The Structure of the Otolith of the Hake

Work on the life-history of the hake Merlucius merluccius Linn, has involved the possibility of assessing the age of the fish by counting the rings which are present in the otolith. The results of this investigation have no place in the present paper, which simply seeks to show to what details in the microscopic structure of the otolith the rings are due.

It is curious that the otolith has attracted so little attention from morphologists and histologists. So far as I know, there has been only one exhaustive description of the microscopic structure of the otolith, that of Immermann On the plaice otolith (1908). Maier (1908) deals less fully with the Cod otolith and Lissner (1925) mentions briefly that his observations on the otolith of the herring confirm Immermann’s on the plaice otolith. I cannot agree with some of Immermann’s observations, at least as concerns the hake otolith.

The structure has been worked out by a careful examination of fresh entire otoliths, of sections cut and ground thin, and of decalcified entire otoliths and sections. For the finest details of the structure I was obliged to obtain thinner sections than I could myself prepare by grinding, and I am indebted to the Lomax Palaeobotanical Laboratories, of Bolton, for two excellent preparations. The photographs which illustrate this paper are due to the skill of Mr. H. Stokes, of the staff of the Fisheries Laboratory at Lowestoft.

Seen in its natural position in the Sacculus, the hake’s otolith (Sagitta) is an irregularly pear-shaped body, strongly compressed laterally, tapering to a point posteriorly, and bluntly rounded anteriorly. The lower edge is fairly smooth, and is slightly curved, whereas the upper edge is much more indented, and rises in a fairly straight line from the pointed posterior end to well forward of the centre of the otolith, after which it falls in a smooth curve to the blunt anterior end. In fig. 1, Pl. 24, a group of five hake otoliths may be seen lying on their side, with their inner surfaces upward, photographed against a dark background.

The otolith is somewhat curved, in such a way that the inner surface (that next the brain) is convex, the outer surface concave. The inner surface is relatively smooth, and is scored by the wide and shallow Sulcus Acusticus. The outer surface is very rugose.

A number of grooves, concentric and parallel to the margin of the otolith, may usually be seen on the smooth inner surface of the otolith, and are clearly marked in wax impressions of this surface. They are usually not visible on the outer surface.

The hake’s otolith is a compound structure, built up of a large number of lobes arranged in one plane, radiating from the middle line, towards the periphery, in such a way that their axes are normal to the margin of the otolith. The lobes may be seen very clearly in fig. 1, Pl. 24; they are fused to one another laterally except at their distal ends, and are broader in the upper half of the otolith than in the lower half. Growth in length and depth of the otolith is accomplished by a branching of the lobes, several of which may be seen dividing, in the otoliths in fig. 1, Pl. 24, and fig. 5, Pl. 25.

A fresh hake otolith, owing to its flattened shape, is very fairly translucent, and, when held against a source of light, is seen to have a discontinuous structure, and to be divided into broad, less translucent zones, separated by narrow, more translucent rings, which are strictly parallel to the margin of the otolith. When placed against a dark background, these more translucent rings appear black, the less translucent areas white, as in fig. 1, Pl. 24, and fig. 5, Pl. 25. The rings are probably not actually continuous, but, since they occur at about the same relative level in all the lobes of which the otolith is built up, they present the appearance of complete rings parallel to the margin. These rings can be seen very clearly in fig. 1, Pl. 24, and fig. 5, Pl. 25. Their similar course in the lobes of which the plaice otolith is built may be seen in Immermann’s fig. 6. As may be seen in fig. 5, Pl. 25, in the present paper, that part of the ring contained in an individual lobe is parallel to the curved border of that lobe.

The internal structure may be investigated by the examination of portions of otoliths rendered transparent by prolonged soaking in, for example, pyridine, or, better, by the cutting and grinding of thin sections. A thin transverse section of the uppermost otolith in fig. 1, PL 24, is shown in fig. 2, Pl. 24, orientated in its natural position. The Sulcus Acusticus is seen as a conspicuous indentation on the inner convex surface (to the right in fig. 2, Pl. 24), and on the same surface may be seen slight indentations, indicated by arrows, which are the concentric grooves on the surface mentioned in Section A. These grooves may also be faintly discerned on the opposite outer concave surface.

The most conspicuous feature in the section is a system of concentric lamellae, extending from the centre of the otolith to the periphery. The lamellae become thick where they cut the mid-sagittal plane of the otolith, becoming thin, almost to invisibility, in the mid-horizontal plane. The lamellae are not so clearly marked in the lower half of the section, because of the fractures about to be described.

The otolith of a small hake is slightly less flattened than that of a larger hake, but the general shape is unchanged during growth, and the otolith therefore grows more rapidly in length and depth than in breadth. Cunningham (1904) describes the similarly shaped lamellae in the plaice otolith as follows: ¹ each successive layer extends over the whole surface ‘(of a transverse section) ‘but is exceedingly thin on the two flat surfaces, and thicker at the edge. The structure is such as would be produced if a sphere composed of concentric uniform layers of plastic material were very much compressed so as to form a flat disk. The thin layers on the two faces being translucent, the surfaces of contact between successive layers are seen as lines approximately parallel to the outer edge’.

Immermann shows that an otolith comprises two constituents, namely, an inorganic crystalline constituent, and an organic fibrous constituent, which is a metamorphosed portion of the gelatinous connective tissue which fills the cavity of the sacculus. The present investigation shows that, of the two, the organic constituent is much the more important. They will now be described, as they occur in the hake otolith.

1. The Organic Constituent.—When an entire otolith is carefully demineralized in weak acid, or, better, in Fol’s solution, as formulated by Immermann, the organic constituent is left behind as a transparent, gelatinous and exceedingly brittle structure, which, though always much shrunken and distorted, has the identical lobed shape of the entire otolith. Moreover, the concentric shells (lamellae) described above can be discerned in the demineralized otolith, and are thus seen to belong to the organic constituent thereof. Unfortunately, as Immermann himself found, even with very gradual demineralization, gas-bubbles form within the otolith, and tear the very frail organic basis, which, also, always shrinks. In a demineralized otolith, therefore, the concentric shells are no longer arranged in the regular manner characteristic of the entire otolith.

The internal structure of the organic constituent may be investigated by the microscopic examination, either of demineralized sections of the otolith, or of serial sections cut from a previously demineralized otolith. The latter are preferable, because they may be cut into much thinner sections than those into which an entire otolith may be ground. On the other hand, it is exceedingly difficult to cut into serial sections a tissue so fragile as the organic constituent of the otolith, and the sections, when cut, break up very easily. Immermann’s own photographs, in his figs. 3 and 4, of sections of the organic substance of a plaice otolith, bear eloquent witness of this difficulty! However, after many attempts, serial sections have been successfully cut, and stained with saffranin, as Immermann recommends.

A photograph of a portion of a section of the organic basis of a hake-otolith, cut transversely to the long axis of the otolith, and stained with saffranin, is shown in fig. 3, Pl. 24.

Three groups of concentric lamellae, cut transversely, are plainly visible, namely, at the top and middle of the photograph (C.L.) and at the bottom of the photograph. These are the thicker lamellae. Thinner lamellae are very faintly to be discerned, in the intervals between the thicker lamellae, as shadows or thin streaks. But the most conspicuous feature of the intervals between the thicker lamellae is the comparatively stout radial fibres (R.F.) which run from the centre of the otolith towards the periphery, and which pass, without interruption, through the concentric lamellae. In places (as to the right of the section) they are torn apart, probably by gas-formation during demineralization.

Under a very high magnification, the concentric lamellae tend to appear as rows of dots, as though they were themselves built up of a network of fibres running in a plane normal to that of the stouter radial fibres. Even under a lower magnification, as in fig. 3, Pl. 24, the concentric lamellae (C.L.) give this impression. The organic basis of the otolith seems to contain no structures other than the comparatively stout radial fibres, and the concentric shells, of varying thickness, which the radial fibres bind to one another.

As I have stated above, owing to shrinkage and distortion, a demineralized otolith is unsuitable for demonstrating the true spatial relations of the concentric shells to one another, or for showing how they vary in thickness. But they are very conspicuous in the entire otolith, as already described; and their true proportions may easily be shown in a thin section of an entire otolith. Such a section is shown, under high magnification, in fig. 4, Pl. 25. The photograph is one of that part of the section, shown in fig. 2, Pl. 24, which is marked with a cross.

In this photograph, which is taken near the vertical middle line of the otolith, we see, in the upper portion, a region of comparatively thick concentric lamellae (thick C.L., marked with an arrow), in the middle, a region of extremely thin lamellae (thin C.L., marked with an arrow) which are just discernible in the photograph, and, finally, at the bottom left-hand corner, the commencement of a fresh region of thick lamellae. The lamellae, while varying so notably in thickness, appear to be spaced out at about equal intervals from one another, this distance being about 2μ.. The thickest lamellae have a maximum thickness of about l·5μ, the thinnest lamellae are almost ultra-microscopic, especially in those regions of the otolith in which their course runs parallel to the flat surfaces of the otolith, and where, therefore, as described above, they thin out. The photograph in fig. 3, Pl. 24, is of a region where the lamellae are nearly at their thinnest in their course around the otolith. As will be described below (Section C), however, the thin lamellae do not have quite the same shape as the thick lamellae, in that they do not thicken out at the vertical middle line to the same extent as the thick lamellae.

Fig. 4, Pl. 25, also gives some support to the hypothesis that the concentric lamellae are themselves compounded of a network of extremely fine fibres. If fig. 4, Pl. 25, is examined minutely, the thin lamellae, especially in the lower centre of the photograph, give the impression of being, really, rows of closely set fine dots, while the thick lamellae, especially in the upper centre of the photograph (about at the level of the upper arrow) tend to appear as rows of fine streaks. Presumably, in the latter case, the fine fibres are cut somewhat obliquely, for, owing to the curved course (parallel to the curved margin of the lobe) which the lamellae take in the lobes of which the otolith is built up (as described in Section A) any section of the otolith, which is not infinitely thin, will contain some or all of the lamellae cut obliquely.

The stout radial fibres, owing to the fact that they run parallel to the crystals which comprise the inorganic constituent of the otolith, are not to be seen in fig. 4, Pl. 25, unless the flecked or dappled radial streaks which are present everywhere in the photograph (and which are carefully to be distinguished from the fissures) are caused by them.

1. The Inorganic Constituent.—Immermann states that the inorganic constituent of a plaice otolith consists of needles of calcium carbonate, arranged in systems, and that a new system of needles seems to commence at each lamella. Maier describes the crystals in the otolith of the cod as fine little needle-crystals, and, in his fig. 9, he draws them as sharply pointed but not very slender.

Owing to the closeness with which the crystals interlock, it is exceedingly difficult to trace the outlines of any one crystal in the hake’s otolith. It was soon realized that they must be minute, and therefore difficult to see for this reason also. If, however, the surface of a section of a hake otolith be etched for a short time with a demineralizing solution, small islets of crystals, which have been less affected by the reagent than others, are left standing out, and, by suitable focusing, some idea of the appearance and dimensions of the crystals in these islets may be obtained.

They are sharply pointed and exceedingly slender. If they are, indeed, entire crystals, they have a maximum length of about 40μ, and a width of less than 1 μ.

Though the crystals are themselves so difficult to see, their arrangement in the otolith can be followed very easily by the course of the radial fissures which appear in every otolith section. They are to be seen among the solidly interlocking crystals clearly in fig. 2, Pl. 24, and are undoubtedly lines of cleavage produced by grinding. They are also well shown in fig. 4, Pl. 25 (Fissures). The fissures arise at the vertical middle line, and run out to the periphery in such a way as to cut through the concentric lamellae at right angles to them. The result of this is that those fissures which arise at the vertical middle line in the central region of the otolith run almost straight to the periphery, while those arising towards the upper and lower edges follow an increasingly curved course (fig. 2, Pl. 24). The two most important points about their arrangement in the otolith are, firstly, that they run through the lamellae normally to them, and, therefore, parallel to the radial fibres, and, secondly, that they are not in any way interrupted at the lamellae_s but pass through them, nor is there any tendency for the lines of cleavage to arise at any particular lamella.

My views as to the microscopic structure of the hake otolith may be summed up as follows, and are illustrated diagram-matically in Text-fig. 1.

The basis is a very frail basket work, consisting of sheets or networks of fibres, which form concentric shells (figs. 3, Pl. 24, and fig. 4, Pl. 25), bound together by stouter radial fibres (fig. 3, PL 24) and supported by needle-like crystals of calcium carbonate, which are orientated normal to the concentric lamellae and parallel to the radial fibres, and which interlock to form a very solid structure. In Text-fig. 1, the lefthand portion of the figure shows the columns of needle-like crystals, among which we must presume the radial fibres to run; the right-hand portion shows the same, demineralized, showing the radial fibres exposed by the removal of the crystals, and rather shrunken. Crystals and fibres run normal to, and pass without interruption through, the concentric lamellae (C.L.), of which groups of thick lamellae are found in the upper and lower portions of the figure, groups of thin lamellae in the central portion. The lamellae are represented as rows of fibres passing between the crystals, in a plane normal to them, and therefore, cut in transverse section, appearing as dots.

One reason for the discontinuous or fibrillar structure of the concentric lamellae seems to be as follows. The concentric lamellae are perfectly definite structures, independent of the crystals, as any demineralized portion or section of an otolith, such as fig. 3, Pl. 24, clearly shows. The crystals are, also, perfectly definite structures, which are quite independent of the lamellae, running normally through them without interruption. It follows, therefore, that the lamellae, which obviously cannot pass through a crystal, must be interrupted in order to pass on either side of the crystal. I have endeavoured to show this, in Text-fig. 1, by placing dots, which represent the concentric lamellae, on the boundary lines between the crystals.

These views on the structure of the otolith are in good agreement, on the whole, with Immermann’s. But Immermann did not observe the reticulate or fibrous structure of the concentric lamellae, whereas Maier, who observed neither the radial fibres nor the concentric lamellae in the otolith of the cod, describes, as layers filled with dark opaque granules, what are obviously the concentric lamellae, and he draws them as such in his fig. 9. He seems to have missed the very fine lamellae which, by analogy with the hake, occur between the thicker lamellae, but his observation clearly supports mine; and, further, Maier draws his crystals as running straight through the lamellae, without interruption, wherein, again, he supports my observation as against Immermann’s.

The structure of the hake otolith, which is an ectodermal secretion, seems to be strikingly similar to that of another ectodermal secretion, namely, the nacreous layer in the molluscan shell, and the pearl. Lyster Jameson (1912) describes the nacre of the Ceylon pearl-oyster as follows: ‘The organic basis which gives it its form, and which retains its iridescence after the calcareous salts have been extracted, consists of a series of parallel lamellae, of extreme fineness, united to one another at intervals by radial connexions, so as to form a series of minute flat or lenticular chambers, separated by organic walls of extreme delicacy. The calcium carbonate appears to be enclosed in these chambers…. This structure is difficult to observe, owing to the distorting effect of the decalcification process, which, owing to the evolution of gas-bubbles, tears some lamellae apart, and forces others tightly together.’ Amarthalingam (1929) publishes a photograph of a decalcified pearl, in which the concentric lamellae are clearly visible. Moreover, he very kindly sent me his preparations of pearl sections, which I was able to examine, and he has allowed me to state that, in the pearls of Pinna, at least, the fibres and crystals are arranged in a manner similar to that in the otolith of the hake.

The rings in the otolith of the hake have been described in Section A, and are portrayed in fig. 1, Pl. 24, and fig. 5, Pl. 25. In fig. 5, Pl. 25, are shown enlarged views of two otoliths, and beneath each is a photograph of a transverse section cut from it. A comparison of the sections with the entire otoliths shows that the more translucent rings, here appearing dark, correspond, exactly, with clear zones among the groups of concentric lamellae visible in the sections. This is confirmed by the examination of many other sections, not reproduced here. In my opinion, the rings visible in the entire otolith owe their greater translucency, undoubtedly, to the presence of these clear zones, which allow of an easier passage of light through the thickness of the otolith at the points where they meet the middle vertical line. The cause of the ringed structure, therefore, is to be found in the cause of the clear zones seen in the transverse section.

The fine concentric lamellae described, in Section B, as visible under very high magnification, tend to occur in groups, according to thickness, a group of thick lamellae being separated from the next group of thick lamellae by a few very thin lamellae. The groups of thick lamellae themselves form, as it were, compound lamellae, which are plainly visible under low magnification—they can easily be seen in fig. 2, Pl. 24. The compound lamellae are, typically, in turn arranged in groups, such that there are regions where thick lamellae predominate, and but few thin lamellae are present, separated by regions where thin lamellae predominate, and thick lamellae are few or absent.

As has been stated, fig. 4, Pl. 25, is a photograph of the region marked X in fig. 2, Pl. 24, under high magnification. A region of thin lamellae occupies the centre of the field, while thicker lamellae occupy the upper and lower parts of the field. The region of thin lamellae appears as a transparent zone under a lower power of magnification (fig. 2, Pl. 24), while the thicker lamellae belong to the relatively opaque regions separated by the transparent zone.

The opacity produced by the groups of thick lamellae is comparable to that of the fat globules in milk, which, though they are themselves individually transparent, produce, in the mass, the effect of opacity and whiteness to the naked eye.

The distinctness of the rings seen in the entire otolith depends upon the sharpness of the grouping of the thick and thin lamellae. In a considerable proportion of otoliths, such sharpness is lacking in a part or the whole of the otolith, and an assessment of the number of the presumed ‘annual rings of growth’ becomes difficult or impossible.

Moreover, if the otolith is allowed to remain dry for any length of time, it becomes much more opaque, and translucency cannot then be restored by moistening. It is impossible to say whether this is due to an efflorescence of the crystals, or to a withering and shrinkage of the lamellae and fibres, or to the penetration of air bubbles. The plaice otolith, owing to its much smaller size, is still easily legible in such circumstances, but the bulkier hake otolith becomes almost useless.

Finally, the thin lamellae do not appear to thicken at the middle vertical fine, at least as much as the thick lamellae. A succession of thin lamellae, therefore, leaves the edges of the otolith relatively more rounded than a succession of thick lamellae; there is a slight change in the shape of the otolith as a whole. This may be seen distinctly in fig. 2, Pl. 24, where the transparent zones are notably more rounded at the middle vertical line than the opaque zones. When the region of thin lamellae becomes covered by a succession of thick lamellae, the slight change of shape causes a concave curvature of the succeeding lamellae, which is responsible for the grooves seen on the surface of the otolith. The grooves occur between the transparent zones, opposite the opaque regions of thick lamellae; but, if the course of growth is followed (in fig. 2, Pl. 24) by tracing the course of the fractures between the crystals from the periphery towards the centre of the section, the surface grooves will be found to correspond, in position, with the transparent zones.

Immermann considers that the lamellae have no part in causing the ringed appearance of the plaice otolith. He considers, however, that they cause, by some internal strain, a bending or twisting of the crystals. The plane of the crystals being slightly altered at the lamellae, there is a differential refraction and internal reflection of the light, causing zones of dark shadow. I cannot find, in the hake otolith, any sign of interruption of the crystals at the lamellae. The crystals, indeed, run straight through the lamellae from the middle vertical line to the periphery, as may be seen by following the course of the fractures in fig. 2, Pl. 24, and fig. 4, Pl. 25, without any bending or twisting. In any case, it is hard to imagine how any effect on the crystals could be produced by lamellae whose distance from each other appears to be not much greater than the width of the individual crystals, and much less than their length. An examination of sections of hake otoliths under polarized fight, between crossed nicols, shows that the crystals are either continuous in their course through the lamellae, or are, at least, optically parallel.

Immermann supports his hypothesis, that the ringed structure is due to the inorganic rather than the organic constituent, by stating that the appearance of the otolith is not affected by the removal of the latter. He recommends two methods for removing the organic constituent, namely, gently warming in alkali, and heating. I have repeated Immermann’s experiments, made on plaice otoliths, with hake otoliths, and my results do not confirm his.

Small entire hake otoliths were warmed for months in strong caustic soda, frequently changed. The otolith was unaffected, except on the very surface, and it was clear that the penetration of the alkali was but slight. A thin section of an otolith, about 100μ thick, was therefore warmed in strong caustic soda for three months. The ringed structure was effaced, at least from the surface layers. Treatment with dilute acid removed the layers affected, and the rings once more became apparent. That the alkali, presumably by removing the organic constituent, destroyed the ringed structure, cannot be stressed, because the inorganic crystals were themselves seriously corroded by such drastic treatment.

The effect of heat has been tried on a great number both of entire otoliths and of sections. When an otolith is heated, there is, at first, a blackening, due to the charring of the organic matter, and in order to remove this completely it is necessary to use dull red heat. The otolith is thereby converted into an opaque, friable mass of lime, retaining the original shape, and even having the surface grooves unaffected. This mass is too brittle to grind down into sections, but if it be broken across, and the fractured surface be examined by reflected light, the original crystalline structure is seen to be represented by parallel columns of lime; but there is no sign of concentric rings, such as are easily seen in an untreated otolith examined in the same way, nor are there any signs of cracking or discontinuity among the columns of lime, corresponding to the concentric rings of the original otolith, such as Immermann claims to have found in the calcined plaice otolith. The result of this experiment is definitely adverse to Immermann’s hypothesis.

It is true that the regularly zoned structure is destroyed by demineralization, but this would certainly result from the shrinkage and distortion of the very delicate organic basis of the otolith. As long as any crystals remain to support the lamellae, the typical structure persists.

Photomicrographs taken with Edinger’s Projection Apparatus (by Leitz).

C.L. Concentric lamellae. R.F. Radial Fibres.

PLATE 24.

Fig. 1.—Five hake otoliths, photographed by reflected light against a dark background, × circa 2.

Fig. 2.—Section of the uppermost otolith in fig. 1, cut transversely to the long axis of the otolith, and ground thin. The arrows mark the surfacegrooves described in the text, × 18.

Fig. 3.—Portion of a section of a demineralized hake otolith (the organic constituent) cut transversely to the long axis of the otolith, and stained lightly with saSranin. Bausch and Lomb 4 mm. ocular, no. 5 eye-piece, using Wrattin B light-filter, × 313.

PLATE 25.

Fig. 4.—Portion of the transverse section of a hake otolith shown in fig. 2, in the region marked X, highly magnified. The conspicuous radial furrows represent lines of cleavage among the crystals. Zeiss oilimmersion objective, × 410.

Fig. 5.—The anterior portions of two hake otoliths, photographed together with transverse sections cut from them. Otoliths and sections are reduced to the same magnification.