Book - Contributions to Embryology Carnegie Institution No.20

The histogenesis and growth of the otic capsule and its contained periotic tissue-spaces in the human embryo

By George L. Streeter

4 text-figures and 4 plates

Contents

1. Introduction 7-8
2. Terminology 8-10
3. Historical 10-17
4. Material and Methods 17-19
5. Development of cartilaginous capsule of ear 19-31
6. Condensation of periotic mesenchyme 20-22
7. Differentiation of precartilage 22-24
8. Differentiation of cartilage 24-26
9. Growth and alteration of form of cartilaginous canals 26-31
10. Development of the periotic reticular connective tissue 31-34
11. Development of the perichondrium 34-37
12. Development of the periotic tissue-spaces 37-48
13. Development of the periotic cistern of the vestibule 38-40
14. Development of the periotic spaces of the semicircular ducts 40-41
15. Development of the scala tympani and scala vestibuli 41-46
16. Communication with subarachnoid spaces ' 47-48
17. Summary 48-.')0
18. Bibliography 51
19. Explanation of plates 52-54

Introduction

During the past year the writer has published two shorter communications regarding the development of the connective tissue and cartilaginous coverings that inclose the membranous labyrinth, one of which dealt with the histogenesis of the cartilaginous capsule and the other with the periotic tissue-spaces that are formed in the interval between the cartilaginous capsule and the membranous labyrinth. In the present pajjer the same matter will be treated in more complete form and a general description will be given of the development of the otic capsule as a whole and of the problems involved in its growth.

In making this studj^ the effort has not been so much toward the determination of the exact form of the capsule as it has been toward the detection of .some of the factors that are involved in the production of this form. These two problems, however, are not to be altogether .separated. It is the distinctive form of the otic capsule that makes it a i^articularly favorable place for determining the histological features of the growth of such a structure. Owing to the fact that it is so well provided with known landmarks, the changes in its size and form can be accurately followed and it is therefore possible to determine deductively at what points, for instance, new cartilage is being laid down and at what jjoints it is being removed.

It was soon recognized that the growth of the otic capsule resolves itself into an increase in its external dimensions and a simultaneous hoUowing-out and reshaping of its contained cavities, the latter being so managed that their general form and proportions are continuously maintained and a suitable space always provided for the enlarging membranous labyrinth. It is particularly the feature of cartilage excavation accompanying the increase in the total mass to which attention will be invited. It is quite evident that such growth can not be explained on the basis of a simple interstitial increase in the amount of cartilage, together with its passiA'e rearrangement to allow for the enlarging cavities, due, one might say, to a mechanical expansive pressure from the growing membranous labj'rinth with its surrounding tissue and fluid. Such a passive rearrangement could only occur in a tissue that is very plastic, whereas cartilage is one of the least plastic of the embryonic tissues. Moreover, the histological picture is not that of mechanical pressure; the cartilagi- nous chambers are always excavated slightly in advance of the space actually required by the membranous labyrinth, and there is no evidence of the labyrinth being cramped or of the creation of pressure grooves in the margin of the cartilage.

Furthermore, it can not be the perichondrium that is the essential factor, either in the deposit of new cartilage or in the excavation of the old, because the perichon- drium, as we shall see, is not formed until after a considerable amount of the growth and hullowing-out of the labyrinth is already completed. Therefore, in the devel- opment of the cartilaginous capsule there is something more than interstitial and perichondrial growth.

As forming at least one element, and an important one, in this process it has been found that there occurs a regression of certain areas of cartilaginous tissue to a more embryonic form followed by its alteration into a different type of tissue. It is this process of dedifferentiation that constitutes the essential factor in the hoUowing- out and reshaping of the otic capsule which take place continuously during its development. Though the significance and wide occurrence of dedifferentiation and redifferentiation have been well known to botanists and to those investigators who have worked with the simpler forms of animal life, this, as far as the writer knows, is the first time that they have been shown to occur in the human embryo. It is not unhkely that these principles will eventually enter into our conception of the growth of other tissues and organs in human as well as in other mammalian embryos. The establishment of this point, of the occurrence of retrogressive as well as progressive differentiation in human embryos, is considered by the writer to be the chief contribution of the following paper.

The fate of the periotic connective tissue that intervenes between the cartilage and the membranous labyrinth and the formation of the characteristic periotic spaces form problems that are naturally of a morphological character. These spaces have been studied by modeling methods and a description will be given of the steps by which the larger spaces acquire their adult form. It will be pointed out that these spaces show a marked individuality. They have constant and definite charac- teristics, including their time and point of origin, the manner in which thej^ spread, and their eventual form and structure. They have a structural individuality which, though less complicated, is just as definite as that of the other parts of this sense- mt'clianism. All of this we will come to later.

Terminology

The writer is not unmindful of a certain feeling of distress that is aroused when it is found on reading a new paper that the author of it is adding to the already difficult matter of following another's description by making a new application of terms or by introducing a whole battery of freshly created ones. Nomenclature constitutes one field in which rock-bound conservatism lias many points of merit and where oiiginality may expect a cold and critical rcH'eption. It is therefore with some embarrassment that the writer api)roaches the subject of terminology, and it is al.so with some apprehension as to whether the " originality " in this instance will i)rove to be justified. It has in fact seemed best to avoid the incorporation of the term "lymphatic" in describing of the tissue-spaces surrounding the mem- branous labyrinth. It has been the custom to designate these as "perilymphatic" spaces since 1833, when the term was introduced by Brescliet, who thus distinguished them from the "endolymphatic" cavities of the membranous labyrinth. These terms, together with the terms "perilymph" and "endolymph" for their contained fluids, seemed particularly appropriate and in practical use have proved to be very convenient. Since Breschet's time, however, the lymphatic vascular system has taken on an increased and individual importance, due to researches in which Amer- ican investigators have taken a particularly active part, and it now seems impor- tant to restrict the term "lymphatic" to it and its associated structures.

Inasmuch as the tissue-spaces surrounding the labyrinth have no known con- nection with the true lymphatic sj'stem, either in their origin or in their ultimate relations, it follows that the use of the term "Ij-mphatic" in connection with them is misleading. It therefore seems advisable to ehminate it, even at the expense of losing such a convenient terminology. As a substitute for " perilj'^mphatic " the term "periotic" was finally decided upon and will be so used throughout this paper. In the formation of this adjective the Greek word ovs, from which it is derived, is used in the restricted sense of representing the essential sense-organ, that is, the otocyst itself and eventually the membranous labyrinth. Inasmuch as numerous words derived from the same source are in commtm use, it is felt that this term will be readil}^ understood.

We shall speak of a periotic connective tissue that everywhere surrounds the membranous labyrinth. This periotic connective tissue includes in part the fine- meshed periotic reticulum, and in part the large walled-off periotic spaces with their contained periotic fluid, the most prominent of which are the scala vestibuli, scala tympani, and the vestibular cistern. For the term " endolj'mphatic " fluid one could substitute "otic" fluid; we would then have "Uquor perioticus" and "Hquor oticus." In all other instances, except when elsewhere specified, the Basillensis Nomina Anatomica terms have been adhered to. The term "semicircular duct" is used to specify the epithelial or membranous canal as distinguished from the carti- laginous semicircular canal. This usage was recommended by Breschet and was adopted in the BNA. It was not taken advantage of, however, by Retzius (1884) in his monograph on the vertebrate ear, who used the term "semicircular canal" for the epithelial channel as well as for the cavity in which it lies. The influence of this great monograph has delayed somewhat the adoption of the BXA recommendation, and one finds subsequent writers still following Retzius in this respect, among whom may be mentioned v. Ebner. R. Krau.se, Rothig, and myself in previous papers on the development of the membranous labyrinth. In a similar manner the usage by Retzius of the term "anterior" canal instead of "superior" canal, as recommended by the BNA, has occurred in relatively recent papers, including, it must be con- fessed, those of my own. In the present paper, however, this usage has been cor- rected. In the historical review which follows the various structures mentioned will be largely referred to in the older terms used by the respective authors.

Historical

The first monographic treatise on the anatomy and physiology of the ear was that published by Breschet in 1833. This work proved to be a very important one, both as regards the new observations contained in it and the constructive manner in wliit'h the facts then known were analyzed. The terminology of the ear region was .â– standardized and most of the names that were used or introduced by Breschet are in use in the literature of to-day. Before stating his views concerning the structures with which we are dealing, reference must be made to the work of some of his prede- cessors, and this will be given essentially as outlined by him.

The early anatomists were familiar with the bonj' labyrinth, but supposed that the spaces contained within it were filled with air. In 1707, however, Valsalva described the normal presence of a fluid in the labyrinth Avhich he compared to the fluid seen in serous cavities. The presence of this fluid was confirmed by Vieussens (1714). His observations were made chiefly on new-born infants, in which he studied the distribution of the labyrinthine fluid and found it present in the vestibule, the cochlea, and in the semicircular canals. The same fluid was also referred to by C'assebohm (1735) and Morgagni (1740). Uptothattinie no author had directed any particular attention to the labyrinthine fluid, nor had anyone attempted to assign any function to it other than that of moistening the auditory nerve. It was Cotugno (1768) who first endeavored to show that the labjTinthine fluid had some connection with the transmission of sound vibrations. He maintained that there was no air in the spaces of the labjTinth, but that it was everywhere filled with the fluid, which according to his description exudes from the ends of the capillary arteries that are distributed throughout the membrane that lines the cavity of the labyrinth. He described the fluid as being drained off by means of the two aqueducts. Because of the completeness of his description and the interest which he attracted toward the subject, the labyrinthine fluid was thereafter known for more than half of a century as Cotugno's fluid.

Any further advance regarding the nature of the labyrinthine fluid required a more detailed knowledge of the soft parts of the labyrinth. Nerve-like cords and semicircular tubes had been seen in the canals and membranous partitions and sacs had been seen in the vestibule, but it remained for Scarpa (1789) to establish the identity of the membranous labyrinth. He showed that in man and other mam- mals the semicircular tubes and the vestibular sacs are of the same nature and form one sj'stem, and that they are distinct from periosteum. He described how they open freely into each other and are filled with a limjiid fluid which distends them. This fluid was thereafter referred to as the fluid of Scarpa. Me recognized it in a general way as di.stinct from the lal)yrinthine fluid, in which all parts of the mem- branous labyrinth floated, but otherwise grants it no further attention.

The relations and significance of the fluid of Cotugno and the fluid of Scarpa were not completely recognized until the jniblication of the noteworthy monograph of Breschet (1833) of which we have spoken. He introduced the terms perilymph and endolymph, by which they have since been known. The existence of the coch- lear duct was unknown to Breschet, but otherwise his description of the labyrinth spaces and their contained fluids is the foundation on which the more recent descriptions are principally based. He showed that the perilymph occupies all the space of the bony labyrinth that is not taken up by the semicircular tubes, the utricle (median sinus), and the saccule. It surrounds these everywhere and separates them from the bony walls of the labyrinth. The perilymph also, according to him, fills the spaces of the cochlea and circulates freely throughout the whole system. The scala tympani is connected at its apical extremity with the scala vestibuli by means of the opening to which he gave the name helicotrema. The scala vestibuli in turn opens freely into the vestibule, into which also the semicircular tubes open. He points out the fact, and discusses its relation to the mechanism of hearing, that any vibrations transmitted to the perilymph by the foot-plate of the stapes would be transmitted freely and evenly to the whole of the membranous labjTinth and to the lamina spiralis. He describes the perilymph as consisting of a thin, watery, sahne fluid containing a small amount of albumin. He believed that it was secreted by the thin, delicate membrane lining the cavity of the labyrinth and that the materials were brought there by the small blood-vessels that .supply this laj'er. The aqueducts, according to him, are not for the transmission of perilymph, but only serve as passages for veins. He regards them of embryological significance; that they represent the remaining strands of connection with the dura mater, of which the inner ear is a part that has been separated off by the enveloping growth of bone. A description is given of the distribution of perilymph in different animals and it is pointed out that in some fish it communicates with the fluid surrounding the central nervous system and how in such cases the oily or gelatinous cerebro-spinal fluid actually serves as the perilymj^h. Entirely distinct from the perilymph is the endolymph, which is the fluid filling all parts of the membranous labyrinth.

Breschet describes the character of the endolymph in different animals. He shows that it always contains calcareous deposits, which he designates as otohths and otoconia, depending on whether they are in the form of lumps or dust. He shows that these are distributed at definite places, at the points of nerve-terminations, and suggests that they act as dampers that tend to check the prolonged vibration of the endolymph. In comparing the ear with the eye, he suggests that the perilymph bears the same relation to the organ of hearing that the aqueous humor bears to the organ of vision. The vitreous body he considers analogous to the endolymph.

Special emphasis has been given to the treatise of Breschet because it marks the beginning of the modern epoch in the anatomy of the ear. Previously the descrip- tions of this organ had been purely fragmentary. Breschet's monograph is both comprehensive and analytic. If his treatise is searched for defects that are revealed in the light of our present knowledge of the anatomy of the ear, one would name perhaps only two major ones. One of these concerns the inaccurate and meager nature of the embryological features as given by him, and the other concerns the membranous cochlea, the existence of which was entirely unknown at that time. It is interesting to note that these two defects prove to be related; that it was through embryolcp-ical investigations that the membranous cochlea was eventually discovered.

At about the time of Breschet's treatise, Huschke (1831) made the discovery that the membranous labvrinth was originally a pit in the skin, a fundamental point that was confirmed later by other embryologists. He also found that in sheep and calf embryos the lamina spiralis is hollow, constituting a spiral tube that is closely attached to the bony walls of the cochlea. He evidently had before him the ductus cochlearis. He supposed, however, that the scalae in their formation flattened out this liollow tube, thereby converting it into the lamina spiralis, and thus he did not grasp the meaning of the structure and just missed being the discoverer of the mem- branous cochlea.

The significance of the spiral tube seen by Huschke in the embryo and its per- sistence as the ductus cochlearis in the adult remained to be pointed out by Reissner and Reichert in a series of communications published in the years 1851 to 1854, being based in large part on embryological studies of the chick and also of mammals. The first communication was the Dorpat dissertation of the former, completed in Reichert "s laboratory'. It contained an account of the ductus cochlearis (canalis cochlearis), which was shown to exist as a definite canal in the adult mammalian cochlea. It was pointed out that the membranous part of the lamina spiralis forms one-half of the wall of the canal and that the other half consists of a very thin and delicate membrane that is usually torn in the preparation of such a specimen. This portion has since been known as the membrane of Reissner.

Rei.ssner and Reichert demonstrated the complete canal in infants at about term. In the embryo they found that the cochlear duct opens into the membranous vestibule, but whether or not it does this in the adult was not definitely determined. They give an account of the development of the ear in the chick, and they describe the formation of a labyrinthine groove in the skin and how this subsequently in vagi- nates to become the otic vesicle to which the acoustic nerve attaches itself later. They divide the early labyrinth into three chief ])arts: (1) recessus labyrinthi; (2) \estibule and its three canals; (3) cochlear duct. In a later paper Reissner (1854) refers to the formation of the scalse. He explains them as two accessory cavities that are formed because the cartilage recedes from the upper and lower surfaces of the cochlear duct. He denies the existence of anj'^ communication between the vestibular cistern (Hohle des Vorhofs) and the scala vestibuli both in the embr3^o and adult. His observations concerning the scalae have proved to be less important than those on the embryology of the otic vesicle and the discoverj^ of the cochlear duct.

Kolliker (1854) had just at this time written the first edition of his"Gewebe- lehre" and had described at some length the finer structure of the cochlea, embody- ing his own and the important observations of Corti (1851). He makes mention of Rei.ssner's dis.sertation, but there is no evidence that he appreciated at that time the .significance of the cochlear duct. In a .sei)arate paper and in his text-book on human embryology, both of which appeared soon after this (Kolliker, 1801 a and b), he definitely establishes the existence of Reissner's membrane and that it forms the boundary of the cochlear canal. He confirmed the Reissner and Reichert embrj'- ological studies showing that the cochlear canal is originally an epithelial tube which is derived from the primitive ear-vesicle and hence from the ectoderm of the embryo. He designates the cochlear canal as the .scala media, a term which persisted for many years, though its inapplicability was promptly ])ointed out by Reichert (18G4).

Kolliker is the first to describe in some detail the formation of the otic capsule and the perilymphatic spaces. The summarj- here given is taken from the second edition of his book (Kolhker, 1879), in which there is some amplification of the account given in the first edition. According to him, after the otic vesicle reaches a certain degree of development it is surrounded by a delicate connective-tissue membrane and there is an outer thicker and firmer mass which takes on the nature of cartilage, which in 19-mm. cow embrj'os forms an integral part of the wall of the skull. In human embryos 8 weeks old the labyrinth capsule consists of true cartilage and completely fuses with the base of the skull. He expresses the opinion that the cartilago petrosa is laid down exactly in the same way as the other parts of the lateral walls of the skull and that its special characteristics, subsequently assumed, are due to the presence of the special sense-organ.

In connection wath the origin of the cavities of the bony labyrinth, Kolliker draws attention to the problem of space formation in general and points out that the space in the otic capsule is of the tj^pe seen in the subarachnoid and other serous cavities. He describes how, along with the growth of the epithelial part of the labyrinth, there is also a rapid growth of its connective-tissue co^•erings, which soon attain considerable thickness. At the same time the periotic tissue becomes differ- entiated into three layers, of which the middle one soon becomes the thickest. This layer consists of a network of anastomosing connective-tissue cells (Gallertgewebe), whose rounded meshes are filled with fluid. From this there is graduallj^ formed the cavity that surrounds the semicircular canals, the meshes becoming larger and finally coalescing. In the process of coalescence, parts of the cellular net are broken and other parts are pressed against the walls of the space, where, even in the adult, one can recognize remnants of the broken net. The same process takes place in the semicircular canals, the vestibule, and the cochlea. In the last-mentioned there are formed the two scalae, in which, in addition to the coalescence of the spaces of the network, there is also involved a disproportionate growth in respect to the cochlear duct and the surrounding cartilage, the latter retracting from the former by virtue of its rapid growth. KoUiker's conception of this process is purelj^ mechanical, and it hypothesizes a protoplasmic network that is entirely passive. He did not con- ceive of an adaptive activity on the part of the protoplasm itself by virtue of which the characteristic changes in form are brought about, as is to be described in the present paper.

Among a series of miscellaneous notes concerning the development of the mam- mahan labyrinth appended at the end of KoUiker's chapter, there is a reference as to the relation of the growth of the cartilage to the growth of the contained spaces, which is particularly interesting, as it shows that he had in mind one aspect of the problem with which we are concerned. He notes there (p. 746) that at first the epitheUal part is directh' surrounded by young cartilage, or better, a cartilage-Uke substance of which the greater part becomes subsequently converted into cartilage. In its further differentiation the tissue lying directly against the labyrinth becomes fibrous tissue and the tissue farther away becomes cartilage. Out of the uniform fibrous tissue there is further differentiated the inner perichondrium, the fiber wall of the labyrinth, and the cell tissue intervening between these. This intervening tissue untlergoes an independent growth characterized by a rich growth of blood-vessels. As a space, i. e., the cavity of the cartilago petrosa, it grows in correspondence to the growth of the epithelial labyrinth. As to the behavior of the cartilage during the growth Kolliker was undecided, giving the opinion, however, that it grew independently at the same time as the space, and was not simply mechanically stretched out.

The embryological studies of His began to appear at about the time of the pub- lication of Kolliker's work on the ear, and one would rather expect that the attention of this keen observer would have been attracted to this subject. In his "Akade- mische Programme" of the year 1865 he outlines (His, 1903) the general problem of the formation of the various body-cavities and describes in detail the formation of the cavities of the middle germ-layer. He includes in this the arachnoid spaces and the cavities of the eye-ball, but he does not refer to the ear.

Inasmuch as the present paper is directly concerned only with the capsule of the ear and the contained periotic connective-tissue spaces, it will not be necessary to trace the further elaboration of a more precise knowledge of the structure of the membranous labyrinth which rapidly took place following the appearance of Kol- iker's text-book and the introduction of the new histological and embryological methods which were devised in such abundance at about that time. A complete survey of such investigations is given in the monograph of Retzius (1884), to which the reader is referred. Our review here of the subsequent literature will be restricted to those publications having a special bearing on the periotic connective- tissue structures and the problem of their development.

The canalis reunions was discovered by Hensen (1863) as a communication exist- ing between the ductus cochlearis and the sacculus. This established the relation of the cochlear duct as a definite part of the closed system of the membranous laby- rinth, and its complete separation from the vestibular space. Using the terminology of Bre.schet, it thus constitutes an endolymphatic space, whereas the scala vestibuli and seal atympani are both perilymphatic spaces. Hensen also described the aquse- ductus cochleae. He regarded it as an invagination from the outer i)erichondrium into the cochlea by a i)rocess similar to the invagination of the aqueous humor of the eye. In the embryo it consists of a connective-tissue tube which is continuous with the primary periosteum. It si)lits into two limbs, the shorter one of which extends towards the round window and forms the lining of the proximal part of the scala tympani. The other limb of the membranous aqueduct extends towards the modio- lus and unites with the dura mater of the acoustic nerve.

Hensen was followed by Odenius (1867) who gives a careful description of the position of the different parts of the membranous labyrinth and of the "peri- lymphatic" spaces surrounding them. He separates the perilymi)liatic space of the vestilnile into two divisions based on the attachment of the ut^ricle to the vestibular wall. The lower and chief division he names sinus perilymphaticus vestibuli. This communicates with the upper division, which surrounds the upper i)art of the utricle and extends along the semicircular canals. This part is narrower and is hardly more than a cleft situated between the membranous labyrinth and the bony wall and is traversed by manj^ trabeculse.

A third author who could be put in this group is Boettcher (1869). He pub- Ushed two papers which bear upon the periotic spaces, but these are not available to the writer and resort has been had to the account of them given by Retzius (1884). He describes the formation of the scalae in sheep embryos. They make their first appearance in embryos 70 nmi. long, beginning in the first turn of the cochlea and gradually extending to the second and third. According to him there is a prelim- inary formation of mucus tissue in the region in which the scalie are to appear; this then undergoes a fatty degeneration, the result being the formation of the spaces. He warns against confusing this special " Schleimgewebe " with ordinary intracap- sular connective tissue and opposes Hensen's theory of its invagination from the outer periosteum. According to him (Boettcher, 1872), it arises in loco out of the original cellular embryonal connective tissue.

Although it was recognized that there must be a provision in the human adult ear for the renewal and drainage of the intralabyrinthine fluid, yet there was no positive evidence of how this was accomplished until the introduction of injection methods. Schwalbe (1869) found that when Berlin blue is injected into the sub- dural space the injection mass passes through the internal auditor}' meatus into the space existing between the bony and membranous labyrinth. Since he could also trace the injection mass from the subdural spaces into the lymph vessels and glands, he therefore believed that the perilymph spaces represented true IjTnph spaces, for which the arachnoid spaces acted as the main drainage-channels.

In order to test out the communication reported by Schwalbe between the arach- noidal spaces and the perilymphatic space, a series of injections were made by Weber (1869), who found that the injected fluid accompanied the acoustic nerve as far as the lamina cribrosa, but did not go through this. It passed rapidh^ however, through the aquaeductus cochleae into the perilymphatic space of the cochlea. Later, this same investigator (Weber-Liel, 1879) invented the aspiration method by which the results were refined, and he was able to avoid the production of artificial paths which conmionly result where strong pressure is necessary for the injection and which apparently had vitiated Schwalbe's experiments. He proved clearly that the aquseductus cochleae was the primary path of communication between the peri- lymphatic and arachnoidal spaces, and that it consists of a free canal lined by an extension of dura mater connecting the scala tympani with the cranial cavity. He was not sjjecific, however, as to whether the communication was with the subdural or subarachnoid space.

Key and Retzius (1875) in their extensive studies on the brain membranes, were able by injection methods to show that the spaces of the brain membranes stood in open communication with the perihonphatic space of the labyrinth, but, although they were able to trace the injection mass along the acoustic nerve, through the lamina cribro,^?, and along the finer filaments of the nerve into the lamina spu-ahs, they were not sure of its communication there with the perilymphatic space. They showed, on the other hand, that the latter could not be injected through the aquipductus vestibuli and scemod convinced that the main communication was through the aqua?ductus cochleae as described by Weber, which view remains the [)revalent one to-day.

It may be added that Retzius (1884) subsequently made some further injection experiments in older fetuses and in the adult, which were published in his large monograph on the ear. He found (p. 330) that in this way the scala tympani com- municates freely through the ductus perilymphaticus with the subarachnoid spaces, and not with the subdural space, which point had been left undecided by Weber. By injecting through either the round or oval window he was able to trace the escape of the fluid into the subarachnoid spaces, but never into the subdural space.

The comparative anatomists gave relatively little attention to the connective- tissue spaces around the ear and there was consequently no great advance secured from this aspect of the problem. Hasse (1873, account taken from Retzius) inves- tigated embryos of various mammals, but his results ar^e confusing. Concerning the lymph tracts of the inner ear he showed that in man and other mammals, in embryonal and adult stages, there exists a channel to which he gave the name ductus perilymphaticus, which is the same channel through which the injectionists had forced their fluids into the scala tympani. Hasse described this as provided with a sac which connects on the one hand with the cavum subarachnoideale (the outer epi- cerebral space after sjjlitting the brain membranes into pia and arachnoid) and into a lymjih- vessel on the other. This drainage path of the "perilymph, " accord- ing to him, is not the only and in fact is not the chief drainage path; a similar path, consisting of a funnel-shaped sheath of arachnoid, projects into the internal audi- tory meatus accompanying the acoustic and facial nerves. In a later paper, Hasse (1881) reverses the importance of these two channels and describes the perilymph as flowing chiefly through the ductus perilymphaticus into the peripheral lymph system in the region of the jugular foramen, the same channel also draining the cerel)rospinal fluid of the subarachnoid cistern. There is also, according to him, some drainage from the subdural space through the internal auditory meatus.

Of more importance is the description of Retzius (1884). In his large mono- graph on the ear the gross and finer morphology of the periotic spaces and espe- cially of the higher mammals, is described in greater detail and completeness than had previously been done. The comparative emliryology of the spaces is referred to by Balfour (1885). He speaks of lymphatic spaces (p. 522) as fonning in the rne.'^ol)last between the membranous labyrinth and the cartilage. These spaces are partially developed in Sauropsida, but become larger and more important in mam- mals, where they form the two scalse and the space surrounding the utricle and semi- circular ducts. According to him the scahc begin to develop at the basal end of the cochlea, the cavity of each being gradually carried forward toward the apex of the cochlear canal by a " jirogressive absorptif)n of the mesoblast. "

The descriptions of Retzius (1884) and of Kolliker (1861 b, 1884) and also the chapter in the .sixth edition of the " Ciewebelehre, " rewritten by v. El)ner (1902), have had a prevailing influence on the present conception of the character and development of the tissues of the otic capsule. There should be mentioned with these also the work of Krause (1901), who studied the development of these struc- tures in a number of vertebrate forms. He finds that the first traces of the formation of a "perilymphatic space" occur some little time before the formation of the scala tympani, in the region lateral to the utricle and saccule. At this point the perilymphatic tissue becomes gradually fluidified and there arises between the lateral wall of the two sacs and the cartilaginous wall of the labj^rinth a large peri- lymphatic space — the cisterna perilymphatica — to which the foramen ovale serves as a direct approach. From this cistern the space-formation spreads into the cartilaginous semicircular canals and simultaneousl.v there begins the formation of the scalse, resulting finally in a cavity system that incloses the entire membra- nous labj^rinth. Reference is made by him to the ductus perilymphaticus of Hasse, which connects this system with the subarachnoid spaces. This duct opens at one end in the vestibular part of the scala tympani and in the jugular fossa at the other. Concerning its development, nothing further was known.

In studying the development of the otic capsule, one is led into the "general question of the growth of hyaline cartilage, for which there is an extensive litera- ture and which is beyond the scope of the present jiaper. For general papers on this subject the reader is referred to those of Retterer (1900), Mall (1902), and Bardeen (1910). Other papers that may be mentioned as dealing particularly with the histogenesis of the skull are those of Solger (1889) and Filatoff (1906). An experimental stud}' bj^ Lewis (1907) should be referred to, in which it is shown that the production of the cartilaginous capsule is dependent upon the presence of the epithelial vesicle and that a transplanted otic plate becomes surrounded by cartilage derived from the tissue of the host.

The gross morphologj^ of the cartilaginous capsule of the car has been described for several embryonic stages in connection with the cartilaginous skull as a whole, and mention may be made in this connection of the work of Gaupp (1906) and Terrj' (1917) on certain vertebrate forms and the papers of Levi (1900) and Macklin (1914) on the human embryo. The writer has also had the opportunity of studying a recon- struction of the otic capsule in a human embryo 21 mm. long made by Professor W. H. Lewis and one of a 43-mm. embryo made by Dr. Macklin, both of which were modeled in this laboratory and have not yet been published.