Paper - The morphogenesis of the follicles in the human thyroid gland

From Embryology
Embryology - 16 Nov 2018    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Norris EH. The morphogenesis of the follicles in the human thyroid gland. (1916) Amer. J Anat. 20(3):411- .

Online Editor  
Mark Hill.jpg
This 1916 historic paper by Norris describes early human embryo thyroid development.

Also by this author:

Modern Pages: Thyroid Development

Endocrine Links: Introduction | BGD Lecture | Science Lecture | Lecture Movie | pineal | hypothalamus‎ | pituitary | thyroid | parathyroid | thymus | pancreas | adrenal | endocrine gonad‎ | endocrine placenta | other tissues | Stage 22 | endocrine abnormalities | Hormones | Category:Endocrine
Historic Embryology - Endocrine  
1903 Islets of Langerhans | 1904 interstitial Cells | 1908 Pancreas Different Species | 1908 Pituitary | 1908 Pituitary histology | 1911 Rathke's pouch | 1912 Suprarenal Bodies | 1914 Suprarenal Organs | 1915 Pharynx | 1916 Thyroid | 1918 Rabbit Hypophysis | 1920 Adrenal | 1935 Mammalian Hypophysis | 1926 Human Hypophysis | 1927 Hypophyseal fossa | 1932 Pineal Gland and Cysts | 1935 Hypophysis | 1937 Pineal | 1938 Parathyroid | 1940 Adrenal | 1941 Thyroid | 1950 Thyroid Parathyroid Thymus | 1957 Adrenal

Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic Textbook" and "Historic Embryology" appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms and interpretations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

The Morphogenesis of the Follicles in the Human Thyroid Gland

Edgar H. Norris

Institute of Anatomy, University of Minnesota


I. Introduction

In spite of numerous investigations, many questions concerning the development of the thyroid gland are still unsettled. This applies particularly to the morphogenesis of the thyroid follicles. The various and contradictory views in the literature on the development of the follicles are doubtless due in part to lack of adequate series of successive embryonic stages (especially of human embryos) available for study. The greatest difficulty, however, has arisen from the use of inadequate methods of investigation. The method hitherto almost exclusively used, that of direct observation of the microscopic. sections, is insufficient. By the use of reconstruction methods, however, it has been found possible in the present study to reach a satisfactory solution of this difiicult and important problem, at least regarding some of the more fundamental features.

This study was undertaken in the Anatomical Laboratory of the University of Minnesota. at the suggestion of Prof. C. M. Jackson, under whose supervision the work was conducted. I wish to thank Dr. Jackson for his valuable aid and criticism.

II. Literature

The literature concerning the thyroid follicle will be consid— ered in chronological order. First, it is desirable to mention briefly the various views which have been held concerning the morphology of the adult (human) thyroid follicle. Then fol— lows a brief statement of the conclusions concerning follicle development (including also prefollicular stages), which have been arrived at by the observers who.have worked upon human material. A few observations, made on lower forms, which have seemed especially pertinent to the present problem, are also included. Unless otherwise indicated, however, all statements refer to human material.

The follicular'str'ucture of the adult human ‘thyroid gland has long been known. According to Boéchat (’73), Lalouette (1750), who was the first to describe the minute structure of the thyroid gland, found vesicles which seemed to communicate with each other. Bardeleben (’41) is said by Zeiss (’77) to have been the first to describe the adult thyroid follicles as isolated structures. Five years earlier, however, Jones (’36) described the thyroid follicles in considerable detail as completely closed vesicles. Although there has been considerable disagreement concerning the structure of the adult thyroid follicle, the majority of the later observers have, like Jones, described the vesicles of the adult gland as closed, spheroidal bodies. Cruveilhier (’43), Virchow (’63), and more recently Boéchat (’73), Zeiss (’77), and Hitzig (’94), however, have followed Lalouette in describing the follicles as forming a system of branched and communicating cavities Within the gland. Still others, like Streiff (’97), have maintained that both branching forms and isolated vesicles occur in the adult gland.

Jones (’36), who was perhaps the first to describe the microscopic structure of the human fetal thyroid, found that in‘a fetus of four and one-half months the cells of the gland had become partially arranged into solid, globular masses; but no Vesicles were observed at this stage.

Remak (’55) described-in chick embryos in the wall of the primitive saccular, epithelial thyroid anlage the formation of thickenings which become separated and later give rise to the thyroid follicles. He also _thought that the original saccular anlage might persist for some time and form new secondary vesicles by a process of constriction. He described similarly the origin of secondary follicles, both by constrictions and by solid budding, in the thyroid of pig fetuses four inches and above in length.

Peremeschko (’67) described the division of primary into secondary follicles in mammalian fetuses. Colloid is described as arising partly by secretion and partly by colloid metamorphosis of epithelial cells.

W. Miiller (’7 1) described a developmental stage in which the thyroid consists of a network of cylindrical tubes. Such tubes were found in a 24 mm. fetus and in decreasing numbers in later fetuses and even in a three year old child. These tubes arise from solid epithelial cords by the development of a central lumen. The segmentation of the tubes with the formation of the gland-vesicles is produced by ingrowth from the mesoblast.

Horcicka (’80) found the thyroid gland of a four months’ fetus to be made up for the most part of solid cell masses with a beginning of lumen formation in the central cells of these masses. Typical gland structure is found after the fifth fetal month.

Wolfler (’80) described the formation of follicles from solid masses of epithelial cells. Toward the end of the fetal period and after birth the peripheral cells of the groups dispose themselves in a circle. The central cells become at. first granular, then degenerate and disappear in .the pale, granular mass which fills the lumen of the vesicle thus formed.

L. Stieda (’81) noted that the anastomosing epithelial cords (‘Epithelstrange’) of the embryonic, mammalian thyroid are at first always solid, but that in the ends of these cords lumina appear, and the resultant vesicles are gradually constricted off.

Baber (’81) described the fully formed follicle as spheroidal in form, but observed also branching follicles which are probably giving rise to secondary follicles by a process of division.

Wolfler (’83) described the process of later development in the human thyroid gland as centrifugal. He distinguished a cortical and a medullary portion, which are respectively youngest and oldest, least developed and most developed portions.

His (’85) described in the thyroid gland of an embryo (Zw) between 16.5 and 22 mm. in length cells grouped to form acini or tubes. The inner ends of the cells have a light, colloidal appearance.

Biondi (’89) found that the (postnatal) thyroid vesicle discharges its contents, collapses and finally rearranges itself in the form of a number of small acini which repeat the process. He held that the colloid arises by cell secretion, and not by cell degeneration.

Ribbert (’89) described a centrifugal growth of the thyroid in embryos and newborn. Follicles are formed by the outgrowth of solid buds or sprouts from the old follicles.

Lustig (’91), who studied the thyroid gland in the pig and other animals, affirmed that colloid and follicles appear synchronously as the result of the degeneration of the central cells of the preexisting solid masses.

Podack (’92) found well formed follicles in a fetus of five months. In some parts of the gland the follicular structure is only suggested and many cell-masses and cell-cords are present.

Marshall (’93) found that the thyroid in chick and frog embryos presents a stage in which the gland is made up of communicating, epithelial tubes. In the rabbit he described the presenceof out-growths, some solid and some hollow, from the primitive epithelial anlage. In the human embryo: “At an early stage the lobes are excavated by a number of detached cavities, which become the vesicles of the adult thyroid.”

Zielinska (’94) found the structure of the thyroid in newborn children variable both in size and number of the follicles, and also in the amount of solid cell masses. The relations “errinern an acinose Driisen und erwecken den Gedanken, dass hier ein sich verastelnder Driisenkanal vorliegt, als dessen Endblaschen die solide Zellhaufen gelten konnen.”

Hiirthle (’94) described in the thyroid of young dogs scattered masses of interfollicular epithelium, in which new (primary) follicles arise by the secretion of colloid into the angles between adjacent cells.

Anderson (’94) described secondary follicles (postnatal) arising from the collapsed epithelium of emptied follicles in various mammals. The new lumina are formed by cell-secretion of chromophile spherules.

According to L. R. Muller (’96), the origin of small secondary follicles from the larger follicles, as described by Ribbert (’89) is clearly evident, even in the human adult.

Tourneux and Verdun (’97) in a careful study of the branchial derivatives in the human embryo described the transformation of the (median) thyroid plate (and later of the lateral thyroid anlage) into a richly anastomosing network of solid epithelial cords by ingrowth of vascular connective tissue in a 14 mm. embryo. This network was likewise observed in embryos of 19 mm. to 37 mm. in length. At 37 mm., the cords become varicose, and follicles develop by the formation of a cavity within each of the enlargements. A similar process of morphogenesis is described in the rabbit embryo by Soulié and Verdun (’97).

Streiff (’97) made wax reconstructions of normal, adult human thyroid tissue, and found it to be made up of closed follicles, ovoidal or spindle shaped. Branched forms due to budding or to secondary fusion were also described; some of these more complex forms he thought may represent persistent branching, a continuation of the embryonic process. He concluded that the thyroid arises as a branched tubular gland, the follicles being formed by constriction of the tubes.

Schreiber (’98), in a fetus of three months, found the thyroid gland for the most part arranged into follicles which contained much colloid.

Kursteiner (’99) in fetuses from 8 to 30 cm. in length found the thyroid lobules made up of round or elongated, solid or hollow follicles. The lumina are few in number up to about 20 cm., but in the older fetuses they are numerous and evenly distributed throughout the gland. Some branching vesicles were also noted at 17.5 cm.

Prenant (’01) (p. 13) stated that in the embryonic thyroid the solid epithelial cords are transformed into a network of tubes from which the follicles arise by a process of constriction.

Von Ebner (’02) found numerous Well developed follicles in older fetuses and newborn. Between the follicles are found, even in the adult, frequent solid strings and nests of epithelial cells, which are in the majority during development.

Elkes (’03), who studied the thyroid in fetuses from four and one—half to six and one-half months in age, found that it presents both solid cords and well developed follicles in variable number. In the newborn the earlier follicles have largely disappeared, leaving only a few at the periphery of the gland.

Hertwig (’10) (pp. 444-446) described in the embryonic thyroid anastomosing epithelial cylinders. These become tubular; "varicose dilations are by ingrowth of the adjacent connective tissue cut off to form the permanant follicles.

Isenschmid (’10) found that in the thyroid of children the gland grows not only by the increase in the size of the follicles, but by the formation of new follicles by two methods: budding and division. He found no evidence that follicles are formed from solid cell-masses (interfollicular epithelium) retained from the embryonal period.

According to Hesselberg (’10): “Die Ausbildung der Thyreoidea in der fotalen Periode erfolgt durch Zerfall der urspri‘1nglich soliden Zellplatte in solide Zellstrange. Diese schniiren sich zu Blaschen ab, die zuerst am kaudalen Pol auftreten.” The normal structure of the thyroid is established from the fourth fetal month on. Desquamation of epithelial cells was found in about half of the cases from the seventh to the ninth fetal month, and the follicles are almost entirely obliterated in the newborn. During the first week of postnatal life the follicles are reformed and increase in number by a process of budding.

Prenant and Bouin (’11) give an account of the development of the median thyroid anlage similar to that given by Prenant (’01).

Broman (’11) described in the differentiation of the thyroid anlage a tubular stage transformed by constrictions into beaded chains and finally into separate follicles.

According to Grosser (’12), the thyroid anlage begins to separate into solid cords in the human embryo of 8 mm. In the 50 mm. fetus the cords, especially in the periphery, appear beaded. The beaded cords become divided into separate cell-masses, the anlages of the follicles. The lumina may appear as independent cavities (no tubular stage), before the follicles are detached, or they may arise later even in early postnatal life.

Simpson (’12) referred to the tubular structure of the thyroid and describes the gland of a seven months old child as tubular in character.

Aschoff (’13) stated that in the developing thyroid gland connective tissue separates round epithelial balls from the anastomosing cords; and it is in these ‘balls’ that the follicular lumina develop.

Sobotta (’15) described the first lumina as appearing in the peripheral parts of the lateral thyroid lobes in a fetus 50 mm. in length. The final breaking up of the cell cords into single groups, which will later form follicles, progresses very gradually, so that the final structure of the gland is arrived at only after birth. Interfollicular epithelium persists, Which may later give rise to follicles.

Kingsbury (’15) in describing the early development of the thyroid states that lumina (follicular cavities) appear within the cell cords in fetuses of 32 mm.,' but colloid is not demonstrable until 40 mm.

III. Material and Methods

This study is based upon the collection of human embryos in the Anatomical Laboratory of the University of Minnesota. Several of the series used are in excellent condition for histological study. The embryos of the collection have been variously fixed and stained. The additional glands specially prepared have, for the most part, been fixed in Formol-Zenker; embedded in paraffin, sectioned at 10 /4; mounted serially and stained with alum~haematoxylin and eosin, or iron-haematoxylin and eosin.

Besides the specimens mounted in the collection, other glands from newborn children and from children in the early years of life were obtained at autopsy. These were used merely for purposes of comparison, and have neither been listed below, nor discussed in this paper.

The following table shows the materials used in this work. The embryos and fetuses used are arranged in the order of their crown-rump lengths. ‘Minn. E. C.’ refers to the Minnesota Embryological Collection. An asterisk (*) following the number signifies that the thyroid gland alone was sectioned. Otherwise the entire embryo was available in serial sections.

The ordinary reconstruction methods, both plastic (Born’s waX—plate method) and graphic, were utilized in the present study. In all cases where a determination of the follicular form or structure was attempted, special precautions were observed in making the reconstructions as accurate as possible.

The drawings for reconstruction were made with the camera lucida on transparent paper. After the drawings were completed, those of successive sections were superimposed upona tracing—table, and each epithelial structure in the section given a letter or number. The drawings were controlled by careful microscopic observations, to determine the frequently complicated relations of neighboring follicles. By this method it was possible to determine with certainty what the limits of any particular mass or follicle might be.


MINN. E. C. N0.




punk»-IsphmwcncnwwwwwoototxbwmwN>Lx>L\ax\31\:m—-»—m-a>->—-¢-—«»—a»—a.-«p—-x-x- -x-*

44* 45* 46* 47* 48* 49* 50*

H 6 H 13 H 60 H 68 H 134 H 1 H 23 H 18 H 28 H 62 H 58 H 260 H 24 H 2 H 7 H 265 H 3 H 15 H 64 H 304 H 56 H 5 H 21 H 29 H 99 H 48 H 10 H 375 H 259 H 108 H 57 H 16 H 313 H 122 H 8 H 12 H 121 H 115 H 290 H 85 H 26 H 285 H 81 H 286 H 75 H 267 H 34 H 49 H 187 H 381

t§3l\’>l\'JN>l\’>b—-t—a>-dp—-)—n>—dr-Av-4\-—)—A»—\'>l\’)V“$@$a7\'C§UiU‘lC7IU1l\'?|-‘D-*‘~TC> $$$$$@$$$3U1$$$$$U‘$


23.0 24.0 24.0 25.0 26.0 26.0 26.0

5?:-‘E'£3'»l-‘<oooooc=o:ac:ro1»:>us.:>oococooooacn:ototo ooooouocnomwcucx-o_o>om3-»—¢>p1wg—«93959951 ooooooooooooooooooooooooo


Zenker Alcohol Alcohol Bouin Alcohol Alcohol Alcohol Formalin Alcohol Formalin Formalin Formalin Zenker Picro-sulphuric


Zcnker Alcohol

Alcohol-formalin Zcnker Alcohol Zenker Alcohol Alcohol Alcohol-formalin Bouin Alcohol Formalin Formalin Zenker

F ormalin Zenkcr Formalin Formalin Formalin Zenker Formalin Formalin Formalin Formalin Formalin Formalin Alcohol Zenker Formalin Formalin Formalin Boui n Formalin Formol-Zenker

10 15 20 15 20 12 10 10 10 20 20 15 12 15 10 20 12 12 20 20 20 15 15 12 20 12 15


IV. Morphogenesis of the Thyroid Follicle

a. Prefollicular period

A brief consideration of the thyroid gland in the prefollicular period is essential to an understanding of the follicular development. For it is during this prefollicular period that the anlages of the primitive follicles are derived from the original epithelial mass; and, to a certain extent at least, the size, form, and arrangement of the earlier follicles are thereby predetermined.

The earlier well-known stages in the development of the (median) thyroid anlage in the human embryo are not considered in the present paper. The original thyroid diverticulum becomes detached and transformed into a small epithelial plate, well shown in the 6 mm. embryo (No. 1 of the present series). This and several succeeding prefollicular stages of the thyroid were carefully reconstructed by Born’s wax-plate method, but for the present purposes it is unnecessary to figure or describe these models.

As is well known, the (median) thyroid epithelial plate soon presents irregularities, as shown in the 7.5 mm. embryo (No. 2), and rapidly becomes transformed into what appears in crosssections to be (as heretofore almost universally described) a network of anastomosing epithelial cords.

With the details of this process of transformation the present study is not concerned. One feature of the end result, however, which comes out clearly in the reconstructed models, is that the cord-like appearance seenin the sections is largely an illusion. Fundamentally the plate-like structure of the thyroid anlage persists for a considerable time, although somewhat modified by a complicated process of fenestration, splitting, and budding during the growth of the primitive epithelial plate. The resultant structure consists essentially in a mass of irregular, branching and fenestrated plates, for the most part longitudinally arranged (parallel to the long axis of the body), so that in crosssection they appear as ‘cords’ of epithelial cells (fig. 1). This type of structure, with varying degrees of complexity is found in the various prefollicular stages of embryos from about 10 mm. to 22 mm. in length. (Numbers 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14', 15, 16, 17, 18 of the present series.)

Such a description, however, does not apply to the so—called lateral thyroid anlage (ultimobranchial body). As described by Tourneaux and Verdun (’97) and others, this body long remains as a compact, deeply staining mass of. epithelium (the ‘inner condensation’ of Kingsbury ’14) on the dorso-medial aspect of each lateral lobe. Whether in the human thyroid the lateral anlage later atrophies or finally becomes transformed into permanent thyroid tissue is still uncertain. This region is therefore purposely excluded from the following descriptions, which apply primarily to the process of morphogenesis in the structures derived from the median thyroid anlage.

Fig. 1 Semi-diagrammatic drawing (camera lucida outlines) of a. transverse section through the left lateral lobe of the thyroid gland and neighboring structures from a human fetus 22 mm. long (No. 18). This shows the complex interrelation of the smooth epithelial plates (Pl.) (which appear as ‘cords’ in the section) and the vascular mesenchyme (_Mes.) found in the interspaces between them. The apparently. detached epithelial masses (Ma.) shown are sec -tions through projections from plates located in the upper part of the isthmus of the gland. B.V., blood vessel; M., muscle; L., larynx; 0., cartilage. X 133.

The thyroid gland in the final stage of the prefollicular period is shown in a fetus of 22 mm. (No. 18). This stage will be described in some detail, in order to make clear the subsequent process of morphogenesis of the thyroid follicles. When studied in sections, the thyroid at this stage is seen to be made up apparently of epithelial ‘cords’ Which are interrelated in a complex manner (fig. 1).‘ A loose, anastomosing network is thus formed, the interspaces of which are filled with a vascular mesenchymal tissue (fig. 1). The ‘cords’ are, in general, only two cells in width, a feature characteristic of the thyroid plate and its derivatives at various stages.

Upon reconstruction (figs. 11 and 12) it is found that the epithelial network seen in sections, and described by so many observers as consisting of ‘rods’ or ‘cords,’ is merely a section of flat, slab—like plates, or bands, representing portions of fenestrated plates. It is true that some few of the epithelial masses are actually cord—like in form, but by far the greater number are better described as bands or plates. A definition of terms is necessary at this point. The term ‘plate’ will be used to signify a structure of relatively slight thickness, presenting expansive surfaces which are more or less smooth, and which may or may not be perforated (fenestrated). A band is a narrow plate. Therefore a fenestrated plate may be considered as made up of a number of anastomosing bands. There is great variability in the way in which these plates and bands are arranged. Some are mere slabs, which may or may not be perforated; others form irregular prisms or rounded cylinders, which open at both ends into the surrounding mesenchyme. The length and width of these plates is also quite variable. In three respects, however, they are in general agreement. They are two cells thick; present fairly smooth surfaces; and are longitudinally placed. Within these epithelial bands (fenestrated plates) the primitive follicles of the thyroid gland develop.

b. Follicular period

A thyroid follicle will be defined as a completely closed sac whose wall is usually made up of only a single layer of epithelial cells. This definition includes all the features of the follicle which may be regarded as absolute and constant. The size and shape of the follicle may vary, and great differences are found in these respects in follicles of the same gland as well as in follicles of fetuses at different stages. Typically the thyroid follicle may be considered spherical or spheroidal in shape; but, as Will appear later, this type is subject to considerable variation. The term primary follicle will be used to include those follicles developing independently of preexisting follicles. The follicles derived from preexisting follicles, by budding or otherwise, are termed secondary follicles.

In the present series, the first primary thyroid follicles appear in a fetus 24 mm. in crown—rump length (No. 20). In this fetus the thyroid gland has essentially the same structure as has that of the last (No. 18) described in the prefollicular period, i.e. it is made up chiefly of longitudinally placed epithelial plates or bands, only two cells in thickness. But in this stage, the plates,

which have in previous stages been characterized by compara tively smooth surfaces, now present surfaces which are more or less roughened by the appearance of scattered hillocks or mounds (figs. 13 and 14). They are placed very irregularly with respect to one another, and may appear for the first time in any part of a plate, at its periphery or in a more central region.

When studied in cross sections (figs. 2 and 3) it is found that these hillocks the immediate anlages of the early thyroid follicles. .It is further seen that the hillocks are apparently produced by the concurrence of four different processes in the epithelium.. The first process is that of cell rearrangement, the second that of cell proliferation, the third that of absolute cell growth, and the fourth that of lumen formation. These processes, although described separately, may occur simultaneously.

The first departure from the two-celled plate arrangement, in the process of follicle formation, is found in a rearrangement of the cells of the plate (fig. 3). The cell outlines can be made out only with difficulty in most cases. But in those places where they can be seen, they bound cells which are more or less columnar in form. The nuclei are ovoidal or elliptical in outline and are placed Withtheir long axes perpendicular to the surface of the plate. Here and there along the course of the plate (fig. 3) it will be noticed that some of the nuclei have shifted their axes and have changed their relative positions.‘ Certain of the nuclei have rotated through an arc of 90 degrees so that their long axes, in their final position, are at right angles to their original position in the plate. As a result of this shifting process, little circlets (really spheres) of nuclei are formed in the plate.

Fig. 2 Semi-diagrammatic drawing (camera lucida outlines) of a section through the right lateral lobe of the thyroid gland and neighboring structures of a human fetus 30 mm. long (No. 30). Shows epithelial plates developing rough surfaces ; also early follicles in various stages of separation from the plates. Note the general increase in width of the plates as compared with those shown in figure 1, and the swellings and constrictions, giving ‘beaded chain’ appearance. Cap., capillary; Fol.L., follicle lumen; S.Ep.Pl., smooth epithelial plate; R.Ep.Pl., rough epithelial plate; Mes., mesenchyme; C., cartilage; L., larynx. X 133.

Fig. 3 Small portion of a cross section of the thyroid gland in a human fetus 30 mm. long (No. 30), magnified to show the cell tructure. The location of the section is indicated (a-b) in figure 14. Note the appearance of lumina (L.) in buds (B.) from the surface of the plate as well as in swellings along its course (L’). In the upper left hand corner a section through one of the hollow epithelial prisms is seen (Ep.P.). Mes., mesenchyme. X 400.

Fig. 4 Drawing of a section through follicle ‘d’ in figure 16. Note the irregular form of the follicle. both the lumen (L.) and the wall. Mes., mesenchyme; S.B., solid bud. X 400.

Fig. 5 Drawing of a section through an epithelial plate taken at the level marked (1-12 in figure 7. Mes., mesenchyme; L., lumen; C'ap., capillary; Ep.T., epithelial tag. X 400.

This shifting of nuclei is but the visible expression of the changed position of the cell. For while it is impossible to observe the cell boundaries in most cases, it is hardly probable that the nuclei shift their axes independent of the cytoplasm; moreover, the few faint cell-boundaries which may be made out show the same changes in position as do the nuclei. Further, it is usually found that at the point from which a nucleus has shifted toward the center of the plate a slight depression appears on the surface of the plate, indicating that the cytoplasm has shared equally with the nucleus in the movement. From these three facts it may be concluded that the first process manifested in follicle formation is the shifting of the axes of certain cells of the epithelial plate through an arc of 90 degrees. There is no evidence that the depressions on the surface of the epithelial plate are due to invasions or activity of the adjacent mesenchyme (fig. 3).

This process results in the transformation of the smooth surfaces of the bands (fenestrated plates) into surfaces which are somewhat roughened. The irregularities are apparently not, at first, due to swellings on the plates, but rather to the slight indentations produced by the shifting of certain cells toward the center of the plate as above described. When studied in cross sections (fig. 2) such a plate appears as a sort of beaded chain, with alternate swellings and constrictions. But, as noted above, the initial swellings due to this process are only apparent and are actually not greater in thickness than is the plate in other parts of its extent Where indentations have not yet occurred. "

The extraordinary cellular activity of the epithelium at this stage is clearly manifested by the large number of mitotic figures to be seen. There is no section of the gland in the 24 mm. fetus (N o. 20) which does not present several cells in process of mitosis. But the localization of these mitoses is even more significant than is their number. It will be noticed that the nuclear figures are usually found only in those places in the epithelial plates where actual thickenings on the plates are being formed. Therefore the little mounds which appear on the plates, as the immediate anlages of the early follicles, may be formed not only by the rearrangement of the already existing cells of the epithelial plates, but also by the formation of new cells as well. Consequently, it can easily be seen how the apparent swellings on the plates, produced by the rearrangement of the existing cells, may be transformed into actual swellings, by the absolute increase in number of the cells found in a localized area. These swellings become roughly spheroidal in form.

The third process referred to above is the absolute increase in size of the cells. While the cells are shifting their axes and proliferating, they are also growing in size. This fact results in the appearance seen in figure 2, where the solid, two-celled plates are found in some cases to be no greater in cross section than the one-celled plate which surrounds the follicle lumen. It might be thought that the cells do not actually increase in size, but only increase i.n height by a closer crowding together. But a study of figure 3 will show that such is not the case. For the cells are not more closely packed together in the newly formed follicles than they are in the two-celled plate.

This progressive increase in the height of the cells corresponds to the progressive stages in the differentiation of the twocelled plate into newly formed follicles. So that the thyroid gland of a 30 mm. fetus (No. 30) presents in different regions epithelial cells varying greatly in height. The lowest cells are found at the beginning of the process, i.n the two-celled plate; the highest being found at the other extreme, in the completely formed follicle.

Three of the four processes above mentioned as apparently involved in the evolution of the follicle from the epithelial plate have now been reviewed in detail. The formation of the lumen remains to be considered. Just preceding the appearance of the lumen, the spherule (in which it is about to develop) appears in cross section as a circlet of very tall columnar cells, whose nuclei are peripherally placed. This arrangement results in the formation of a striking picture. The nuclei are regularly placed at the periphery of the circle and form a dark band, which surrounds an expansive, clear, central cytoplasmic portion. The magnitude of this cytoplasmic area and the sharp contrast between the two portions (in the stained preparations) are usually striking features (fig. 3). It is in the center of this cytoplasmic area that the lumen makes its appearance as a tiny spherical space outlined by a definite and regular margin. It is as though the cells had but drawn a little apart, so that their central ends, instead of remaining in contact one with another, might be separated by an interval. The relation of the early lumina to one another is well shown in figure 6, which is a graphic reconstruction of a plate. It is important to note that no tubular stage is found in the process of lumen formation. The lumina appear as absolutely independent spaces.

As the lumina first appear they apparently have no content; but undoubtedly they contain some substance which is not stained with the ordinary methods, and which increases in amount with the size of the follicular cavity. Certain of the larger lumina (not all of them), which are perhaps older, are found to contain a hazy, granular substance. Typical colloid does not appear until later, in the 60 mm.. stage (No. 40).

The various possibilities as to methods by which the follicular lumen may arise will be considered later in the discussion.

Fig. 6 A graphic reconstruction (surface view) of parts of two fenestrated epithelial plates from the thyroid gland of a human fetus 30 mm. long (No. 30) to show the relative position of the lumina (Fol.L.) as they appear in the plate. Lumina indicated by stippled areas. Note that in all cases these early lumina are quite distinct and never connected with one another. Perf., perforation. X 267.

Fig. 7 A graphic reconstruction of a number of epithelial masses from the thyroid gland of a human fetus 60 mm. long (No. 40). Note the various degrees in the breaking up of the plates and the relative positions of the lumina. a—b indicates level of section in figure 5. X 267.

Fig. 8 A graphic reconstruction of a follicular complex shown in the lower part of figure 17. Lumina indicated by stippled areas. Note the solid buds. X 267.

Fig. 9 Model (reconstructed by Born’s wax-plate method) of a large follicle from the peripheral part of the lateral lobe of the thyroid gland in a human fetus 60 mm. long (No. 40). The three buds shown have lumina communicating with the lumen of the main follicle. X 267.

From the time of their first appearance, considerable Variability in the size of the lumina found in any particular gland is to be noted. The first isolated follicles are found in the more peripheral parts of the thyroid gland, and it is in these regions that they first attain large size and considerable complexity. For some time after the formation of follicles has begun, all stages previously described may be found in different parts of the same gland, a considerable portion of which retains the irregular plate~like type of structure characteristic of the prefollicular stages.

In the foregoing account, the cell masses in which the lumina develop have been described as spherules whose cross section is circular in outline. While this is true for typical follicles and in most cases, some variation within comparatively narrow limits is found. Ovoidal or somewhat irregular follicles occur, but these are not more numerous than would be expected in a rapidly growing tissue.

The foregoing descriptions of the early primary follicles have been taken in large part from observations made on two fetuses, one of 24 mm. (No. 20) and one of 30 mm. (No. 30). The excellent condition of these specimens has made possible studies of considerable detail. The members of the series (Nos. 21, 22, 23, 24, 25, 26, 27, 28, 29) intervening between these two, although not favorable for such intensive studies, show substantially similar structure. These stages may be summarized very briefly. The comparatively smooth epithelial plates of the prefollicular period have been transformed into plates with rough surfaces. The roughenings on the plates are the early indication of the follicles about to be formed. With the progressively increasing number of follicles the plates are transformed into irregular bands, which in turn give rise to groups of solid or hollow masses of cells. In the 30 mm. stage, however, the thyroid gland is still largely made up of anastomosing bands (fenestrated plates) (figs. 3, 6, 13, 14), although some entirely isolated follicles are found.

While the fetus is increasing in length from 30 mm. to 50 mm. the thyroid gland presents merely a continuation of the process above described; so that by the time the 50 mm. stage is reached the band or plate formations are relatively insignificant, while the isolated epithelial masses make up the greater part of the organ. All stages in the breaking up of the bands or plates may still be found, however. At this stage (50 mm.-) the gland is therefore characterized by the presence of few epithelial plates or‘ bands, very many isolated, solid cell-masses (anlagesof follicles) and relatively few well developed follicles. 1 At 56 mm. (No. 39) the thyroid gland presents slightly more follicles than at 50 mm. Not only are the follicles increased in number, but they also show some changes in both size and form. In the previous stages the follicles have ranged from 16 to 55 p. in diameter (average diameter about 40 ,u.), and have been typically spherical in form. But in this stage many of the follicles at. the periphery of the gland have enlarged considerably and have departed from the more usual spherical form of previous stages. This condition is more strikingly seen in the next fetus of the series (60 mm., No. 40) where both increase in size and change in form of the follicles are very evident (fig. 15). The number of follicles at this stage appears about equal to the number of solid epithelial masses. Figure 7 is a graphic reconstruction of a number of epithelial masses from which follicles are being formed and of certain isolated follicles. A study of sections at this stage will show follicles situated at the "peripheral portions of the lobe whose diameters are three or even four or five times as large as those of greatest magnitude located more centrally. The average diameter of the centrally placed follicles is‘ about 15 microns, while the average diameter of the follicles in the peripheral regions is approximately 55 microns. There is much less absolute variation in the size of the central follicles than in those of the periphery. Centrally placed follicles range from 13 to 17 microns, while peripherally placed follicles range from 20 to 103 microns in diameter. It will further be seen that these same large peripheral follicles are quite irregular in form. Instead of being circular in outline, they have a more or less regularly elliptical or ovoidal form, and may present a number of solid or hollow buds. Figure 9 represents a reconstruction of a typical one of these large budding follicles. Figure 7 is a graphic reconstruction of a number of epithelial masses found in the thyroid gland of the 60 mm. fetus (No. 40) to ‘demonstrate stages in the breaking up of the plates and the formation of isolated follicles, which are present at this time.

The 65 mm. fetus (No. 41) is the first member of the series in which practically no epithelial plates or bands are found. From this stage on through the remainder of the series the thyroid gland is made up almost entirely of independent, solid or hollow, more or less irregular epithelial cell—masses. The stages (Nos. 42 and 43) included between this embryo and the one of 86 mm. (No. 44) may be passed over briefly with the statement that the processes of budding and formation of new follicles as previously described are advancing rapidly parallel with the increasing length of the fetus.

Fig. 10 Microphotograph of a portion of a cross section of the thyroid gland in a human fetus 86 mm. long (No. 44). Note the numerous branching follicles, also the elongated follicles in various stages of division by constriction. This section is in the region of the reconstruction hown in figure 15. X 100.

The 86 mm. fetus .(No. 44) deserves special attention for it exhibits what appears to be the culmination of the budding processes referred to above. A- study of figure 10 reveals the extreme complexity of the gland at this stage. The hollow and solid epithelial masses are about equal in number. The extreme variability in the form of the follicles is the most striking feature (figs. 4, 10 and 16), and the reconstructions reveal more clearly the complexity of these structures than is apparent in sections. Elongated follicles with numerous hollow or solid buds, as well as numerous varieties of other more or less complex arrangements are to be seen. The whole picture is one of active growth. There is great variability in the size of the follicles, the largest being for the most part located in the peripheral zone ; but the extreme variation in the follicular form makes absolute measurements for comparison of follicles of little value.

This condition of the gland prevails in the remaining members of the series, up to and including the 158 mm. fetus (Nos. 45, 46, 47, 48, 49). There are but two differences to be noted. In the first place, the relative number of follicles is increasing from stage to stage, so that by the time the fetus is 158 mm. in length, the follicles in number are far in excess over the solid, interfollicular masses. Secondly, the number of irregular and complex forms, although still occurring, is becoming fewer (fig. 17), and the spheroidal follicles are relatively numerous.

The thyroid of the 163 mm. fetus (No. 50) has quite a different structure from that described for the members of the series just preceding (Nos. 44, 45, 46, 47, 48, 49). The irregular and branching follicular complexes are comparatively few in number, while small, spherical follicles make up almost the Whole of the tissue. There are, however, a few solid, interfollicular epithelial masses still present at this stage. The picture has changed from one in which the epithelial structures, instead of being complex in their form, have become relatively simpler in character and for the most part are organized into small follicles.

My material from this stage on up to and including newborn children shows to a greater or less extent the process of epithelial desquamation in the thyroid follicles as described by Elkes (’03), Hesselberg (’10), Isenschmid (’10) and others. Whether this process is physiological or pathological is as yet undetermined. Because the material at hand is not suflicient to warrant the drawing of any conclusions in this matter, it is not listed with the materials nor discussed in this paper.

In four of the fetuses studied (Nos. 40, 42, 43, and 44) a number of cyst-like follicles located in all cases in the lower and posterior (dorsal) part of the lateral lobe, were observed. Some of these follicles measure as much as 200 microns in diameter. The size of these structures is not much greater than that of some of the larger normally appearing follicles. But in structure they are quite different, having walls made up of very much flattened epithelial cells, whose nuclei cause the cells to bulge and are separated from one another by much greater distances than is the case with the nuclei in the more usual follicles. The lumina of these cysts are quite regularly circular in outline and in many cases contain a granular substance. It is as though a follicle had been greatly distended, the cells of the wall being stretched and flattened.

Fig. 11 Model (reconstructed by Born’s wax—pl-ate method) of the left lateral lobe of the thyroid gland from a human fetus 22 mm. long (No. 18). Antero— lateral surface view to show the perforated (fencstrated) plates, which present a relatively smooth surface. Line a—b indicates the level of the section shown in figure 1, and the level at which the model was divided to show the relations pictured in figure 12. The carotid arteries are shown on the right hand side of the figure. X 150.

Fig. 12 The lower section of the model pictured in figure 11. The upper portion has been removed in order to demonstrate the relations to the gland mass of the appearances found in cross section. The separation of the two portions is made at the level of the section shown in figure 1 and indicated ((1-12) in figure 11. X 150.

Fig. 13 Model (reconstructed by Born’s wax—plate method) of the upper portion of the left lateral lobe of the thyroid gland from a human fetus 30 mm. long (No. 30). Plates are more broken up than in figure 11 and present surfaces which are relatively rough and irregular. The upper parathyroid (P.) is shown. Lateral view. X 100.

Fig. 14 Model (reconstructed by Born’s wax-plate method) of a part of a plate from the thyroid gland of a human fetus 30 mm. long (No. 30) at a higher magnification, to demonstrate more exactly than shown in figure 13 the appearance and relations of the mounds forming on the surface of the plate. The plane a—b indicates the position of the section shown in figure 3. Lateral view. X 265. 436 MORPHOGENESIS or THE FOLLICLES 437

V. Discussion and Conclusions

Remak’s theory of the derivation of the thyroid follicles directly from the primitive saccular thyroid anlage has not been confirmed. In the prefollicular stages, the thyroid is by recent investigators quite generally described as assuming the form of irregular, anastomosing ‘cords’ or masses of epithelium. This undoubtedly appears to be the case when sections of the gland are observed (figs. 1 and 2). But the reconstruction methods used in the present investigation reveal a surprisingly different condition. It is found that, as a matter of fact, in the great majority of cases the cords are illusions and in reality are merely sections of fenestrated epithelial plates longitudinally arranged.

As to the further steps in the process of morphogenesis of the follicles from these anastomosing ‘cords’, widely divergent views have been held, as noted in the section on ‘Literature.’ While differing considerably in detail, these views may be classified according to their principal features. The morphogenesis of the primary follicles and of the secondary follicles will be considered in their order.

According to the apparently most widely accepted view, based perhaps ’argely upon preconceived ideas or theories concerning the evolution and morphogenesis of the thyroid gland, there are two d'.stinct stages in the transformation of the epithelial ‘cords’ into" follicles. First the anastomosing ‘cords’ acquire lumina, so that the gland becomes a more or less definite network of hollow epithelial tubes. The tubes then become constricted (by ingrowth of vascular connective tissue) into spheroidal segments, each of which becomes a small sac or follicle whose cavity represents a portion of the originally continuous lumen of the tube. This view has been advocated by W. Muller (’71), Marshall (’93 (in chick and frog), Streiff (’97), Prenant (’01), Hertwig ('10), Prenant, Bouin and Maillard (’11), Broman (’11) and others. Some like Streiff (’97) and Simpson (’12) have described this branching tubular condition as persisting in part throughout fetal life and even in postnatal life.

Fig. 15 Wax reconstruction of 2. portion of a peripheral lobule of the thyroid gland of a human fetus 60 mm. long (No. 40). This model is from the region shown in figure 10. The structure placed across the top of the figure is a blood vessel. The oblique lines designate cut surfaces. Viewed from above. X 270.

Fig. 16 A series of follicles (a, b, c, d, e) reconstructed by Born’s wax-plate method to show the varying degrees of complexity. All are taken from the thyroid gland of a human fetus 86 mm. long (No. 44). X 135.

Fig. 17 Wax reconstruction (Born’s method) of a peripheral lobule of the thyroid gland in a human fetus 158 mm. long (No. 49). The variable form and mutual relations of the follicles are evident. The structure at the left of the figure is a blood vessel. The oblique lines designate cut surfaces. X 130.

Other investigators, however, have described the -lumina of the primary thyroid follicles as appearing directly and idependently, with no preceding tubular stage. The anastomosing solid cell-cords are usually described as becoming varicose, with successive enlargements and constrictions, so as to present an irregular beaded chain appearance. Sooner or later each of these spheroidal swollen masses acquires a lumen and becomes separated so as to form an independent follicle. This method of follicle formation (with no tubular stage) has been described by Tourneux and Verdun (’97), Soulié and Verdun (’97), Grosser (’12), Aschoff (’13), Sobotta (’l5) and Kingsbury (’15).

It is impossible to decide from direct observations of sections which of the preceding theories is correct. By reconstruction methods, however, both graphic reconstruction and wax-plate models, evidence has been secured in the present investigation which definitely disproves the tubular theory and establishes in the human thyroid the independent origin of the lumina of the thyroid follicles. The follicles, however, appear not in epithelial ‘cords’ as described by earlier observers, but in the fenestrated epithelial plates above mentioned. The view that the thyroid is a modified branching tubular gland (Zielinska, Streiff, Simpson, and others) therefore obtains no support from its morphogenesis, aside from the initial stage of the primitive diverticulum.

In the glands studied in the present series the first follicles appear in a fetus of 24 mm. in length (No. 20). This is earlier than the time of appearance described by most observers. His (’85), however, described follicles in a fetus (Zw) whose absolute length is not recorded, but is placed in series between fetuses of 16.5 and 22.0 mm. in length. Kingsbury (’15) described follicles in a fetus of 32 mm.; Tourneux and Verdun (’97) in one of 32.4 mm.; and Grosser (’12) and Sobotta (’15) in fetuses of 50 mm. These are the only cases found in the literature where the presence of early follicles has been noted in fetuses of definite length. Several observers refer to the age of fetuses in which the thyroid follicles appear, but in terms too indefinite to be of value for comparison.

Although by definition the prefollicular period ends abruptly with the appearance of the first follicles, it is not true that the structure of the gland undergoes any corresponding sudden change with the advent of the follicular period. The epithelial bands (fenestrated plates) are only gradually replaced by the primary follicles and structures characteristic of the prefollicular period may be present through a considerable part of the follicular period, at least until the fetus has attained a length of 65 mm.

Concerning the first three processes (rearrangement of the cells, cell proliferation, and increase in the size of the cells) involved in the development of the follicles from the epithelial plate, no further discussion is necessary. The fourth process, however, that of lumen formation, calls for further consideration. The follicular lumen might arise in various ways, which have been suggested by earlier investigators. Most of the workers, -however, do not mention the process by which the lumen of the follicle is formed.

Wélfler (’80) and Lustig (’91) have described the formation of lumina in the solid cell masses by a degeneration of the Inore centrally placed cells. The present investigation does not support this view, however. In the first place none of the so—called central cells have been found; and secondly no evidences of degeneration have been observed in the primary follicles.

It might be supposed, as Hiirthle (’94) and Anderson (’94) have suggested, that the lumen is formed by the accumulation of colloid between the angles of the cells which compose the solid mass. Such a process would leave a colloid-containing space surrounded by epithelial cells. But in no case in the present observations was ‘colloid’ found within the very early follicles; although the accumulation of some other (precolloidal?) secretory product between the angles of the cells, might result in lumen formation. The fact is that the smaller lumina, which are probably the most recent in formation, have been universally found to be devoid of any demonstrable content, and that some of the larger and supposedly older lumina do contain a stainable substance. As to the nature of the precolloidal substance or substances, nothing definite is known.

One Inight suggest (as thought by His?) that the lumen could be formed by a degeneration or liquefaction of the central ends of the cells which later form its outline. In this case it would be expected that the early follicles would present a lumen outlined by an irregular or ragged margin. Without exception, however, the lumina of the early follicles are clearly outlined and marked off by a very sharp margin.

Having studied the way in which the follicle forms within the plate, it is of further interest to determine the method by which the follicle frees itself from the plate and comes to take up an isolated existence. Earlier observers, in describing the formation of follicles from anastomosing rods or tubes, have laid emphasis upon the activity of the adjacent mesenchymal tissue as the factor operating in the separation of the follicles. In the present study no definite morphological evidence appears in favor of this View. For as shown in figure 3 there is no special differentiation of the mesenchyme or increased vascularity in the regions in which follicles are being separated off from the plate. The evidence seems rather to indicate that the follicles themselves are the active agents in their separation. Thus as certain of the cells leave their original positions to assume a position more nearly in the center of the plate, the indentations previously described appear on the plate. These may be considered as weak points. And as the cells increase in number between these indentations it is not difficult to see how the increased pressure due to the increasedmass might force certain follicles out from the row in which they formed and thus isolate them from the parent plate. The vascular mesenchyme doubtless takes some slight part in the process, however.

As previously pointed out, the primary follicles may not in every case be at once separated completely from the plate. Instead of being in all cases sharply outlined and spheroidal in form, small portions of the plate (epithelial tags) may be left hanging to the follicle wall. The significance of such epithelial tags is readily understood when it is considered how easily they might be mistaken for epithelial buds arising from the primary follicle in which secondary follicles were about to form. These structures will be mentioned later in the discussion of the secondary follicles. It has been seen that there is no sharp line of demarcation between the prefollicular and follicular periods. Similarly, in the origin of follicles, the period of primary follicle formation is not sharply marked off from that of secondary follicle formation.

The number of thyroid follicles is apparently not absolutely established or finally limited at any stage of development (as is the case, for instance, with the glomeruli of the kidneys). In the earlier stages, the number of follicles increases by the formation of additional primary follicles. Soon, however, these primary follicles begin togive rise to secondary follicles (at 56 mm.). In later stages the formation of primary follicles apparently ceases-, although their occurrence even in the adult has been claimed, e.g., by Hiirthle (’94) and ‘Sobotta (’15), the newformed follicles being all "secondary in character. Various methods of secondary follicle formation have been described.

1. Origin from buds or sprouts. Ribbert (’89), L. R. Muller (’96), Streiff (’97), Isenschmid (’10) and others have described this process. The bud is‘ usually described as a local thickening of the follicle wall, which continues to increase in size by the proliferation of cells, until a solid bud, projecting into the stroma, is formed. Directly, through the concentric rearrangement of the cells, the form of the lumen can be made out, even while the young follicle is still in contact with the mother-follicle.

2. Originfrom collapsedfollicles. Biondi (’89) Anderson (’94), and others have described the process as follows. After filing the vesicle discharges its contents, collapses and finally rearranges itself in the form of a number of small acini which repeat the process.

3. Origin by fusion of follicles. Streiff (’97), V. Ebner (’02), Isenschmid (’10) and others have observed follicles which are apparently formed by the secondary fusion of two or more preexisting follicles.

4. Origin by division of follicles. Isenschmid (’10) has described the formation of daughter-follicles by the growth of an epithelial spur across the lumen of the mother follicle; and Peremeschko (’67) has noted the formation of secondary follicles by the constriction of the parent follicle.

In discussing my observations concerning the formation of secondary follicles in comparison with those of earlier observers, it may be said at the outset that nothing to support either the second or the third methods just outlined has been noted. These, however, have been described chiefly by previous observers upon postnatal material. The other two methods (budding and division of follicles), however, are in general agreement with the findings of the present investigation.

There appear to be three general methods by which secondary follicles arise in the fetal thyroid, the third of which might be regarded as a modification of the second. But each of these types is subject to a wide degree of variation, so that many intermediate and modified forms are found.

1. Solid epithelial buds may develop on the follicle wall (figs. 4, 8 and 16d). These may become separated from the parent follicle while in the solid state, or they may develop lurnina while connected with the wall of the mother-follicle, and subsequently be constricted off. This method is essentially that advanced by Ribbert (’89), L. R. Muller (’96), Streiff (’97), and Isenschmid (’10). It is difficult, especially in the earlier stages, to distinguish these solid buds from the ‘epithelial tags’ representing persistent portions of the original epithelial plates remaining attached to the earlier primary follicles.

2. Hollow buds whose cavities are continuous with that of the mother lumen are also found (figs. 4-, 9, 10). It might be suggested that these were originally solid buds whose lumen appeared independently and later established a secondary connection With the lumen of the parent follicle. While it is difficult to disprove such an occurrence, it would tend to reduce rather than to increase the number of follicles and seems improbable. The parent follicle apparently sends off extensions or branches, which represent both the wall and the lumen and gradually become constricted off to form new follicles. All possible stages in such a process, from the slightest outpouching of the wall to the finally separated follicle are easily observable.

3. The third method of secondary follicle formation is the simple division of the parent follicle somewhat as described by Peremeschko. The process appears similar to that of hollow bud formation. The follicle first takes on the form of an elongated ellipse which becomes constricted about its center. No cases of division by the ingrowth of epithelial spurs as described by Isenschmid (for postnatal stages) have been observed in my fetal material, though such may possibly occur.

Having considered the primary and secondary follicles separately it remains to consider them in their relations to each other and to the gland as a whole.

The formation of the secondary thyroid follicles begins when the fetus has reached a length of about 56 mm. From the 70 mm. stage on the formation of secondary follicles progresses with such rapidity that the total number of follicles is very greatly increased within a short period of time, and the relative number of primary follicles becomes progressively smaller.

It might be thought that such elongated, irregular, branched and budding forms as appear in large numbers in certain of the older members of the series (especially in the 86 mm. fetus, No. 43) are derived from irregular, branching bands (fenestrated plates) retained from the prefollicular period. But the evidence disproves that view. After the appearance of the first follicles, the bands and plates are being progressively broken up until the gland parenchyma in a fetus of 65 mm. is practically devoid of such structures, and is almost entirely made up of isolated solid or hollow masses of cells. So closely does this resolution of the plates parallel the increasing length of the embryo up to 65 mm. that it is very improbable that individual variations could explain the presence of these structures in such great numbers at the relatively late stage of 86 mm. Further, as shown in figure 16, forms of all degrees of complexity may be found from simple spheroidal follicles to those of extreme complexity. The conclusion is therefore reached that these branching structures, which are better described as follicular complexes than as follicles, are developmentally only follicles which have grown excessively and attained a high degree of complexity.

Such a rapid increase in the number of thyroid follicles as occurs in'fetuses between 65 and 158 mm. in length might be expected to produce a marked increase in the size of the gland. But. according to Jackson (’O9) the growth curve for the prenatal thyroid gland shows no remarkable increase in the size (weight) of the gland during this period. These two observations, which at first may appear contradictory, are readily explained when the size of the follicles is taken into account. The secondary follicles formed by budding and division of the primary follicles are Very small and arise from follicles which are in most cases relatively of much greater size. So that while the number of follicles is greatly increased during this period, the gland mass is not correspondingly larger. As previously described, the formation of secondary follicles becomes less rapid before the fetus has reached 163 mm. in length. From this point on, the number of thyroid follicles apparently increases but slowly, the subsequent growth of the gland being due rather to the increase in the size of the individual follicles than to a further increase in their number.

The significance of the large cyst-like follicles described in four of the fetuses is uncertain. Kursteiner (’99) has described the presence of similar follicles in four fetuses. The remarkable regularity with which they were found, in his cases as well as in those of the present series, located in the lower and posterior (dorsal) part of the lateral lobe of the gland, is a striking fact. Possibly they may be related to the cysts of the thyroid gland, frequently met in pathological conditions of postnatal life.

VI. Summary

By methods of reconstruction (both graphic and wax—plate), the complicated process of morphogenesis of the follicles of the prenatal human thyroid gland has been worked out and several mooted points definitely established.

1. The so—called ‘cords’ forming the anastomosing network in sections of the thyroid (median anlage) in the later prefollicular stages represent chiefly sections of epithelial bands, two cells in thickness, and forming irregular, fenestrated plates.

2. The frequently described stage in which the ‘cords’ are transformed into an anastomosing set of epithelial tubes from which the follicles are derived does not exist. The process of follicle formation gives no evidence or indication that the thyroid has been derived from a branching tubular gland.

3. The primary thyroid follicles arise directly as isolated and independent structures from the epithelial plates of the prefollicular period, by the rearrangement of cells, cell proliferation, increase in the size of the cells, and lumen formation.

4. The primary follicles appear in fetuses about 24 mm. in length. The epithelial bands (fenestrated plates) have practically disappeared in a fetus of 65 mm., but a few solid interfollicular epithelial masses are still present in fetuses 163 mm. in crown-rump length.

5. Secondary thyroid follicles are formed from preexisting follicles apparently by three methods: by solid buds; by hollow buds; and by constriction of the parent follicle.

6. The first secondary follicles appear in fetuses about 56 mm. in length, but are formed most rapidly in stages when the fetus is between 80 mm. and 158 mm. long. After 163 mm., the growth of the gland probably takes place largely by the increase in size of the individual follicles, rather than by increase in their number.

7. Large cystic follicles were observed in the lower and posterior (dorsal) parts of four glands from the older fetuses. Their significance is uncertain, as is likewise the apparent involution of the follicles with desquamation of epithelium observed in the later fetal and newborn stages. 446 EDGAR H. NORRIS

VII. Literature Cited

ANDERSON, 0. A. 1894 Zur Kenntnis der Morphologie der Schilddriise. Archiv f. Anat. u. Entw. S. 177.

ASCHOFF, L. 1913 Schilddri'1se—Normale Anatomie. In Pathologische Anatomie. 3 Aufl., Bd. 2, S. 928-9.

BABER, C. E. 1881 Researches on the minute structure of the thyroid gland. Philosophical Transactions, vol. 2.

BARDELEBEN, S. 1841 Observationes microscopicae de glandularum ductu excretorio carentium structura cleque earundem functionibus experimenta. Diss. inaug. Berlin. (Cited by Zeiss ’77.)

BIONDI, D. 1892 Contribution a 1’étude de la glande thyroide. Arch. ital. de biol., T. 17.

BOECHAT, P. A. 1873' Structure normale du corps thyroide. These, Paris.

BROMAN, J. 1911 Entwicklung der Schilddriise. In Normale und abnormale Entwicklung des Menschen. S. 288-9.

CRUVEILHIER, J. 1843 Glande Thyroide. In Traité d’Anatomie Descriptive. 2 Ed. T. 3. Paris.

V. EBNER, V. 1902 Schilddriise. In Kolliker’s Handbuch d. Gewebelehre. Bd. 3.

ELKES, C. 1903 Der Bau der Schilddrijse um die Zeit der Geburt. Dissert., Konigsberg. i. Pr.

GRoss1:R, O. 1912 The Development of the pharynx and of the organs of respiration. In Keibel and Mall’s Human Embryology, vol. 2, pp. 453, 468, 469.

HERTWIG, O. 1910 Die Schilddriise. In Hertwig’s Lehrbuch der Entwicklungsgeschichte des Menschen und der Wirbeltiere. 9 Aufl., pp. 444-446.

HESSELBERG, C. 1910 Die menschliche Schilddrijse in der fotalen Periode und in den ersten 6 Lebensmonaten. Frankf. Zeitschr. f. Path., Bd. 5.

His, W. 1885 Anatomie menschlieher Embryonen. Bd. 3.

HITZIG, 1894 Beitriige zur Histologie und Histogenese der Struma. Diss. Zurich. (Cited by Streifi" ’97.)

HORCICKA, J. 1880 Beitréige zur Entwicklungs- und Wachstumsgeschichte der Schilddriise. Prager Zeitschr. f. Heilk., Bd. 1. (Cited by Elkes ’03.)

HI"IRTHLE, K. 1894 Beitréige zur Kenntnis des Secretionsvorgangs in der Schilddriise. Pfl1"1ger’s Archiv. f. d. ges. Physiol., Bd. 56.

ISENSCHMID, R. 1910 Zur Kenntnis der menschlichen Schilddriise im Kindesalter, mit besonderer Beriicksichtigung der Herkunft ans ve1fschiedenen Gegenden im Hinblick auf die endemisehe Struma. Frankf. Zeitschr. f. Path., Bd. 5.

JACKSON, C. M. 1909 On the prenatal growth of the human body and the relative growth of the various organs and parts. Am. Jour. Anat., Vol. 9, No. 1.

JONES, C. H. 1836 Thyroid Gland. In Todd’s Cyclopaedia of Anatomy and Physiology, vol. 4, pp. 1102-1118. MORPHOGENESIS OF THE FOLLICLES 447

KINGSBURY, B. F. 1914 On the so-called ultimobranchial body of the mammalian embryo: man. Anat. Anz., Bd. 47, pp. 609-627.

1915 The development of the human pharynx. Am. Jour. Anat., Vol. 18, p. 329.

KURSTEINER, W. 1899 Die Epithelkorperchen des Menschen in ihre Beziehung zur Thyreoidea und Thymus. Anatomische Hefte, vol. 2.

LALOUETTE. 1750 Recherches anatomiques sur la glande thyroide. Mémoires de mathematiques et de physique présentésa1’a.cademie royale des sciences (savants etangers). T. 1., p. 159. (Cited by Boéchat ’73).

LUSTIG. A. 1891 Contribution a la connaissance de Phistogenese de la glande thyroide. Arch. ital. de Biol. T. 15.

MARSHALL, A. M. 1893 Vertebrate embryology.

MULLER, L. R. 1896 Beitrfége zur Histologie der normalen und der erkréinkten Schi1ddrfise._ Zieg1er’s Beitréige, Bd. 19.

MULLER, W. 1871 Uber die Entwicklung der Schilddrijse. Jena. Zeitschr. f. Med. u. Na.turw., Bd. 6.

PEREMESCHKO. 1867 Ein Beitrag zum Ban der Schilddriise. Zeitschr. f. wis~ sensch. Zoo1., Bd. 17.

PODACK, M. 1892 Beitréige zur Histologie und Funktion der Schilddriise. Dissert., Konigsberg. i. Pr.

PRENANT, A. 1901 in Poirier et Charpy, “Traité d’anatomie humaine,” T. 4, p. 13.

PRENANT, BOUIN ET MAILLAR15 1911 Traité d’histo1ogie,' T. 2, Histologie, p. 971.

REMAK, R. 1855 Untersuchungen fiber die Entwickelung der Wirbelthiere. Berlin. S. 122-123.

RIBBERT. 1889 Ueber die Regeneration des Schilddriisengewebes. Virchow’s Archiv, Bd. 117.

SCHREIBER, L. 1898 Beitréige zur Kenntnis der Entwicklung und des Baues der Glandulae parathyreoideae des Menschen. Archiv f. mikr. Anat., Bd. 52.

SIMPSON, B. T. 1912 Growth centers of the benign blastomata with especial’ reference to thyroid and prostatic adenomata. Jour. of Medical. Research, vol. 22.

SOBOTTA, J. 1915 Anatomie der Schilddriise (Glandula. thyreoidea). 29 Lieferung des “Handbuchs der Anatomie des Menschen” (von Barde1eben’s). Bd. 6, Abth. 3, Teil 4.

SoULI£: ET VERDUN. 1897 Sur les premiers développements de la glande thyroide, du thymus et des glandules satellites de la. thyroide chez le lapin et chez la. taupe. Jour. de1’Anat. et de la Phys. T. 23.

STIEDA, L. 1881 Untersuchungen fiber die Entwickelung der Glandula. Thymus, Glandula. thyroidea und Glandula carotica. Leipzig.

STREIFF, J. J. 1897 Uebcr die Form der Schilddrfisen Follikel des Menschen. Archiv f. mikr. Ana.t., Bd. 48, S. 579.

TOURNEUX ET VERDUN 1897 Sur les premiers développements de la thyroide, du' thymus et les glandules parathyroidiennes chez Phomme. Jour. de 1’Anat. et de la Phys. T. 23.

VIRCHOW, R. 1863 Die krankhaften Geschwiilste. Bd. 3, 1 Héilfte.

WCSLFLER, A. 1880 Ueber die Entwicklung und den Bau der Schilddriise mit Riicksicht auf die Entwicklung der Krépfe. Berlin. (Cited by Elkes ’03.) 1883 Ueber die Entwickelung und den Bau des Kropfes. Archiv f. klin. Chirurgie. Bd. 29.

ZEISS, O. 1877 Mikroskopische Untersuchungen fiber den Bau der Schilddrijse. Dissert., Strassburg, 1877.

ZIELINSKA, M. 1894 Beitréige zur Kenntniss der normalen und strumésen

Schilddriise des Menschen und des Hundes. Virchow’s Archiv. Bd. 136.

Cite this page: Hill, M.A. (2018, November 16) Embryology Paper - The morphogenesis of the follicles in the human thyroid gland. Retrieved from

What Links Here?
© Dr Mark Hill 2018, UNSW Embryology ISBN: 978 0 7334 2609 4 - UNSW CRICOS Provider Code No. 00098G