On the New Zealand Lamprey, Geotria australis, Gray.
Part 2.—On the Mid-gut Diverticula, the Bile-duct, and the Problem of the Pancreas in the Ammocoetes Stage.
[Issued separately, 25th November, 1930.]
Northern Hemisphere Ammocoetes.
Ammocoetes of Geotria.
Although some observers have had at their disposal a few specimens of the Ammocoetes stage of Geotria—Kner (1869), Smitt (1901), Plate (1902), Dendy (1902)—not one, so far as I can find, has examined the internal anatomy of this interesting form. Smitt (1901) has remarked on its external similarity to the European Ammocoetes, from which, however, it differs in the greater number of pre-anal myomeres. What was my surprise to find, on dissection, and on examination of serial sections, that at the junction of oesophagus and mid-gut, there are constantly present two forwardly directed diverticula, a right and a left! The right is comparatively short and blind, the left is quite long, about half as long as the oesophagus, and into it upon its dorsal surface and near its anterior end opens the bileduct. Below will be given, first, an account of the relations which the bile-duct exhibits in the Northern hemisphere Ammocoetes, and, secondly, will be described and figured the relations of bile-duct and diverticula as observed in the Ammocoetes of the Southern hemisphere lamprey, Geotria. It will be noted that the diverticula are lacking
[Footnote] * During the course of this work, I attempted to extract insulin from the insular organ (Cotronei) of Geotria australis, but without success. Professor J. J. R. Macleod, of Aberdeen University, kindly suggested the method of extraction, which was followed and Dr. Lynch, of the Wellington Hospital, tested the extract on mice. The mice appeared perfectly well after the injection and showed none of the expected symptoms. Only six lampreys were available for the experiment and the insular organ is very small, so the amount of material to work with was also very small. So far as I know, then, physiological evidence of the presence of insulin in this organ is still lacking.
in Northern hemisphere Ammocoetes, and this is the first time, to my knowledge, that they have been observed and described in the Ammocoetes of Geotria.
I take this opportunity of expressing my thanks to Mr. Kevin Rix-Trott for assistance in collecting Ammocoetes, to Mr. A. Waterworth for the micro-photographs, and to Professor H. B. Kirk for criticism and advice.
2. Northern Hemisphere Ammocoetes.
Brief mention of the bile-duct in these forms is made by Langerhans (1873, p. 41), Schneider (1879, p. 90), and Goette (1890, p. 75). Fig. 1 is copied from Fig. 140 of Goette, representing a dorsal view of the heart-liver region of a larval stage (neither age nor length stated) of P. fluviatilis. As text-books rarely describe the anatomy of the Ammocoetes, the following brief account of the bile-duct (Nestler) is given.
Nestler ((1890, p. 107) says “Beginnen wir die Beschreibung bei Ammocoetes. Eine sehr grosse und weite, von ganz niedrigem Cylinderepithel ausgekleidete Gallenblase liegt in der vorderen rechten Leberhälfte. Aus ihr kommt der Gallengang hervor. Nachdem sich die Arteria coeliaca zu ihm gesellt hat, ziehen beide, zu einem gemeinsamen Strang vereint, schräg rückwärts über den Magen hinweg auf dessen linke Seite. Hier, fast am Leberende, mündet der Gallengang in den Anfangsteil des Mitteldarmes, während die Coeliaca in die Darmfalte eintritt.”
Further I have examined an Ammocoetes (length—10.5 cms.) of Entosphenus appendix, from Maple River, Cheboygan County, Michigan, U.S.A. For this specimen I am indebted to Professor Carl Hubbs of Michigan University. Here the bile-duct leaves the liver about half-way down its length, runs parallel with the coeliac artery, and both in a common strand run obliquely backwards, crossing dorsally the oesophagus and entering the mid-gut on the left dorsal side at the junction of oesophagus and mid-gut.
The above figure and descriptions will suffice to show conditions as they are found in European Ammocoetes (Petromyzon) and N. American Ammocoetes (Entosphenus appendix).
3. Ammocoetes of Geotria.
A dissection from the dorsal surface, which requires the removal of nerve-cord, notochord, aorta, kidneys and fat bodies, exposes the diverticula and the bile-duct. The conditions exposed are essentially similar in Ammocoetes of 1.1 cm. length, the smallest I have yet found, and in larger specimens, till metamorphosis sets in. Metamorphosis occurs usually at a length of about 10 cms. At metamorphosis both diverticula disappear, as well as gall-bladder and bileduct. Figs. 2 and 3 are drawn to illustrate conditions as found in Ammocoetes respectively of 2.8 and 9 cm. length.
At the junction of oesophagus and mid-gut (intestine) two forwardly-projecting diverticula, a right and a left, are given off. The right is comparatively short, extends forwards only a short way below the posterior tip of the liver and ends blindly. The left diverticulum is much longer and runs forward parallel to the oesophagus. Before it ends blindly, the bile-duct opens into it on its dorsal surface. The bile-duct leaves the liver about the middle of its length, crosses the oesophagus and debouches into the dorsal surface of the left diverticulum. Though it is correct to speak of right and left diverticula, at the junction of oesophagus, diverticula and midgut, the oesophagus is rather more dorsally situated than the two diverticula, which lie one on either side of it and at a slightly lower level, i.e. more ventrally. (Fig. 8).
In an Ammocoetes of 4 cm. length, the left diverticulum measured 2.5 mm. long, the right 0.5 mm., in a specimen of 7.5 cm. length, the left 5 mm., the right 1 mm.
The liver is found immediately behind the heart region. Its ventral surface is convex, its left dorsal surface slightly concave, the oesophagus and left diverticulum resting in this concavity. (Fig. 10).
On the right side it attains its greatest development. It is broadest anteriorly, posteriorly it tapers to a point on the right side, where the portal vein enters. The anterior surface of the liver is slightly concave and fused with the posterior wall of the sinus venous. A median ventral ligament binds the liver to the ventral body wall—this, however, is quite short and developed only in the anterior region of the liver.
The gall-bladder is located in the right anterior portion of the liver. In small Ammocoetes, it is relatively immense (Fig. 2), and occupies about half the volume of the liver. In such a specimen it is largely naked, i.e. not clothed by liver tissue. But as the liver grows, the gall-bladder does not keep pace with it and in a 9 cm. Ammocoetes it is, relatively to the liver, quite small. (Fig. 3). In
Figs. 2a and 2b.—These represent dorsal and ventral views of dissections of the liver-region of Ammocoetes of Geotria. The scale on Fig. 2a applies to all. In all figures a = dorsal view, b = ventral view. Figs. 2a and 2b. Ammocoetes 2.8 cm. Lettering as in Fig. 3.
the smaller specimens (up to about 5 cms.) the epithelium lining it is very thin and flattened, in the larger specimens it is cubical to cylindrical.
The bile-duct enters the dorsal surface of the left diverticulum near the anterior end of the latter. From here we may trace it back over the oesophagus, which it crosses dorsally in the same strand of tissue as the mesenteric (coeliac) artery, to the liver. In small Ammocoetes it crosses the oesophagus transversely, in larger ones usually obliquely. (Figs. 2 and 3). On reaching the liver, it may be followed a short distance posteriorly before it is lost to the eye in the liver tissue.
The ventral hepatic vein with several tributaries may be observed easily on the ventral surface of the liver. (Fig. 3). It discharges into the posterior ventral region of the sinus venosus. On the right dorsal surface of the liver, close to the gall-bladder, is a much smaller
dorsal hepatic vein, which discharges into the sinus venosus. This vein runs close alongside the mesenteric artery in its short course. It is not shown in the figures.
Anteriorly, the gut-vein runs dorsal to the gut on the side opposite to the spiral fold. A short distance behind the junction of oesophagus, diverticula and mid-gut, it receives a large branch from the dorsal wall of the mid-gut. This branch is made up of veins from
Figs. 3a and 3b.—Ammocoetes 9 cm. The liver is stippled, the gall-bladder clear and the bile-duct black. OES, oesophagus; AM, mesenteric artery; GB, gall-bladder; L.DIV and R.DIV, left and right diverticula; PV, portal vein; INT, mid-gut; SF, spiral fold; BD, bile-duct; VHV, ventral hepatic vein.
both diverticula, and a vein from the spiral fold. Otherwise for some distance immediately posterior to the liver it lies free from the gut. After receiving this branch, it continues forward as the hepatic portal vein to the tip of the liver which it enters. It may then be followed for a short distance along the dorsal surface of the liver, in which organ it breaks up. (Figs. 2 and 3).
The mesenteric artery (coeliac, coeliaco-mesenteric) is given off from the right side, more dorsally than ventrally, of the dorsal aorta at the level of the venous confluent. It courses to the right, obliquely outwards and backwards, through the venous confluent, which it now leaves. It then bends ventrally and runs obliquely inwards and backwards, and along the dorsal surface of the liver to which it is attached. Its course is now more or less parallel with the oesophagus. When it reaches the bile-duct, in the same strand of tissue, it crosses the oesophagus dorsally, then leaves the bile-duct which enters the dorsal surface of the left diverticulum, and now runs along the inner (i.e. right) side of the left diverticulum to the junction of diverticula and oesophagus with mid-gut. Here it enters the spiral fold as artery of the spiral fold or intraintestinal artery.
Fig. 4.—Ammocoetes, 1.2 cm. Horizontal (frontal) section at junction of oesophagus, diverticula and mid-gut seen from above. OES, oesophagus; L.DIV and R.DIV, left and right diverticula; INT, mid-gut; AM, mesenteric artery.
The hepatic artery is a branch of the mesenteric artery. It originates from the mesenteric after the latter in a common strand of tissue with the bile-duct has crossed the oesophagus, and either just before or just when the mesenteric becomes attached to the left diverticulum. In this common strand of tissue are present, then, bileduct mesenteric and hepatic arteries. (Fig. 10). Thus the hepatic artery has to run back (i.e. towards the right) first, and traverse this strand over the oesophagus before reaching the liver. It is a very small artery. Arterial pads are present at its origin from the mesenteric.
Figs. 5, 6, 7, 8 and 9.—Ammocoetes, 1.2 cm. long, cross sections. If Fig. 5 be regarded as Section 1, then Fig. 6 = Sect. 13, Fig. 7 = Sect. 17, Fig. 8 = Sect. 113 and Fig. 9 = Sect. 132. Sections cut at 7 microns. The scale on Fig. 5 applies also to Figs. 6, 7, 8 and 9. OES, oesophagus; AM, mesenteric artery; LIV, liver; L.DIV and R.DIV, left and right diverticula; BD, bile-duct; PV, portal vein; SF, spiral fold. Fig. 5: T.S. in region oesophagus and liver and before left diverticulum. Fig. 6: T.S. at anterior end of left diverticulum; note bile-duct and mesenteric artery passing over oesophagus. Fig. 7: T.S., bile-duct opening into left diverticulum, the actual opening is missed in this section. Fig. 8: T.S. just before junction of oesophagus, diverticula and mid-gut. Note position of mesenteric artery in what will become spiral fold. Note dividing nuclei near lumina of diverticula. Fig. 9: T.S. midgut with spiral fold. The caudal face of these sections is shown.
A series of cross-sections (Figs. 5, 6, 7, 8, 9) from an Ammocoetes 1.2 cms. long and a single section from a specimen 8.7 cms. long are given. (Fig. 10).
Fig. 4 is a horizontal section from a 1.2 cm. Ammocoetes.
The spiral fold originates at the junction of oesophagus, diverticula and mid-gut. Its position is then in the mid-ventral line. (Figs. 2, 3, 9).
The oesophagus is lined by a simple epithelium, the individual cells being tall and cylindrical. (Fig. 5). In cross-section the lumen is not circular, but drawn out into four bays. The cells lining the bays are somewhat shorter than the cells between two bays, and these latter cells are ciliated.
If the elements composing the epithelial wall of a diverticulum be isolated, they will be found to consist of two types of cell, which may be spoken of as columnar and glandular. (Fig. 12). Both types extend from the base of the epithelium to the lumen of the diverticulum, i.e. they are of equal length and lie side by side, so that the epithelial wall is only one cell layer deep, though exhibiting two types of cell.
The columnar epithelial cell is tall and extremely thin and attenuated—so much so that it is practically impossible to obtain a correct picture of it from sections, and it was only by isolating the two types of cell that it was possible to realise its shape. For isolation I found specimens which had been fixed in Bouin's fixative and then preserved in 10 per cent. formalin most serviceable. This type of cell is swollen at its free end, i.e. at the end lining the lumen, and these swollen ends fit in between the pointed or gently rounded ends of the glandular cells. The striated border (Stäbchensaum) which lines the lumen appears to be formed entirely at the free surfaces of the columnar cells and the glandular cells not to be concerned in it—the latter seem to end just below the striated border. The basal end of the columnar cell is also sometimes slightly swollen. The nucleus is placed about the middle of the cell—it is narrow and elongate and resembles in shape and structure the nuclei of the columnar epithelial cells lining the remainder of the mid-gut. These
columnar epithelial cells found in the mid-gut diverticula appear to correspond with the columnar epithelial cells lining the remainder of the mid-gut, e.g. in possession of a striated border and in nuclear structure. But they have been slightly modified in the diverticula owing to the great number and relatively greater individual size of the glandular cells present which are packed between them. The modifications referred to are the attenuated shape and the position of the nucleus—in the mid-gut it is basally situated, in the diverticula it is nearer the lumen.
The glandular cells, found only in the mid-gut diverticula, I propose to describe under the following heads: Shape—tall, columnar, slightly wider at the base than at the end near the lumen, with gently pointed to rounded distal (i.e. near lumen) ends which appear to end within or just below the striated border, of the same height as the columnar cells described above. (Fig. 12). Nucleus—large, in section circular to oval, situated in the basal portion of the cell, with
Fig. 10.—Ammocoetes, 8.7 cm. long, cross section. Fig. 10 shows liver, oesophagus, and bile-duct opening into left diverticulum. In same strand of tissue as bile-duct and dorsal to oesophagus are indicated mesenteric and hepatic arteries. Lettering as in Fig. 11.
prominent nucleolus, rarely two nucleoli. The nucleus has a characteristic vesicular appearance, i.e. the nuclear membrane stains, but the content of the nucleus, with the exception of the nucleolus, stains very faintly. (Fig. 11). Thus in section is seen a circular or oval margin (nuclear membrane)—the deeply stained nucleolus frequently appears attached to the inner side of the nuclear membrane—the remaining content appears clear, though high magnifications exhibit a sparse faintly-staining reticulum.
Staining Reactions: After fixation in Bouin, Zenker or Tellyesnickzy with subsequent staining in haematoxylin, the basal or proximal half of the cell stains much more deeply than the distal half. The cytoplasm in this proximal half appears as a coarse darkly staining mass, frequently exhibiting a coarsely fibrillar structure. This is particularly the case after Tellyesnickzy (acetic-bichromate) and iron-haematoxylin. The distal half of the cell after fixation as
Fig. 11 shows opening of bile-duct into left diverticulum more highly magnified. (Same section as Fig. 10). OES, oesophagus; LIV, liver; BD, bile-duct. In this figure (11) the two types of cell (nucleus) in the wall of diverticulum are shown—glandular cells with basal nucleus with prominent nucleolus and columnar epithelial cells with nuclei much nearer lumen. Note flagella in bileduct. Fixation—Bouin.
above and staining with haematoxylin is much clearer and lighter than the distal half—it contains faint but distinct indications of granules—the granules appear as if partially dissolved out. With iron-haematoxylin the granules are rendered more distinct—it is then seen that they are closely packed at the free end of the cell, ready, no doubt, to be cast into the lumen. From this closely packed area
the granules may be traced in diminishing number to about one-third way down the length of the cell. Benda's fixation test for secretion granules (Bolles-Lee, 1921, p. 315) followed by Mallory's triple stain—makes evident the granules in the closely packed area near the
lumen—they are stained red. Strangely enough, the other end of the cell—the region below the nucleus—shows the same staining reaction as the granules—it is also red. The significance of this will appear when the mitochondria are discussed.
Mitochondria: The fixatives of Regaud, Champy and Gatenby followed by iron-haematoxylin were used to demonstrate mitochondria. The method of Gatenby (Flemming-without-acetic and iron-haematoxylin or briefly, F.w.a.) was most useful. By these methods the cytoplasm is stained a homogeneous gray, while granules and mitochondria stain in the same way, i.e. both appear blue-black on the gray background. The mitochondria are abundant at the basal end of the cell, i.e. between the nucleus and the base of the cell, the granules are abundant at the free end of the cell, so that these two opposite regions of the cell are more deeply stained than the intervening region. At the basal end the mitochondria form a tangle—they frequently give the appearance of a cap or a wig of closely interwoven threads seated on the proximal (basal) end of the nucleus. They are very thin structures and appear sometimes as threads, at other times rather as linear aggregations of dots. From this tangle isolated threads emerge, passing up around the nucleus towards the middle region of the cell. These isolated threads may be observed only in very thin sections and may be traced distally to about one-third the length of the cell. Here they come into relation with a structure which I shall refer to as the clear area or vacuole. This at first puzzling structure was first noticed in F.w.a. preparations. Above the nucleus there is frequently to be seen a large and distinct vacuole, as large as the nucleus or larger. (Fig. 13, b, c, d). Within this vacuole any one of the following appearances may be seen (1) a faint reticulum with granules on it, (2) a collection of granules, (3) a number of fairly large globules, deeply stained, (4) two, three or four deeply stained masses, or finally (5) the vacuole may contain a single immense deeply stained mass.
The mitochondria which emerge from the tangle and extend up alongside the nucleus can be seen in favourable cases to reach the vacuole, and at or in the vacuole they appear to give rise to the granules. The large deeply-stained globules or the deeply-stained masses, which are seen at times in the vacuoles, I am inclined to attribute to the fusion of granules—such fusion perhaps being a result of fixation or depending on the particular state of the granules at that time. In all cases fixation was as rapid as possible—the head was cut off, the organs rapidly dissected out and placed in the fixative.
A re-examination of material preserved in Bouin showed that the vacuole was present here also—in any one section a number of cells exhibited the vacuole. The vacuole sometimes appeared empty—usually it contained a mass of stained material, which did not fully fill the vacuole. A more deeply-stained dot was usually noticeable in the middle of the stained mass. (Fig. 13, e). Vacuoles were also abundant in material fixed by Lane's method mentioned below, they were rare in acetic-bichromate material. More careful examination of F.w.a. material now revealed the fact that in cells, which did not
Fig. 13.—Glandular cells of the left mid-gut diverticulum. a, b, c, and d = from Flemming-without-acetic and iron-haematoxylin preparations. Note basal tangle of mitochondria, nucleus with nucleolus, clear area (vacuole) with mitochondria reaching to it and granules. In a the clear area above nucleus was vague and ill-defined. e = from Bouin to show vacuole; f, g, h = from F.w.a. and iron-haematoxylin showing indications of canalicular system; j = from Regaud (formol bichromate) showing canalicular systems with prozymogen granules of three cells—mitochondria and nucleus shown in middle cell; k = from Mann-Kopsch preparations. Golgi network of three cells.
exhibit a vacuole, there was nearly always present in the corresponding region a clear area—this might be just a vague clear area, two or three small circular areas, vague indications of clear canals (canaliculi) or a definite system of canaliculi. Thus in F.w.a. material structures varying from a canalicular system to a vacuole may be seen—all occupying similar positions above the nucleus and about the middle of the cell. (Fig. 13b, c, d, f, g, h). In material prepared by Regaud's formalin-bichromate method, vacuoles are practically absent. I could find only two or three—on the contrary, the intracellular canalicular system is excellently shown (Fig. 13j). It takes the form of a system of clear canaliculi—three examples are shown. An attempt was next made to impregnate the Golgi apparatus—by Cajal's method (Da Fana cobalt nitrate modification) no success was obtained, though several trials were made. By the Mann-Kopsch method, however, an apparatus was blackened about the middle of the cell, occupying a position very similar to that occupied by the canalicular system. Three examples are shown (Fig. 13k). Many observers are of the opinion that Golgi network and canalicular system represent different pictures of the same apparatus, i.e. in formalin-bichromate the material is dissolved out leaving clear canals, in osmic acid impregnations the material is heavily blackened, leaving black cords. If this be so, the most accurate picture of the structure in question is given as Golgi network (Mann-Kopsch) or canalicular system (Regaud's formol-bichromate) and the various structures seen in F.w.a. preparations, varying from canalicular system to vacuole, together with the vacuole itself, represent apparently distortions of the canalicular system. One further point of interest is this—granules are always found associated with this vacuole, clear area or canalicular system, both in F.w.a. and in Formol-bichromate preparations—this is the most proximal region of the cell in which granules are to be seen—from here they may be followed distally to the free end of the cell. That the first-formed granules (prozymogen of Saguchi) are formed under the influence of the canalicular system (Golgi apparatus) there seems no reason to doubt, and from what has been said before, there is reason to believe that the material of which they are formed is derived from the mitochondria.
The thesis I wish to maintain is this—that the glandular cells, which characterise the diverticula, are to be regarded as pancreatic cells comparable to the exocrinous pancreatic cells of vertebrates: in other words, there is in the Ammocoetes stage of Geotria no definite and compact pancreatic gland, but the pancreatic (exocrinous) cells are still scattered in the gut wall in the diverticula. The endocrinous constituent of the pancreas (Islets of Langerhans in vertebrates) is represented by the Follicles of Langerhans in the Ammocoetes.
Schafer (1916) in discussing the pancreas remarks:
1. That the inner two-thirds of the cells are filled with granules;
2. That in haematoxylin-stained sections the outer part of the cell is coloured more deeply than the inner;
3. That pancreas cells frequently exhibit a rounded mass of mitochondria near the nucleus.
Horning (1925) has shown that in the guinea-pig the zymogen granules are constricted off from the ends of the mitochondria. Saguchi (1918) has made an intensive study of the glandular cells of the frog's pancreas. He finds—
1. The nucleus is provided with usually only one nucleolus.
2. After fixatives containing a large amount of acetic acid the basal portion of the pancreatic cell exhibits a fibrillar structure and stains more deeply with haematoxylin. I find the same in the Ammocoetes, particularly after fixation in acetic-bichromate. (Tellycsnickzy). According to Saguchi, this is due to the presence of his “protofibrillae,” which he regards as morphological constituents of the cell. In fixation some of the plasm is removed, the protofibrillae thus individualised and a certain amount of adhesion of them follows, hence the fibrillar structure seen in sections. The protofibrillae are not to be confused with mitochondria.
3. The mitochondria have the form of rods or filaments and are crowded round the nucleus. The mitochondria are used up in the formation of the zymogen granules. Saguchi derives the mitochondria from the nucleus.
4. Zymogen granules are never found between nucleus and basement membrane. Above the nucleus is a clear area—“secretogenous area”—to this the mitochondria converge and here small granules (prozymogen) are formed by the disintegration of the mitochondria. The granules leave this area, increase in size and collect at the free end of the cell as zymogen granules.
5. The Golgi apparatus is placed above the nucleus. It appears to occupy the same position as the “secretogenous area,” where the granules are formed from the mitochondria.
6. After Regaud's fluid (and others) a system of canaliculi with clear lumina may be observed in the “secretogenous area”—within the meshes of this system are prozymogen granules—Saguchi considers the canalicular system is to be identified with the Golgi apparatus, the former being the negative of the latter.
From my own observations I can affirm that the nucleus of the pancreatic cells of trout, frog (Hyla aurea) and rat appears clear and vesicular with a prominent nucleolus. Lane's method for pancreatic islet tissue (Carleton, 1926, p. 279) stains also the zymogen granules of the acinous cells. Pieces of frog pancreas and the Ammocoetes diverticula were prepared by this method. The zymogen granules (frog) and the granules of the glandular cells (Ammocoetes) reacted to the stain in just the same way—both were stained purple.
From the comparisons made above I venture to say that the glandular cells found in the diverticula (Ammocoetes) may be compared to the exocrinous pancreatic cells of the vertebrates. If this be a true comparison, an interesting conclusion follows. We then have in the Ammocoetes the most primitive condition of the pancreas, i.e. the pancreatic cells are still scattered in the gut wall and not compacted into a definite gland. So far as I can find, no pancreatic cells have been described in Amphioxus. Further, such cells (glandular, pancreatic) almost certainly occur in European Ammocoetes, as I shall attempt to show below.
One further point should be noted here—as mentioned before, the epithelium lining the diverticula is provided with a striated border. Within this border may be seen tiny dots or rods (Fig. 11), and in tangential sections it can be seen that these are really sections of a network placed in the striated border. Krause (1923) figures a similar structure in the “Stäbchensaum” of the gut-epithelium of the pike. This structure is the “Schlussleistennetz” or network of terminal bars.
After studying these glandular cells in the mid-gut diverticula of New Zealand Ammocoetes, the question naturally arose—do such cells occur in European Ammocoetes? The mid-gut diverticula are absent in such forms—the anterior portion of the mid-gut suggests itself then as the most probable spot. I have not been able to observe European Ammocoetes myself—but the reports of two earlier investigators—Brachet (1897, 1897a), and Picqué (1913)—have convinced me that such cells occur.
Brachet (1897), at the end of his paper, draws attention to a point in European Ammocoetes which he considers of importance. It is this: “La texture de l'épithélium de l'intestin moyen, est toute différente, dans la partie antérieure, qui fait immédiatement suite au ‘Vorderdarm,’ de ce qu'elle est dans le reste de son étendue.” Here among the ordinary epithelial cells, “on trouve un très grand nombre de cellules toutes spéciales,” which stand out strongly when stained (borax-carmine, safranin) from the others. “Ces cellules, très allongées, s'étendent de la membrane proprà la surface libre de l'épithélium: leur extrémité tournée de ce dernier côté est garnie d'un plateau strié, moins élevé, semble-t-il que celui des cellules ordinaires de l'intestin moyen. Le corps de ces cellules, fortement coloré en rouge, par le carmin ou la safranine, se montre constitué de deux moitiés assez nettement distinctes. La moitié externe est homogène ou très finement granuleuse. C'est elle qui contient le noyau. La moitié interne dirigée vers la surface libre de l'épithélium, se colore moins fortement par le carmin, est moins homogène, montre des granulations plus ou moins nombreuses, et des taches claires irrégulières” … “Ce qui caractérise encore ces cellules, c'est l'aspect tout particulier de leur noyau. Il est constitué d'une mince membrane, très peu chromatique. Au centre du noyau se voit un très gros nucléole, absorbant très fortement les matières colorantes. Il semble que toute la chromatine du noyau s'est condensée dans ce corpuscule.” Brachet remarks first on the zone which these cells occupy “dans la portion initiale de l'intestin moyen,” secondly on their characteristics which suggest those of pancreatic cells. He remarks on Mayr's conception—that in the ancestors of the Selachians, having as yet no pancreas, there must have existed a zone, occupying the dorsal region of the mid-gut, and containing the materials at the expense of which the pancreas forms itself among actual Selachians. “Chez l'Ammocoetes, cette zône paraît exister,
mais elle n'occupe pas seulement la partie dorsale de l'intestin moyen; on la retrouve sur toute sa circonférence. Si c'est là réelement une zône pancréatique, ce que je crois, on peut admettre qu'elle contient les matériaux aux dépens desquels s'édifieront non seulement le pancréas dorsal, mais aussi le pancréas ventral.”
The agreement between these cells described by Brachet and those described by me in Geotria Ammocoetes is striking—position of nucleus, nuclear structure, nucleolus, division of cell body into two halves, presence of granules in inner half. It will further be obvious that the thesis maintained here is not original—it was suggested first by Brachet so far back as 1897, but no one since appears to have followed it up. It was Brachet's paper which suggested to me an explanation of these glandular cells and gave me a means of linking them up with similar cells in European Ammocoetes.
Picqué's paper (1913) is longer, and I do not agree with his main conclusions, but certain points brought out are of interest as regards the glandular cells. He considers a pancreas is present in Ammocoetes and in adult—in his view, with which I disagree—it is formed in the Ammocoetes by the follicles of Langerhans, in the adult by the organ constituted by an aggregation of such follicles, which Cotronei (1927) has since identified as an insular organ. Picqué studied the origin of this organ in very young Ammocoetes (5 mm. to 7 mm. long in P. planeri) and I copy his scheme as Fig. 14. Where the oesophagus is continuous with the mid-gut, two pads, a dorsal and a
Fig. 14.—Scheme showing the disposition of the “bourrelets (pads) d'invagination de l'intestin antérieur (1A) dans l'intestin moyen (IM).” Copied from Picqué (1913). This scheme refers to young larval forms of P. planeri and P. fluviatilis. C.ch. = bile-duct; BID and BIV, dorsal and ventral pads.
ventral, are formed by the oesophagus being slightly invaginated into the mid-gut. The ventral pad is the larger. It is at the level of these pads that the follicles of Langerhans (pancreatic rudiments of Picqué) are formed. Picqué's sections through these pads show externally follicles of Langerhans and internally the intestinal epithelium. The interesting point is this—that in this epithelium two kinds of nuclei are present—nuclei such as occur in the remainder of the intestine and nuclei with large nucleoli, standing out sharply from
the others. On p. 21 we read: “Puis le bourrelet (pad) se dessine (Pl. 1, Fig. 1), constitué par une masse d'éléments cellulaires, dont les noyaux sont surtout caractérisés par un nucléole en général unique et assez volumineux. Ces cellules à gros karyosomes font manifestement partie de l'épithélium intestinal.” Farther down on the same page, “Les coupes suivantes nous conduisent en plein bourrelet. Ici les noyaux innombrables tranchent absolument sur ceux de l'épithélium intestinal par leur nucléole unique, gros et de volume croissant vers le dehors.” Picqué does not explain the presence of these nuclei with large nucleoli. He recognises clearly, however (p. 41), the presence of two types of cells in the epithelium of the pads and thinks the follicle cells (his pancreatic cells) may come from the cells with big nucleoli, but is not sure. He does not discuss these nuclei with big nucleoli any further.
I believe they are the nuclei of the cells which Brachet (1897) had already described in older Ammocoetes. Again Picqué's sections showing the mid-gut epithelium (pads) with two kinds of nuclei show a striking agreement with my sections of the diverticula.
To summarise—in these three cases—the pads of Picqué in Ammocoetes of 5 to 7 mm. length—the epithelium of the most anterior portion of the mid-gut in European Ammocoetes (Brachet 1897)—the mid-gut diverticula of Geotria Ammocoetes—we find a feature common to all—characteristic cells denoted in particular by a nucleus with a large prominent nucleolus. As I have indicated above, these cells are to be compared to the pancreatic (exocrinous) cells of Vertebrates.
It seems possible that Picqué's “pads” and the diverticula of Geotria Ammocoetes are to be regarded as homologous structures, but that in the former case they develop very slightly while in the latter they increase enormously. However, the pads are dorsal and ventral (Fig. 14), while the diverticula are right and left. I have not been able to obtain Ammocoetes smaller than 1.1 cm. long, so as to whether there is any change of position as regards the diverticula in earlier stages I cannot say. That a rotation of the gut may occur is not improbable. A rotation of the gut (Geotria) certainly occurs at metamorphosis, as I have been able to observe.
The position of the follicles of Langerhans is another point of agreement. Picqué finds them in the furrow or groove between oesophagus and the pads of the mid-gut—in Geotria they occupy a similar position, i.e. in the furrow between oesophagus and the origin of the mid-gut diverticula.
Keibel (1927) in his Fig. 8 gives a frontal section passing through the passage of fore-gut into mid-gut of an Ammocoetes, 7.5 mm. long, of Lampetra fluviatilis. The section shows two structures (Wülste) which strongly suggest pads (bourrelets) or the rudiments of mid-gut diverticula. From the section they are right and left, and the left is the larger. Their histological structure cannot be observed. It would be interesting to know if the two types of nuclei (cells) described above are present in them. In his Fig. 10 Keibel shows a sagittal section at the passage of fore-gut into mid-gut of an Ammocoetes 9.2 mm. long, of Lampetra planeri. In the legend he remarks
“Keine Wülste zwischen Vorderdarm und Mitteldarm, aber Epithelgrenze.”
Considering these two sections, therefore, we see that the “Wülste” (structures which I have referred to as suggesting “bourrelets” or rudiments of mid-gut diverticula) are right and left, and absent dorsally and ventrally. They therefore certainly suggest structures, only slightly developed and homologous with the mid-gut diverticula of Geotria larval forms.
Finally it would have to be supposed that in Geotria one pad or diverticulum (the left) had carried forward with it in its development the opening of the bile-duct, since this no longer opens at the junction of oesophagus and mid-gut, but far forward into the left diverticulum.
In this paper the following view is taken regarding the question of the pancreas in lampreys:
Both constituents of the pancreas (exocrinous and endocrinous) occur in the Ammocoetes stage. The exocrinous constituent is represented by the glandular cells occurring in the anterior region of the mid-gut—in Geotria in the mid-gut diverticula. The endocrinous constituent is represented by the follicles of Langerhans.
Only the endocrinous constituent persists in the adult. At metamorphosis the follicles of Langerhans increase in number and become aggregated together to form a definite gland. This is the insular organ which Cotronei (1927) has already described as being composed of insular tissue. In Geotria the diverticula are lost at metamorphosis and with them the glandular cells (exocrinous constituent). Probably these cells are lost also in European Ammocoetes at metamorphosis, but this would have to be ascertained.
The Ammocoetes stage of Geotria australis possesses two forwardly-directed mid-gut diverticula, a left and a right. Into the left opens the bile-duct. Such diverticula are unknown from other Ammocoetes.
The diverticula contain characteristic glandular cells which are to be compared to the exocrinous pancreatic cells of Vertebrates.
In the Ammocoetes stage the pancreas is not yet differentiated as a special gland—we have a stage in which the pancreas cells are found distributed in the epithelium in a certain section of the gut wall. This view was first expressed by Brachet (1897a, p. 773). This paper attempts to confirm this view and to bring fresh evidence in support of it from the study of Geotria.
Addendum: Through the kindness of Dr. Horst Boenig I have received a copy of his recent paper—Studien zur Morphologie und Entwicklungsgeschichte des Pankreas beim Bachneunauge (Lampetra (Petromyzon) planeri). III Teil. Die Histologie und die Histogenese. Zeitsch. für mikr-anatom. Forschung—17 Band, ½ Heft, 1929. Unfortunately I have not got part 1, nor is it obtainable in New Zealand, consequently in the criticism which I offer I feel rather handicapped by not knowing its contents. Dr. Boenig considers that the Follicles of Langerhans, formed as buds from the gut wall and from the epithelium of the bile-duct, constitute in the lamprey a pancreas “Morphologisch dem Pankreas der Höheren Wirbeltiere homolog” (p. 181). Essentially the same view has been expressed before by Picqué (1913) and by Keibel (1927). For the following reasons I cannot subscribe to this view: 1. In Vertebrates the liver and the pancreas originate at practically the same time—here, however, the first indications of the “pancreas”? appear after the liver is completely formed—according to Boenig in 1.2 cm. Ammocoetes—further during the whole larval life “pancreatic” buds continue to be formed, while at metamorphosis the epithelium of the bile-duct proliferates strongly, forming many more “pancreatic” buds (caudal pancreas of Boenig), the bile-duct, as such, disappearing. 2. Though, in the adult lamprey, the cranial and caudal sections of the “pancreas” are respectively dorsal and ventral in position and hence are compared by Keibel and Boenig to the dorsal and ventral pancreas of higher Vertebrates, Boenig (Part 2, p. 591, 1927) himself admits, “dass der Entstehungsmodus des ‘dorsalen und ventralen Pankreas’ bei Lampetra planeri ein ganz anderer ist als der bei den anderen Wirbeltieren” and again “Beim Ammocoetes findet sich weder eine ‘dorsale’ noch zwei ‘ventrale’ Pankreasanlagen.” In fact, the formation of the lamprey “pancreas” will not fit in with any scheme of pancreas-formation, which holds good for other Vertebrates. 3. There is no pancreatic duct—at no time is there ever any indication of one. 4. There is no zymogen tissue nor every any sign of such.
Since the main characteristics of the pancreas, as we know them from other Vertebrates, appear to be missing here, the homology with the Vertebrate pancreas drawn by Picqué, Keibel and Boenig seems to me of very doubtful value. There remains, however, the view originated by Brachet (1897) and advocated in this paper. So far as I know, there is nothing incompatible in the view that these follicles or buds represent islet tissue, either as regards their origin or structure. Cotronei (1927) has already described the “pancreas” of Boenig as an insular organ. The remarkable cells present (a) in the anterior region of the mid-gut in European Ammocoetes (b) in the mid-gut diverticula in Geotria Ammocoetes, bear so many and striking resemblances to pancreatic (exocrinous) cells that it seems reasonable to regard them as such. Neither Picqué, Keibel or Boenig are able to offer any explanation of these cells.
Two further points in Boenig's paper call for notice. Dr. Boenig constantly speaks in Part 3 of the follicles (his pancreas) as being derived “durch Wucherung des Vorderdarmepithels” (p. 179), also on pages 137. 145, 146, 164, etc. This appears strange, since both
Fig. 15.—Microphotograph: T S wall of left diverticulum. F.w.a. and iron-haematoxylin. Note basal tangle mitochondria, nucleus with nucleolus (black), clear areas (circular) and granules in inner half of cell.
Picqué (1913) and Keibel (1927) speak of the follicles as being derived from the mid-gut epithelium, while Boenig himself in Part 2 derives the follicles from the mid-gut epithelium e.g. “Das Pankreas legt sich zunachst; au [ unclear: ] dem Epithel des Mitteldarmes enstehend, multipel in Form einer Spange …” (p. 590). Again on p. 144 Dr. Boenig speaks of Brachet's pancreatic cells as being in the foregut epithelium, but according to Brachet (1897) and Picqué (1913) these cells are located in the mid-gut epithelium. In Ammocoetes of Geotria they are in the mid-gut diverticula.
Brachet, A., 1897. Sur le développement du foie et sur le pancréas de l'Ammocoetes. Anat. Anz., Bd. 13.
—— 1897a. Die Entwickl. u. Histog. der Leber u. des Pankreas Ergeb. d. Anat. u. Entwickl., 5.
Carleton, H. M., 1926. Histological Technique. Oxford University Press.
Cotronei, G., 1927. L'organo insulare di Petromyzon marinus. Pubblicazioni della Stazione Zoologica di Napoli, vol. 8, Fasc. 1.
Dendy, A., 1902. On a pair of ciliated grooves in the Brain of the Ammocoete apparently serving to promote the circulation of the fluid in the Brain-cavity. Proc. Roy. Soc. London, vol. 69.
Giacomini, E., 1900. Sul pancreas dei Petromizonti con particolare riguardo al pancreas di Petromyzon marinus. Verhandl. d. Anat. Gesellschaft auf der vierzehnten Versammlung in Pavia.
Goette, A., 1890. Entwickl. d. Flussneunauges (Petromyzon fluviatilis). Abhandl. zur Entwickl. d. Tiere, Heft. 5.
Horning, E. S., 1925. Histological observations on pancreatic secretion. Aust. Journ. Exp. Biol. & Med. Sci., vol. 2, part 3.
Keibel, F., 1927. Zur Entwickl. d. Vorderdarmes u. d. Pankreas b. Bachneunauge Lampetra (Petromyzon) planeri u. b. Flussneunauge Lampetra (Petromyzon) fluviatilis. Zeitschrift für mikr.-anatom. Forschung. Bd. 8, Heft. 3, 4.
Kner, 1869. “Reise der Österreichischen Fregatte Novarra um die Erde.” ‘Zoologie,’ Bd. 1, Fische, p. 421.
Krause, R., 1923. Mikroskopische Anatomie der Wirbeltiere. Vierte Abteilung. Berlin und Leipzig.
Langerhans, P., 1873. Untersuchungen über Petromyzon Planery. Ber. Verh. Naturf. Ges. Freiburg. i. B., Bd. 6, Heft. 3.
Lee, A. Bolles, 1921. The Microtomist's Vade-Mecum. Eighth edition. London.
Nestler, K., 1890. Beiträge zur Anat. u. Entwickl. von Petromyzon Planeri. Arch. f. Naturgesch, Bd. 56, Heft. 1.
Picque, R., 1913. Recherches sur la Structure et le Développement du Pancréas chez Petromyzon. Mém. de la Soc. Zool. de France, Tome 26.
Plate, L., 1902. Studien über Cyclostomen, 1, Systematische Revision der Petromyzonten der südlichen Halbkugel. Faun. Chil. Zool. Jahrb. Suppl. 5, 4 Heft., 2 Bd.
Saguchi, S., 1918. Studies on the glandular cells of the frog's pancreas. Amer. Journ. Anat., vol. 26.
Schafer, E. A., 1916. Essentials of Histology. Tenth edition. London.
Schneider, A., 1879. Beitrage zur verg. Anat. u. Entwickl. d. Wirbeltiere. Berlin.
Smitt, F. A., 1901. Poissons d'eau douce de la Patagonie. Bih. K. Svensk. Vetensk.-Akad. Handl., Bd. 26.