Explanation of Plate XI (figs. 1-15):

'Note. In the preparation of all the figures (except 15) Abbe's camera was employed and all, with the exception of two, are illustrated as they were seen with an immersion apochromatic objective (Zeiss 3 mm., 2 mm. or 1-5 mm. focus). The exceptions are Figs. 1 and 2, in the drawing of which Zeiss D. was used.

Fig. 1. Section of a villus from the pyloric end of the small intestine of a guinea-pig kept on ordinary diet. Alcohol, acid ferrocyanide mixture, balsam. x 305.

Fig. 2. Optical section of a slightly compressed villus from a guinea-pig after the administration of "peptonate" of iron. l, the lacteal vessel. Alcohol, ammonium sulphide, glycerine. x 305. 

Fig. 3. A portion of the mucosa of the intestine in a lake-lizard. e, epithelial cells, l, iron-carrying leucocytes, r, red blood corpuscles, also shown to contain inorganic iron. Alcohol, acid ferrocyanide mixture, balsam. x 620.

Fig. 4. A portion of the epithelium and underlying elements of an intestinal villus of a guinea-pig after the administration of "peptonate" of iron. l, leucocyte, bc, blood capillary. Alcohol, acid ferrocyanide mixture, balsam. x 1240. Drawn with the diaphragm of Abbe's condenser removed from the microscope.

Fig. 5. Portion of a section of the liver of a guinea-pig fed with "peptonate" of iron. l, leucocytes, hc, hepatic cells, bc, blood capillary. x 1240. Drawn with the diaphragm of the condenser removed.

Fig. 6. A portion of the tip of an intestinal villus of a guinea-pig kept on its ordinary diet, to show the distribution in the cells of the organic iron compounds (chromatins). e, epithelial cells, l, leucocytes, a, nuclei of adenoid elements. In the cytoplasm of two of the leucocytes are found granules of an inorganic (?) iron compound. Alcohol, nitric acid alcohol, acid ferrocyanide mixture, balsam. x 1240. Drawn with the diaphragm of the condenser removed.

Fig. 7. Epithelium and underlying elements from the side of a villus of a guinea-pig on the second day of the course of yolk-feeding, to show the distribution of organic iron compounds (chromatins). l, leucocytes, a, the sub-epithelial "membrane." Alcohol, nitric acid alcohol, acid ferrocyanide mixture, balsam. x 1240. Drawn with the diaphragm of the condenser removed.

Fig. 8. Epithelium and underlying elements of a villus of the same animal. a, sub-epithelial "membrane," s, the secretion from epithelial cells. Corrosive sublimate, haematoxylin, eosin, balsam. x 1240.

Fig. 9. Portion of a Lieberkuhnian gland of a guinea-pig to show the distribution of organic compounds of iron (chromatins) and especially of those connected with secretion. Alcohol, nitric acid alcohol, acid ferrocyanide mixture, balsam. x 620. Drawn with the diaphragm of the condenser removed.

Fig. 10. Portion of a fresh villus of a guinea-pig killed five hours after being fed with stale yolk. The epithelium has been removed. Ammonium hydrogen sulphide, glycerine. x 620. Fig. 11. Intestinal epithelial cell of a villus from the same animal. Alcohol, ammonium hydrogen sulphide, glycerine. x 620.

Fig. 11. Intestinal epithelial cell of a villus from the same animal. Alcohol, ammonium hydrogen sulphide, glycerine. x 820. (Zeiss oc. 4, apochr. imm. 1.5 mm.).

Fig. 12. Epithelium and underlying leucocytes of a villus from the same animal, to show the absorption of the yolk chromatin. Alcohol, ammonium hydrogen sulphide, glycerine. x 620.

Fig. 13. A liver cell of an Amblystoma fed artificially for four days with egg-yolk. Alcohol, sulphuric acid alcohol, acid ferrocyanide mixture, balsam. x 820. Drawn with the diaphragm of the condenser removed.

Fig. 14. Yolk spherules from a hard-boiled egg, to show the distribution of the iron (of the haematogen). In the spherule on the left the elements are fewer and coarser. Sulphuric acid alcohol (for 48 hrs.), ammonium hydrogen sulphide, glycerine. x 820. (Comp. oc. 4, imm. apochr. 1.5 mm.)

Fig. 15. Free border of an intestinal epithelial cell of an Amblystoma, fed with yolk. p, cell protoplasm, h, hyaline border, y, elements of yolk. Alcohol, ammonium hydrogen sulphide and glycerine (at 60° C. for eight hours). (Comp. oc. 4, imm. apochr. 1.5 mm.).' (295-297)

Fig. 4 in text:

'In very thin sections treated with the ferrocyanide mixture and mounted in balsam, the distribution of the iron was more clearly seen. Sometimes in the epithelial cells the blue reaction was a diffuse one with blue granules collected in groups here and there in the cell, in some instances it was fouind in the inner end of the cell chiefly, while again the protoplasmic processes in the hyaline border gave an intense reaction. Fig. 4 shows some of these details distinctly. In this are represented three cells, in two of which the inner ends appear loaded with iron and they were fixed in the act of transferring it to the underlying tissue.' (272)

Fig. 5 in text:

'When the dose of iron given was not great, then the iron was mainly if not wholly confined to the peripheral zone. With large doses a greater portion of the lobule was impregnated with iron, and other elements came prominently to view, especially in the "peptonate" preparations. There were leucocytes in the angles of the capillaries in all parts of the lobule but very frequently in the central portion, and their occurrence was manifested under the low power by the strong reaction which they gave for iron (fig. 5). Sometimes each cell was a mass of blue material or it contained large blue masses, in others again, the cytoplasm had a diffuse blue tint with one or more clumps of ironholding substance.' (273)

Figs. 6-8 in text:

'In the guinea-pig, as ordinarily fed, the "masked" iron exhibits in its distribution in the epithelium of the intestine very little difference from that represented in fig. 6, in which the iron is shown in the chromatin of the nuclei and in a narrow zone immediately about some of the nuclei, but in preparations from animals fed with yolk for two or three days, the epithelial cells situated on the sides of the villi and below the tips of the same have the iron distributed as represented in fig. 7, in which the whole of the protoplasm in the lower half of each cell and in the leucocytes below give a uniformly diffuse Prussian blue reaction. The epithelial cells at the tips of the villi are so much distorted by the fat present in them, that a division of each into an internal and external part is impossible, except in some cases where the absorption of fat has ceased to take place. In preparations stained with haematoxylin and eosin, the cells immediately below the tip give usually the appearance represented in fig. 8, but with this exception, that the bodies enclosed in cavities of the protoplasm in the external half of each cell shown in the figure are not present in the preparations from all the animals fed with egg-yolk.' (283-284)

Fig. 9 in text:

'Now as in the secreting cells of all sorts, and especially in those of the Lieberkühnian glands, secretory activity is associated with the presence of a chromatin in the part of the cell remote from the lumen (fig. 9), it might be urged that the increase of the "masked" iron in the inner ends of the superficial epithelial cells of a villus was due not to an iron compound absorbed, but to secretory activity bringing about an increase of the substance governing that process.' (284-285)

Fig. 11 in text:

'It may, in fact, be that the oval vesicles were not connected with absorption at all and that they were merely appearances in the rodlets, although, on this view, the green reaction in their envelopes would be difficult to explain. That the rodlets are not always simple structureless elements, I have, several times, found to be the case in preparations from the guinea-pig (fig. 11), in which each rodlet appeared to be a series of beadlets or granules.' (293)

Fig. 12 in text:

'There was evidence of the transference of yolk chromatin from the epithelial cells to the underlying elements, and this was found in those villi in which the process of fat absorption had not distorted the cells. Sections of these villi, obtained from material hardened in alcohol, when treated with ammonium hydrogen sulphide, gave preparations like that of which fig. 12 is an illustration. The inner portions of some of the cells at the extreme tip of a villus gave a faint greenish reaction immediately after the reagent was added, but in the corresponding portions of other epithelial cells the reaction was given also by granular elements lying between and among the fat droplets.' (290)

Fig. 14 in text:

'The failure of all these experiments led me to use a less abnormal kind of food; and, since, according to Miescher [note: 'Miescher. Hoppe-Seyler's med. chem. Unters. Pt. 4, p. 502, 1871.'], egg-yolk itself contains 1 to 1-5 per cent of nuclein (haematogen), that food substance appeared likely to yield the best results.

I used unboiled egg-yolk, for when egg-yolk is hard-boiled the yolk spherules become thereby fixed in form, and the chromatin-holding particles are set free only when the spherules are digested, but wben the yolk is administered fresh the spherules readily undergo fragmentation and the chromatin-holding particles are liberated and put in a form in which the epithelial cells, if they possess the power, can invaginate them. In the spherules the chromatin is partly in a granular form [note: 'Miescher (loc. cit.) localised the nuclein which he discovered in egg-yolk in the granules of the yellow yolk spherules.'] and, apparently, partly as envelope material to its fat globules, the latter varying in size and shading off into the small granules in such a way as to suggest that the latter are also fat globules of almost infinitesimal size surrounded by chromatin. Fig. 14 gives a representation of two yolk spherules which were fixed by heat and in which the iron, set free by sulphuric acid alcohol, was converted by ammonium sulphide into the sulphide. In it can be seen smaller and larger fat globules surrounded by an iron-holding envelope. The fat is, therefore, closely associated with the chromatin, and as we know the former is in some way absorbed by the intestinal epithelium, the conclusion did not appear to be a strained one that both constituents are absorbed together.' (282)

Fig. 15 in text:

'With 15 mm. apochromatic immersion (Zeiss) and compensation ocular 4, the vesicles could be seen connected by a grayish line. On the protoplasmic side of the margin were also minute vesicles with greenish envelopes, apparently of the same character as those in the striated border and the reticulated protoplasm itself had a slightly greenish tinge (fig. 15).' (292)

Explanation of Pkate XI, figs. 1-3:

'Fig. 1. A portion of the transverse section of an ox's rib. Shewing the effect of a slight digestion by trypsin. Bone hardened in chromic acid and -decalcified by the careful addition of nitric acid. p. Periosteum. pl. That part of shaft wall mainly occupied by periosteal lamellse; the structures are intact beyond the sharper definition of the contour lines where a removal of the cement has occurred. i. Inner part of shaft, in which the effect of the digestion is most marked, The Haversian systems and internal masses of lamellse as also those of the cancellous tissue are intact. a. Gap where the intermediate lamellie are gone. b. Channel for blood-vessel in these lamellae.

Fig. 2. Periosteal lamellae from ox's rib after brief digestion. Bone treated same as that in fig. 1. l. A lamina in longitudinal section. t. One in transverse section. s. Space for lacuna. The outlines of lacunae and canaliculi are omitted.

Fig. 3. Part of a Haversian system. The matrix has become homogoneous and transparent. a. Membrane indicating former limi of the Haversian canal. b. Later membrane lining permanent channel.' (383)

Fig. 1 in text:

'there were to be seen disturbances in the arraigement of the intermediate lamellae in the inner part of the shaft; these presented a confused mass near the inner surface, and in some places were entirely wanting, there being gaps in the section (P1. XI. Fig. 1). The Haversian systems stood intact amidst the debris, as did also the masses of lamellae forming the inner surface of the shaft and cancellous tissue.' (372-373) 

Fig. 2 in text:

In Schäfer's characterization, ''two sets of fibres constituting the lamella are held to be separable from their neighbours along an ideal plane any point of which is indicated by the occurrence of a lacuna' [note: 'A similar opinion is held by Klein. Atlas of Histology. Pt. iv., PI. xII. fig. 2, and p. 55.']

This arrangement is not supported by an examination of the structures in question, for lacunae are found in the thickness and course of one set of fibres, in support of which there are the statements already quoted of v. Ebner (l.c. [' Sitzb. der math. Naturwiss. kais. Acad. der Wiss. Wien, Bd. LXI., Abth. II. 1876.'] p. 70), and his drawings especially Taf. I. Fig. 4, and Taf. III. Fig. 25. b.  [note: 'Ranvier. Trait technique d'Histologie, 1875, p. 314, fig. 101.']. Ranvier [note: 'Ranvier. Trait technique d'Histologie, 1875, p. 314, fig. 101.'] gives drawings unconsciously supporting this view. My own observations have convinced me of the almost constant nature of the fact. Pl. XI. fig. 2.' (365-366)

'Alternate lamellas are similar to each other, and such a view is represented in Fig. 2 (Pl. XI.). One set has the appearance of continuous bands, whilst the other consists of small masses in single rows, separated by bridges that seem to join the band-like set. Both are lamellk of the same constitution, only they are seen in different section.' (366)

'I would maintain that the lamella consists of one set of fibres having a parallel arrangement and that these fibres send bundles of fibrils to one another occasionally, as has been shewn by v. Ebner. The position occupied by the lacunae is one which has a more definite relation to one set of fibres than to two, as specially indicated by their occurrence in the thickness of one set of fibres (Fig. 2, s).' (367)

'Either two sets of fibres are empirically and unnecessarily to be assigned to a lamella, or the more natural division is to be followed of accepting one set of fibres having a common direction, and lying in the same plane, of greater or less extent, as constituting a bone lamella.

In conclusion I would here allude to Fig. 2, which is a sketch of a number of subperiosteal lamellae from the rib of an adult ox. The sketch was made before this particular relation of the lacuna, to the lamellwe was in my mind, and has remained unaltered since.

The alternate layers of bands and masses are lamellee seen in different section. Three lacunae are visible; the two on the right hand are situated in lamellae which have different directions, the upper one being in a situation corresponding to that shewn at s, but not so marked, whilst the lower one lies in a thin longitudinally-cut lamella.' (368)

'In the subperiosteal and intermediate lamelle the fusion into each other of similarly directed lamellke is more frequent than in the Haversian systems. This is well shewn in Fig. 2.' (369)

'Examination with a power of 300 diam. shewed the lamellar structure very plainly in the coherent parts, the alternating arrangement in the matrix of bands and rows of bodies being more distinct than before digestion and very sharply defined...

The appearance was so distinct that there could be no doubt as to the construction to be put upon the observed arrangement. Where the band and dot series were most markedly seen, that is where the plane of section had followed the course of one set of fibres, these alternating layers of bands exhibiting longitudinal fibrillation, and others composed of little blocks separated from each other by almost empty spaces occurred; the blocks exhibited a granular appearance, the sectional view of the fibrils composing the bundles to whose cut extremities the appearance of blocks is due (Fig. 2).

Particularly well seen, in the same locality, were the frequent junctions or fusions which lamellae of a similar direction undergo with each other:... Gentle curves may and no doubt do occutr in the fibres of lamellie, insufficient in themselves however to alter the relation of parts. In Fig. 2 are represented a number of such lamelle, of periosteal formation, these being especially chosen on account of the behaviour of the lamellie and the little regularity of arrangement which obtains in them. (373-374)

'In the figure the lacunar outlines have been omitted; the spaces occupied by them are indicated by the occurrence around them of the bundles of fibrillae; the canaliculi are also omitted. These spaces are in every case, of the three represented in the figure, in the thickness of a lamella, the most marked being at $ (Fig. 2);' (374)

'A point particularly noticeable is the difference in the thickness of the lamelle; the great irregularity seen here (Fig. 2) is one which appertains to the periosteal lamellaw, those of the Haversian systems being much more regular, a fact which has already been alluded to.' (374)

Fig. 3 in text:

'In some systems membranes similar to that lining the canal were found further in the matrix towards the periphery, sometimes at no very great distance from the central canal. Such a membrane is shewn in (PI. XI.) Fig. 3, where b is the internal, and a the outer one, situated in the substance of the matrix.' (379)