Fig. 3). The splanchnic oxyphile cells of the hedgehog resemble those of the rat in type. In the carnivora, ferret, cat and dog, the splanchnic oxyphile cells depart very widely from the types found elsewhere (Fig. 4). The granules are small and have a much diminished affinity for acid dyes, in point of fact the granules agree in size and reaction to dyes much more with the finely granular oxyphile cell of the haemal system than with the coarsely granular oxyphile cell of the extra vascular fluids. On the other hand the cell has a rounded or curved nucleus and never shows the characteristic irregular nucleus of the haemal cell. The granules are smallest and the cell is most extreme in type in the dog and cat, while in the ferret the granules have a very appreciable size and lustre, so that the structural features more closely resemble those found in the splanchnic oxyphile cells of the rat and hedgehog.' (494-495)

Figs. 5 and 7-8 in text:

'The striking changes which the basophile cells exhibit are limited to changes in their granules. The cells are markedly granular in well-nourished animals and become less granular during, starvation though the latter change is not readily brought about and it is difficult to produce extreme exhaustion of the granules (Fig. 8) [note: 'Small though they sometimes are these granules show signs of being complex in substance. Strictly speaking the change above mentioned is limited to that portion of the granule which fixes the basic dye.'].

If however the condition of starvation be complicated with increase in the micro-organisms of the gut, or in any case where micro-organisms are abundant, the basophile cells are much distended with granules which are often much more stable than those found normally (compare Fig. 7 with Fig. 5). (514-515)

Figs. 5-6, 10 and 12-13 in text:

'The splanchnic basophile cells themselves vary much in shape, and in sections of the gut they may appear elongated and flattened or rounded and lobed. They are usually more or less flattened in a plane parallel to the internal surface of the gut and are irregular in shape (Fig. 10). The cell substance is commonly so crowded with granules as to obscure the nucleus. When that structure can be seen it is ovoid or spherical.

In Amphibia and Reptilia, these cells resemble those found elsewhere both in the size of the cells and of the granules. In Mammalia the splanchnic basophile cells always differ from the coarsely granular (coelomic) basophile cell in being smaller and possessing usually very much smaller granules. In the rat for instance the difference between the large basophile spherules of the ccelomic basophile cells and the fine dust-like granules of the splanchnic basophile cells is especially striking (Figs. 5 and 6).

The arrangement of these cells in the villi of Carnivora is very remarkable. It can be most readily studied by mounting entire villi which have been fixed in absolute alcohol and stained with niethylene blue in 80% alcohol, in Canada balsam. The whole villus then appears closely studded with basophile cells and, in optical sections, one sees that they form a defined layer lying immediately under the epithelium (Figs. 12 and 13). Each cell is flattened in the plane of the surface of the villus and is of an irregular shape. Allowing for contraction in preservation it is clear that during life they form a practically continuous layer. We thus find this feature, characteristic of the villi, and of the villi only, namely that a layer of flattened granular cells underlies the epithelium, and to this layer from its histological characters the name "basophile layer of the villi" might fitly be given. ' (498-499)

Figs. 14 and 16 in text:

'As Heidenhain [note: 'Pflüger's Archiv 43, Suppl. 1888.']  pointed out the nuclei sometimes stain a deep opaque colour with nuclear stains, or they may show an open network with or without nucleoli (Fig. 14).

The dark staining nuclei lie in small cells with scanty cell protoplasm, the open, lighter staining nuclei are larger and occur in larger cells. Two extreme conditions of the splanchnic hyaline cells may be recognised, in the first condition the cells present appear to be all of one order, resembling one another in size (about 7 µ) and in the fact that the nuclei all show the open network and do not stain intensely (Fig. 14 b). Here and there however are a very few of the smaller cells with deeply stained nuclei. In the second condition cells are present in all the varying stages from small young cells with darkly staining nuclei, so small and with such scanty cell substance as to deserve the term "lymphocyte," to hyaline cells larger than common and reaching to a diameter of 10 µ (Fig. 14). Under these conditions the percentage of small cells with dark nuclei varies markedly in different regions of the gut and is maximal in the neighbourhood of solitary follicles or Peyer's patches. To this point we shall return later. The larger cells in the second condition sometimes manifest phagocytosis and when swollen with ingesta they may be larger than 10 µ in diameter (Fig. 16). We believe that these hypertrophied hyaline cells are identical with the large phagocytes which Metchnikoff includes under the term " macrophage."' (501)

Fig. 15 in text:

'Direct observation may be brought to bear by observing the staining of spherical cells in a hanging drop faintly tinged with some dye. If for instance a trace of methylene blue be used the cells die anid stain so slowly as to afford ample time for observation and the making of sketches. A quiescent cell was brought into the field of the microscope before the addition of methylene blue. Figures 15 a, b and c represent stages in the slow appearance and staining of the nuclei. The drop was from "inflamed" lymph of a frog and the cell happened to be one of the apparently truly multinuclear oxyphile cells sometimes found in that condition.' (496)

Figs. 16-17 in text:

'It was found that clean staining of the oxyphile granules was possible after the application of the ferrocyanide test to sections. Very beautiful preparations can be made in this way. After the application of the iron test, the section should be washed fairly quickly in distilled water and then very thoroughly in several changes of re-distilled spirit (95%). They are then lightly stained with haematoxylin, washed in tap water, stained with very dilute aqueous eosine, dehydrated, cleared in cedar wood oil, and mounted in balsam. Bismark brown is perhaps a more effective nuclear stain than hematoxylin. In these various ways we demnonstrated the presence of iron in large numbers of wandering cells in the villi and in the spleen. These iron-holding cells were in all cases hyaline cells (Figs. 16 and 17).' (513)

Fig. 18 in text:

'A comparison of all the mammals examined by us leads us to the conclusion that the moving upwards of the oxyphile cells from the basal layer towards the free surface of the gut is at first associated with an increase in their numbers. If therefore we are right in regarding the basal layer as the place of origin of the oxyphile cells, and in viewing the movement thence towards the lumen as an indication of activity on their part, we may say that, at the onset of a period of activity, the cells migrate towards the lumen of the gut and increase in numbers, and some of them pass into the epithelium and through it into the lumen of the gut (Fig. 18).' (505)

'The most obvious structural change which the oxyphile cells manifest within the epithelium or in the lumen of the gut is a diminution in the number of the oxyphile granules even to the total disappearance of these structures. This appears in figure 18, where a small portion of a crypt of Lieberkühn is shown containing oxyphile cells some of which are completely charged with granules, others have only a few, while one has lost its granules. If we limit ourselves strictlv to the evidence furnished by stained sections it is impossible to decide whether this change in the oxyphile cells is or is not necessarily followed by death and complete disintegration.' (506)

'The results of an attempt to determine experimentally the effects of a flesh diet are given on page 516. The most marked oxyphile granulation met with by us among dogs, when the granules were large and the cells very nuimerous, occurred in a portion of small intestine from a fat bitch. The muicous coat however was injured by the presence of worms, and migration was particularly marked, the number of oxyphile cells in the crypts of Lieberkühn being very striking. Fig. 18 is drawn from this animal.' (511)

Fig. 19 in text:

'If we trace the oxyphile cells into the epithelium we find that they thrust themselves between the bases of the cells and move between the cells towards the lumen. Some of the cells suffer degenerative changes while still within the epithelium (Fig. 19). The granules disappear and the nucleus alters. The latter loses all trace of a network and passes into a condition in which it stains evenly. It may also break up into two or three masses. These degenerated nuclei frequently stain a remarkably intense and brilliant green with Ehrlich-Biondi's fluid. Sometimes the nucleus suffers these extreme changes long before the granules have completely disappeared. The appearances presented force one to the conclusion that the oxyphile cells perish in the epithelium and indeed they may shrink there, as though dissolved, and leave cyst-like spaces.' (505)

Fig. 20 in text:

'Oxyphile cells. It is clear from what we have already said that the splanchnic oxyphile cells of mammalia betray in their structural features a close affinity to the ccelomic oxyphile cells. In Herbivora, Rodents and Insectivora the differences between the two are very small and it is not until we come to the Carnivora that any nmarked divergence appears. This suggests that the splanchnic oxyphile cells are closely connected with, or are a specialised portion of, the cceloirnic cells, and we accordingly find in the frog that a focus of proliferation of oxyphile cells may supply both the smiall intestine and the peritoneal cavity (Fig. 20).

The walls of the body spaces - pleural, pericardial and peritoneal - of all animals contain areas, usually related to lymphatic capillaries and to blood vessels, which are crowded with wandering cells. In some areas the wandering cells are oxyphile, other areas again, notably about the diaphragm, contain only vast numbers of basophile cells... In the frog, a focus of oxyphile cells exists in the connective tissue which accompanies the hepato-pancreatic duct, and in sections taken through the small intestine at the point where the hepatic duct opens into it one sees that this focus of oxyphile cells extends into the walls of the gut, and continues a short distance upwards and downwards, gradually thinning out until it merges into the scattered oxyphile cells of the gut walls (Fig. 20). In other words the splanchnic and coelornic oxyphile cells in this region clearly have a common origin.' (518)

Figs. 21-22 in text:

'In the course of our work on the wandering cells our attention was called to changes in the extent of the granularity of Paneth's cells and a comparison of various preparations showed that those from well-fed animals agreed in possessing a scanty granulation (Fig. 22), while those from hungry animals agreed in possessing numerous and large granules (Fig. 21). This change was observed in rats, and it would appear to show that the granules, like those of the salivary glands, pancreas, and other digestive glands, suffer loss during digestion.' (523)