nature of the vessels, which very frequently were torn in injecting them with water, and partly from the soft, pulpy, and exceedingly fatty condition of most of the livers of this class of animals.

In the flounder the connexion is shown in fig. 66, b c and e. The secreting tubes represented are much smaller than is usual in this class, depending principally upon exosmose of their fluid contents, owing to which the tube shrinks, and its outline is clearly distinguished. I have succeeded in injecting the ducts of the sturgeon and frog-fish, and the connexion between the very narrow ducts and the cell network is represented in fig. 64, a c, and in fig. 65, g h k.

Even in the fatty liver of the cod I once traced the continuity of the narrow ducts, with the very wide tubular network distended with cells containing oil and free oil globules.

In injecting the livers of fish, the injection must be diluted with weak spirit, or it does not penetrate to the smallest branches. The colouring matter which is employed is the same as in other cases. Often the particles of the injection accumulate in some of the finer ducts, forming what appears to be rounded and slightly dilated extremities; for the further continuity of the tube cannot be detected. Indeed, so perfect is the resemblance, that it is only by carefully examining many different specimens that one becomes convinced of its fallacy.

In all four classes of vertebrate animals, the arrangement of the ducts, and the relation which they bear to the secreting cells, is very similar. I have seen both in injected, and also in uninjected specimens, the communications between the finest ducts and the cell-containing network. Of the nature of this continuity there can, I think, be no doubt. I can conceive no other explanation of the facts I have observed, or of the appearances presented by my preparations. The observations upon uninjected specimens, shown in figs. 41, 42, and 43, were made early in 1854, many months before I had succeeded in injecting the ducts. The arrangement of the most minute ducts varies somewhat in different animals, as has been described. In some they form a network of very narrow tubes, continuous with those in which the livercells are contained; in others, these communications are exces

sively few in number; while in some, I do not think they exist at all. But I would not express myself positively upon this point, for I feel persuaded that in the most perfect injection which I have yet made, the minute ducts have not all been injected. According to the observations just described the cells of the liver correspond in all essential characters to the secreting cells of other glandular organs. They lie within a cavity of basement membrane, which is here arranged so as to form a network, the tubes of which are directly continuous at various points with very narrow efferent ducts. Now, this narrowing of the duct before it becomes connected with the secreting portion of the organ, is seen in other glands. In the kidney, the total diameter of the straight and ductal portion of the renal tube is considerably less than that of the convoluted and glandular part, although the central cavity is wider, which allows of a very rapid removal of the secreted products formed in the convoluted portion of the tube. The cavity of the very narrow ducts of the liver, although so small, would doubtless admit the passage of a larger quantity of fluid, within a certain time, than the variable and irregular interstices existing between the cells and the basement membrane in the secreting portions of the network. A somewhat similar arrangement occurs in many other glands. In the liver, where the secretion is highly elaborated, and slowly removed from the secreting structure in a concentrated form, we should naturally expect to find the contrast between these two different portions of the gland even more remarkable than in the examples referred to. This is really the

case.

Epithelium of the Smaller Ducts.-The larger ducts, as is well known, have a thick lining of columnar epithelium; but the cells become shorter towards the smaller channels. In the smallest tubes, the cells of epithelium are somewhat flattened, and are of a circular or oval form, which latter, in many instances, is due to the ducts being much stretched in preparing the specimen. These minute cells have a pale granular appearance, and it is not often that a nucleus can be distinctly seen within them. The epithelium of the ducts is not dissolved by caustic soda so readily as the liver-cells; indeed, the former cells are scarcely altered by weak solutions, while the latter are rendered very soft and trans

parent (page 40). The quantity of this epithelium in the smaller ducts varies very much; sometimes it completely lines the tube; in some instances it is so abundant as apparently to leave no distinct cavity in the duct (a condition I have met with in the rabbit, turkey, and fowl); while it is not uncommon to find some of the finest ducts, containing only a very few cells, scattered at irregular intervals over the basement membrane, of which the walls of the small ducts are entirely composed. In a perfectly normal condition, when the minute ducts are undisturbed by manipulation, and are examined in a proper medium, they are generally seen to be lined by epithelium; but, from the extreme minuteness of these ducts, and the tenuity of their walls, and not less from the very delicate nature of the epithelium itself, there is no wonder that we should fail in making out distinctly their epithelium in every instance in which we search for it.

It is not easy to lay down with precision the exact point at which the change in the character of the epithelium of the ducts occurs; but it appears to me that the alteration is a gradual one, and that the cells become shorter and shorter as the diameter of the ducts diminishes. In ducts of the 1-600th of an inch, and in smaller ones, the epithelium presents the characters above described.

It will be found that this ductal epithelium does not pass by insensible gradations into the secreting epithelium, or hepatic cells, but ceases abruptly at the point where the narrow duct becomes continuous with the wide secreting cell-containing network. Nowhere, that I know of, can be seen, in so small a space, a more striking contrast between secreting epithelium, and that which lines an efferent canal. The large characteristic cell of glandular epithelium ceases at the point where the small cell of protective or ductal epithelium commences.

The arrangement here referred to, is very similar to that which is met with in the gastric gland. It is only in the lower part (stomach tube) that the cells of spheroidal epithelium,-which alone there is every reason to believe take part in the secretion of the gastric juice, are found. The upper portion or duct (stomach cell) is lined with columnar or subcolumnar epithelium.* In these glands the secreting cells are not arranged with any order or

* Todd and Bowman's Physiology.

regularity round the basement membrane of the tube, as is the case in the kidney; but appear, in the ordinary state of the parts, almost to fill its cavity, so that the secretion, having escaped from the cells, must pass off by the slight interstices which exist between them and the membranous wall of the tube. As I have before observed, the same irregularity occurs in the arrangement of the secreting cells in the tubes of the liver, and also, but in a less remarkable degree, in the secreting portion of many other glands, as the pancreas, the lacteal glands, and sweat glands.

In a few instances I have seen tubes containing liver-cells apparently lined with the delicate epithelium of the ducts, an observation which would tend to confirm those of Mr. Wharton Jones, who has seen liver-cells in the smaller ducts. It is probable that in these cases the cells have entered the ductal portion of the tube accidentally. The tubes in the cases referred to, however, were uninjected, and therefore I am not disposed to lay much stress upon the observation.

The epithelium of the small ducts presents much the same character in all the animals which I have examined. In figs. 42 and 43, g, the characters of the epithelium in the more minute ducts of the human subject are represented. In the cat it is shown in fig. 35; and in fig. 34, in the lamb. In the latter figure an opportunity is afforded for contrasting the characters and size of these small granular cells with the large hepatic cells.

Dr. Handfield Jones' Observations.-My friend, Dr. Handfield Jones, has arrived at a totally different conclusion with reference to the nature of the smallest ducts, and the epithelium which is contained within them. The latter he regards rather in the light of nuclei, "which are set close together in a subgranular or homogeneous basic substance;" and to these nuclei he assigns the important office of absorbing the material which is secreted by the liver-cells, and which he considers bathes the terminal extremities of the ducts in which this epithelium is contained, or with which it is in direct continuity. Dr. Handfield Jones also observes, that there is seldom any trace of basement membrane to be found in ducts having epithelium of this character, and, he considers that, the basement membrane imperceptibly ceasing, the ducts terminate in this epithelium. Upon such a view, however, I should be

quite unable to explain a number of appearances which I have often seen; and I have been led to a totally different conclusion upon this point, as has been fully discussed. That view seems to me to be the only one, with reference to the termination of the ducts, compatible with the facts which have been observed. When the ducts are injected, according to the method described in Chap. I., the epithelium and the injection may be seen very distinctly with a quarter of an inch object-glass, in narrow tubes, often not more than the 1-3000th or 1-2000th of an inch in diameter, and I have never been able to satisfy myself of the existence of blind extremities. The investigation, however, must be admitted to be a most difficult one, and the question at issue of such a delicate nature, that I am very unwilling to convey an impression of speaking too confidently in favour of the results of my own work, or in the least degree to appear to disparage the conclusions which other observers have arrived at,-no matter how essentially these conclusions may be at variance with my own, or how strongly I may feel convinced in my own mind of the truth of my deductions, and of the correctness of the data upon which they have been founded.

Diameter of the finest Ducts.-The diameter of the finest ducts can only be obtained approximatively; for when not injected, they can only be demonstrated distinctly in fortunate specimens, and are probably somewhat narrower than during life. When injected, on the other hand, they are usually distended, and sometimes to a very considerable extent. In the pig, the smallest branches containing a little injection are not more than the 1-3000th of an inch in diameter; in the human subject, about 1-2500th; in the seal, 1-3000th; in some fishes, not more than 1-5000th.

The diameter of the cavity of the tube and the total diameter of ducts of different sizes are shown in the following Table :