li.i L i g n n n. I u i i ii c u n i T t n o i i i ui UHLII LIBRARY OF THE UNIVERSITY OF CALIFORNIA IF THE UNIVERSITY OF CALIFORNIA /ft) -- LIBRARY *<=5 F THE UNIVERSITY OF CALIFOR ^ THE UNIVERSITY OF CALIFORN THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA PRESENTED BY PROF. CHARLES A. KOFOID AND MRS. PRUDENCE W. KOFOID QJ/~^\) THE UNIVERSITY OF CALIFORNIA LIBRARY OF THE UNIVERSITY nF fiiMFnnim Oaei I I DO 1 D V f MEMOIRS OF THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY No. 1 THE ANATOMY AND DEVELOPMENT OF THETSYSTEMIC LYMPHATIC VESSELS IN THE DOMESTIC CAT GEO. S.jHUNTINGTON PROM THE ANATOMICAL Itt^ffEATOBY OP COLUMBIA UNIVERSITY PHILADELPHIA, PA. MAY, 1911 MEMOIRS OF THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY No. l THE ANATOMY AND DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS IN THE DOMESTIC CAT GEO. S. HUNTINGTON FROM THE ANATOMICAL LABORATORY OF COLUMBIA UNIVERSITY PHILADELPHIA, PA. 1911 COMPOSED AND PRINTED AT THE WAVERLY PRESS BY THE WILLIAMS & WILKINS COMPANY BALTIMORE, I 1 . B. A. CONTENTS Introduction 5 Part I. The development of the systemic lymphatic vessels in their relation to the blood-vascular system 9 Historical review of theories of lymphatic development 10 Material used in present investigation 16 Comparison of haemal and lymphatic vascular development 19 Ontogeny of mammalian systemic lymphatic vessels 24 Mutual relations of developing systemic lymphatic vessels and em- bryonic veins 27 Phylogenetic relations of the abdominal veins and axial lymphatics in mammals 33 Development of systemic lymphatic vessels in the mammal independ- ent of topographical association with embryonal veins 49 Summary and conclusions of Part I 53 Figs. 1 to 28 Part II. The development of the preazygos and azygos segments of the thor- acic ducts 55 1. The thoracic duct approach of the jugular lymph sacs 60 Figs 29 to 91. 2. The preazygos segment 83 Adult conditions 84 Figs. 92 to 99. A. The development of the broncho-mediastinal trunk 91 Figs. 100 to 158. B. The development of the preazygos segment of the thoracic duct 107 Figs. 159 to 169. C. Junction of preazygos segment of thoracic duct and broncho- mediastinal trunk with each other and with the thoracic duct approach of the jugular lymph sac 113 Figs. 170 to 187. 3. Azygos segment 121 A. General analysis of the development of the thoracic duct in the azygos region 123 Figs. 188 to 193. B. Detailed consideration of the individual stages in the development of the azygos segment of the thoracic duct 129 Junction of azygos and preazygos segments of the thoracic duct 150 Figs. 194 to 275. Summary and conclusions of Part II 153 INTRODUCTION I have recently published in a preliminary communication, 1 a re'sume' of the results obtained in an investigation of mammalian lymphatic development and organization extending over the past six years, and now, in its main chapters, concluded. The paper above quoted was presented, with demonstrations of slides, at the 25th session of the Association of American Anat- omists held in Boston during Convocation week of 1910, and is intended as an attempt to definitely establish what I believe to be the genetic principle upon which all systemic lymphatic development in the mammalian embryo is based. In outline this matter was also presented and demonstrated to the Section of Anatomy and Embryology of the XVIth Inter- national Medical Congress held at Budapest, August-September, 1909, and published in the Proceedings of the Congress. 2 Owing to the character of the problem and its complexity, a detailed consideration of the same exceeds the reasonable limits of an article suitable for publication in our current anatomical periodicals, and the unavoidable number of microphotographic illustrations demanded makes publication through the ordinary channel still more unadvisable. For these reasons I have ar- ranged, with the cooperation of The Wistar Institute of Anatomy, through Director Greenman, to publish the details of my obser- vations on mammalian lymphatic ontogeny in the form of a series of monographs, in which the subject can be handled with 1 G. S. Huntington: "The Genetic Principles of the Development of the Sys- temic Lymphatic Vessels in the Mammalian Embryo." Anat. Record., vol. iv, no. 11, 1910, pp. 399 to 403, with 32 illustrations. (18 plates.) 2 G. S. Huntington: "Ueber die Entwicklung desLymphatischen Systems beim Sauger-Embryo." Compte-Rendu, xvi. Congres International de Medecine, Section 1, Anatomic, Embryologie, 2. Fascicule, pp. 127-142, Budapest, 1910. 6 INTRODUCTION less restraint than in one of the current publications. I have been led to the undertaking largely by nay conviction of the value of the work which has been done within the last decade in this field by American investigators. The development of the lym- phatic system is one of the very few broad morphological problems as yet incompletely solved, and I believe that the pains- taking, able and conscientious work of the relatively large num- ber of interested investigators of the subject in this country will eventually furnish a satisfactory answer to the question, as a national contribution to the advancement of anatomical science. I have been obliged to differ, on the basis of my own investi- gations, from the conclusions reached by most of my American colleagues. I hence welcome the opportunity of placing my results fully on record, in such a manner that they can be readily examined and verified, if correct, or refuted, if found to be erroneous. In carrying out this purpose it is of course necessary, in order to avoid repetition and economize space, to simply refer to those parts of the subject which have been already fully covered in the existing publications, and to include these articles as part of the entire record. The main problem then is narrowed down to the developmental history of the mammalian systemic lymphatic channels, as distinguished from the jugular lymph sacs, or other homologous structures of like origin and equivalent functional significance, wherever situated. The proposed series of publications will include the following topics in the order given: Part I. The development of the systemic lymphatic vessels in their relation to the blood vascular system. Part II. The development of the preazygos and azygos segments of the thoracic ducts. Part III. The development of the visceral lymphatic drainage, and especially of the lymphatics of the abdominal cavity; the forma- tion of the receptaculum, and of the postazygos segment of the thor- acic ducts, as well as the lymphatic return from the pelvic and caudal regions and from the posterior extremity. Part IV. The development of the main lymphatic trunks, other than the thoracic ducts, draining into the jugular lymph sacs, and through them into the venous system, viz., the cervical, jugular and supra-scapular lymphatics, and their mediastinal connections, and the lymphatic return from the anterior extremity along the sub- clavian vein. Part V. The interpretation of adult normal and variant lym- phatic organisation on the genetic basis, and the interdependence of the adult venous and lymphatic systems. The present publication includes Parts I and II of the above list. PART I THE DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS, IN THEIR RELATION TO THE BLOOD- VASCULAR SYSTEM The question as to the origin of the lymphatic vessels has, especially since 1902, occupied the attention of a number of American observers. These investigations have followed the older work on the same subject of Langer ('68), Budge ('80-'87), Gulland ('94), Ranvier ('95-'97) and Sala (1900), and the results have been published chiefly in the American Journal of Anatomy* and in the Anatomical Record.* During the progress of these researches a number of facts of primary importance bearing on the problem of lymphatic develop- ment and organization have been discovered. Some of these facts have been worked out in detail and are based on sufficiently extensive material and accurate observation to carry con- viction by their constancy and consistency and to warrant their acceptance as definitely established ontogenetic conditions in the mammalian embryo. Other observations still lack complete confirmation, and in some others the methods employed in their determination create a doubt as to their validity, and tend to confuse the subject. Finally there are other conditions concern- ing which there still exists an honest difference of opinion, and which hence require further study and definite determination. On the whole, however, the American work of the last six or seven years has led to considerable and permanent advance in Vol. i, 1902; vol. iii, 1904; vol. iv, 1904; vol. v, 1905; vol. vi, 1907; vol. ix, 1909; vol. x, 1910 4 Vol. ii. 1908: vol, iv, 1910- 9 10 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS our knowledge of the genesis of the mammalian lymphatic system. The results so far obtained, and the views based thereon, may, with inclusion of the older work on the subject of vertebrate lymphatic development in general, be briefly summed up as follows in the form of short theoretic statements: I. The lymphatic system is developed independently of the blood-vascular system. It is formed by the confluence of inde- pendently developed mesenchymal spaces, and, in case of the avian thoracic duct, by canalization of preformed solid strands of differentiated mesenchyme. 6 - 6 > 7 The works of Brachet and the combined researches of Brachet and Swaen, 8 in their relation to the interpretation of lymphatic development, also support the independent origin of the lymph- atic system from the mesenchyme. The interesting observations of Marcus on the development and organization of the lymphatic system in Hypogeophis 9 place the conclusions of this investigator in regard to the genesis of lym- phatic structures in this general division, with the addition of the conception of the coelomata as primary lymphatic spaces (vide infra, pp. 25 and 26), and the phylogenetic derivation of the peripheral lymphatic system from the same. 8 Budge: "Ueber ein Canalsystem im Mesoderm von Hiihnerembryonen." Arch, fiir Anat. und Phys., Anat. Abth., 1880, s. 320. "Untersuchungen iiber die Entwicklung des Lymphsystems beim Hiihnerembryo." Arch. f. Anat. u. Phys., Anat. Abth., 1887, s. 59. 6 L. Sala: "Sullo sviluppo del cuori linfatici e del dotti toracici nell' embryone di polio," Ricerche fatta nel Laboratorio di Anatomia Normale della R. Univ. di Roma, vol. vii, p. 263-269, April, 1900. 7 G. Lovell Gulland: "The Development of Lymphatic Glands", Jour. Path, and Boot., vol. ii, 1894, pp. 447-485. 8 A. Brachet, "Recherches sur le developp. du coeur, des premiers vaisseaux et du sang chez les amphibiens urodeles," Arch, d'anat. microscopique, ii, 1898. "Recherches sur 1'origine de 1'appareil vasculaire sanguin chez les amphibiens," Arch, de Biologic, xix, 1903. A. Swaen et A. Brachet, "Etude sur les premiers phases du developp. des organs derives du mcsoblast chez les poiissons teleostiens," Arch, de Biologie, xvi, 1899-1900. 9 H. Marcus, "Beitrage zur Kenntnis der Gymnophionen; II. Ueber interseg- mentale Lymphherzen, nebst Bemerkungen iiber das Lymphsystem," Morphol. Jahrbuch, Bd. xxxviii, Heft 4, 1908. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 11 II. The lymphatic vessels are directly derived from the venous system, certain embryonic venous channels being trans- ferred in toto to the lymphatic system. 10 ' 11 ' 12 - 13 III. All systemic lymphatics are formed by union of multiple direct derivatives from the embryonic veins, 14 or only the thoracic ducts are so developed, while the other systemic lymphatic ves- sels arise independently. 15 IV. The mammalian lymphatic system as a whole is developed by blind ducts which "bud off" from the embryonic veins of the cervical, and later from those of the inguinal region, widen out to form sacs, from which lymphatics grow to the skin and "cover its surface, while at the same time a growth of ducts occurs along the dorsal line following the aorta to make a thoracic duct from which lymphatics grow to the various organs." The theory under- lying this conception of lymphatic development assumes the primary formation of a number of sacs, derived from the veins, and lined by embryonic venous endothelium, from which, as the 111 C. Langer: "Ueber das Lymphgefasssystems des Frosches." Sitzb. d. Akad. 168 34 144 . . . . 31.5 90 . . 35 104... . 51 These embryos are contained in the Embryological Collection of Columbia University. The majority of the preparations were fixed in Zenker's fluid and stained differentially on the slide with Hsemotoxylon (Delafield) and Orange-G. I have also had, through the courtesy of Professor McClure, the opportunity of carefully examining three very interesting cat embryos of the Princeton Collection, series 34 and 37, each of 14 mm. crown-rump measure, and series 53, a 15mm. embryo. A series of 180 adult animals with successful injection of the main systemic lymphatics served as control for the embryologi- cal determinations, and for the comparison of normal and variant adult conditions of the venous and lymphatic systems with the corresponding ontogenetic stages. The results of my observations on the development of the mammalian systemic lymphatic vessels, as distinguished from the jugular lymph sacs, studied jointly with McClure, may be briefly summed up as follows: The systemic lymphatic vessels of the entire body are formed through confluence of numerous originally separate intercellular mesodermal spaces, which develop iij great part in close apposi- tion to the walls of the embryonic venous channels, and in exactly the same way as the primary anlages of the haemal vascular system, but independent of the latter. The endothelium lining these first anlages of the lymphatic vascular channels is from the RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 19 beginning independent of the haemal endothelium, and develops with the first appearance of the lymphatic spaces, from the indifferent mesodermal cells lining these spaces. In my opinion the lymphatic and haemal vascular endothelium have the same ge- netic derivation from the modified mesodermal cell lining the tissue spaces. The primary stage of endothelial differentiation is the same, whether the resulting channel system is to be assigned to the haemal or the lymphatic division of the vascular system. We have therefore two generations of the embryonic vascular endothelial cell, a haemal and a lymphatic. Both develop in the same way and as the result of identical genetic factors from the indifferent mesodermal cell. Both are from the very begin- ning of the process independent of each other in the mammalian embryo. I desire again to emphatically aver my conviction that all the systemic lymphatic vessels of the mammalian embryo, includ- ing the thoracic ducts and their tributaries, are neither in their genesis continuous " outgrowths" or "buds" from sacs of venous origin, wherever situated, nor derived from multiple outgrowths from the embryonic veins, such outgrowths subsequently separat- ing from'the veins and fusing into continuous lymphatic channels. They are, in my opinion, on the contrary, from their very first inception, independent of the haemal vascular system, and their endothelial lining is not derived from the blood vascular endo- thelium. They develop as independent intercellular mesodermal spaces, which become confluent with each other to form larger and larger communicating channels. These finally attain their entrance into the venous system through the intervention of the jugular lymph sacs, in the manner outlined in the publica- tions above quoted. 22 > 26 > 27 Before taking up the details of the development of the first lymphatic anlages in the mammalian embryo, it seems advisable to refer briefly to a resume of known facts in regard to the earliest formative stages of the blood vascular system, in order to facili- tate the comparison between haemal and lymphatic development. Phylogenetically, the earliest form of a closed circulatory system in multicellular organisms consists of intercellular canals 20 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS conveying a clear plasmatic fluid without cellular contents. The same picture is presented in the earliest stages of haemal development in vertebrates. Thoma's 28 investigations of the histogenesis of the blood vascular system in chick embryos have furnished us with a very clear picture of the process. The first histogenetic inception of the vertebrate haemal vascular system is marked by the condensation of the mesoderm into cellular strands. Between the cells of these praevascular strands multiple oval or round spaces develop, which enlarge, elongate and become confluent, forming a network of inter-com- municating channels, the hcemal capillary anlages. These chan- nels contain a clear colourless fluid, with no, or only very scat- tered, cellular elements. This fluid, obtained by secretion from the free surfaces of the cells limiting the spaces, is evidently under a certain definite pressure, which exerts an influence on the form of the cells lining the channels. These limiting cells lose their earlier isodiametric, more or less regular cuboidal form, and appear flattened, and on optical section spindle-shaped. They have begun to assume the endothelial character. Hence from its earliest inception the endothelial lining of vascular channels appears as an environmental adaptation of the meso- dermal cell. One surface of a cuboidal cell is freed from con- tact with adjacent cells by the development of an intercellular cleft, and this free surface is subjected to the pressure of the fluid contained in the earliest capillary anlages, modified by the tension pressure of the organism as a whole. This mechanical adaptation to the altered cellular milieu results in the formation of endothelium, and the process is identical in all portions of the mesoderm, independent of the question as to whether the resulting endothelial lined space shall subsequently be incor- porated in a haemal or a lymphatic system of vascular channels, or shall remain as a closed non-vascular mesodermal space. It seems to me futile to try to speculate on an ontogenetic stage in which endothelium acquires a " specific" character. It 28 R. Thoma: "Untersuchungen iiber die Histogenese und Histomechanik des Blutgefasssystems." Stuttgart, 1893. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 21 develops de novo in the adult under appropriate normal con- ditions. Furthermore, this endothelial characterization of modified mesodermal cells is from the beginning a multiple process, start- ing independently at innumerable separate and discrete points of the vascular area, and becoming only subsequently continuous by confluence of the individual separate anlages. This fact is of importance in drawing general conclusions as to the later extension of vascular endothelium, whether haemal or lymphatic. Up to this point the histogenetic and physical characters of all developing vascular structures are identical. The picture just described applies equally to the earliest definite anlages of the haemal capillary system, and, as I shall show, to the first appearance of the earliest lymphatic structures of the body. In the case of the definite blood-vascular channels of the verte- brate embryo, however, a further developmental change occurs, namely, the addition of free, specially modified, mesodermal cells, as the red blood cells, to the clear plasma circulating in the channel-system of the earliest capillary anlages in response to the first pulsations of the heart. The first blood vessels of the area pellucida appear, at least in part, to develop independentlyof the so-called " Blood islands." These latter, originally, form broad cellular strands composed of closely packed uncolored cells, which are only distinguished from the solid strands of the earliest vascular anlages of the area pellu- cida by greater size and massiveness. After the vascular cell- strands of the area pellucida have developed in their interior the intercellular closed oval or round spaces of the first capillary an! ages, similar spaces also appear in the more massive cell strands of the peripheral portion of the area vasculosa. In the subsequent confluence of these discretely developed spaces to form the early capillary reticulum, the blood islands become more and more surrounded by the forming channels and are thus separated from the adjacent tissues. New endothelial-lined spaces continue to develop on the surface of the blood-islands, enlarge and join the system of connected channels. The cells forming the walls of these primitive capillaries become, as above stated, transformed 22 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS into the typical flattened endothelial cells of the vascular intima. Excepted from this endothelial transformation are only those cells of the vascular walls which differentiate as young blood cells. Thus eventually the confluence of the originally individual and separate spaces produces a continuous and connected channel system, lined by endothelium, which nearly encircles the blood islands. The latter are therefore now in large part included within the lumen of the capillaries, with whose walls they are from place to place continuous. It only requires a further and complete solution of this continuity, and the accompanying freeing of the blood cells, to add the latter to the plasma circulating in the preformed channels. In the chick, according to Thoma's obser- vations, the resolution of the blood islands into separate blood cells occurs between the 45th and 55th hour of incubation, while the acquisition of haemoglobin by the cells occurs between the 40th and 45th hour. With this occurrence the development of the primary blood-vascular channels has reached its definite accomplishment. The general picture presented by the earliest development of the blood-vascular system may therefore be summarized as follows: (1) Differentiation of certain mesodermal areas and lines by the multiplication of mesodermal cells to form cell-strands of varying density and size (vascular strands). This appears to be a common antecedent condition not only of all vascular mesodermal structures, but also of other meso- dermal derivatives eventually destined to obtain a lumen and enter into the formation of canals, as the Wolffian tubules and the cell-strands of the gonad. It possibly explains the conditions described, for example, by Sala in the development of the avian thoracic duct, is seen in the developing aortae of early chick embryos, and is especially significant in the pictures furnished by Sabotta of the develop- ment of the aortse in fishes. (2) Development in the interior of these cellular strands of intercellular spaces in large numbers. (3) Bio-mechanical modification of cells lining these spaces to produce typical flattened vascular endothelium. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 23 (4) Confluence of these endothelial-lined spaces to form larger and larger vascular areas of intercommunicating channels, containing a clear plasmatic fluid, which circulates in the chan- nels in response to the establishment of cardiac pulsation. (5) Addition to the plasmatic contents of these channels, coincident with their further growth and extension into a con- tinuous circulatory system, of cellular elements, derived from the mesoderm and specially modified to acquire haemoglobin and function as red blood cells. These cells, by solution of tissue continuity, are liberated from the blood islands which are first surrounded by the confluent spaces of the capillary anlages. (6) Subsequent differentiation of the adventitia, with speciali- zation of districts by cardiac concentration, amalgamation of the plexus into larger arterial and venous channels, definition of permanent capillary areas, valve and septal formation, etc. The striking features of this ontogenetic history of the blood vascular system are: 1. The relatively late addition to the preformed non-cellular circulation of free cell elements, which, as the red blood cells, stamp, from the period of their liberation and inclusion in the circulating plasma, the resulting vascular system as haemal. 2. The common origin from mesoderm of both characteristic components of the haemal system, viz., the vascular endothelium and the red blood cell. Both the vascular intima and the free cell contents of the channels lined by this intima appear as highly modified derivatives of the same mesodermal cellular ancestors, which constitute the cell-strands of the earliest period of vascular development. In their first inception the systemic lymphatic vessels of the mammalian embryo, as distinguished from the jugular lymph sacs of venous origin, repeat in every detail the primary stages of the developing haemal capillaries, prior to the inclusion within the lumen of the latter, of the cellular contents of the blood islands. They can be identified as distinct struc- tures as soon as the blood channels proper have differentiated. Before that period direct observation cannot determine, in case of individual mesenchymal spaces, whether they are eventually to become part of the lymphatic or of the blood vascular system. 24 DEVELOPMENT OF THE* SYSTEMIC LYMPHATIC VESSELS It is therefore quite possible that in the mammalian embryo both sets of intercellular mesodermal spaces develop side by side and simultaneously, although prior to the setting free of the haemo- globin cells and their appearance within the lumen of the haemal capillaries, there is no absolute criterion which would serve to distinguish intercellular spaces as belonging definitely to either the haemal or the lymphatic division of the general vascular anlage. It hence appears to me futile to base serious conclusions in regard to the genesis of vascular structure on observations laade on the vessels seen in the transparent tails of living anure am- phibia. Phylogenetically, from the urodele standpoint, such larvae are adult organisms. We all know that, once established, all divisions of the vascular organization are, under the stimulus of normal or abnormal growth, capable of further increase and extension. The observations above referred to may offer, if correctly interpreted, interesting side lights on the method of vascular growth, but they cannot, in any valid sense, bear on the problem of vascular genesis, either haemal or lymphatic. But in mammalian embryos of the proper stages, and specifi- cally in embryos of the Cat between 10 and 12 mm., the first systemic lymphatic anlages are clearly differentiated in the cir- cumscribed areas of their first appearance, coincident with the definition of the early intra-embryonic blood channels. Thus in the omphalo-mesenteric district, and cephalad and caudad of this point, on each side of the aorta, isolated intercellular mesenchymal spaces appear at this period, closely applied to the walls of the neighboring venous plexuses of the postcardinal and mesonephroic veins, but not connected with the same. These intercellular clefts enlarge rapidly to form numerous oval or round spaces, closely interwoven with the venous network and later with the sympathetic anlages. The cells limiting these early lymphatic anlages become, with the further extension of the spaces which they line, flattened and assume typical endothelial characters. Fig. 1 shows a transverse section of a 10 mm. cat embryo (series 111, slide viii, section 4) magnified 175 diameters, cephalad RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 25 of the mesonephros and the subcardinal cross anastomosis, and gives a topographical view of the region in which the first meso- dermal intercellular lymphatic tissue spaces (77) appear, in the in- terval between aorta (73), postcardinal vein (68} and coelom (78}. Fig. 2 is the same section of this embryo, magnified 300 dia- meters, and shows the area to the left of the aorta. Between this vessel (73} and the well defined left postcardinal vein (68} dorsad, and the ccelom cavity (78} ventrad, appear a number of clear mesenchymal spaces (77) which do not communicate with the adjacent venous channels. Some of these spaces extend from the ccelom angle dorsad along the lateral aspect of the postcardinal. They are the first distinctly recognizable anlages of the lymphatic system, and they develop, from their first inception, as independent mesenchymal spaces, closely related to the adjacent haemal channels, but genetically independent of the latter. Fig. 3 shows the same field in a magnification of 600 diameters. The spaces are clearly cut, separate and distinct, and the limit- ing cells are beginning to assume endothelial character. The relation between the haemal channels and the develop- ing adjacent lymphatic spaces can be clearly traced in succes- sive sections of this same embryo proceeding caudad. Figs. 4, 5, 6, and 7 show, respectively, sections 7, 8, 9 and 10 of slide viii of series 111, magnified 300 diameters. In all of these sections the uniformity, and the distinct struc- tural character of the primary lymphatic tissue-spaces is clearly visible in the same situation and in identical relation to surround- ing structures. For comparison with the preceding series the same region is shown in another 10 mm. embryo (series 120, slide ix, sections 25 and 26) in two successive sections, magnified 300 diameters, in figs. 8 and 9. In these sections the same independent mesodermal spaces (77) are seen in their typical relation to aorta (73), post-cardi- nal vein (68} and ccelom cleft (75). The interesting question of the relationship between these early mesenchymal spaces and the coalom cavity can only be 26 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS considered superficially at this time. Yet, in some regions, very suggestive pictures are obtained. Thus in section 4 of slide viii of series 111 (figs. 1, 2 and 3) the distinct appearance of a communication between the coelom cavity proper (78} and the early mesenchymal spaces (77) above described is given by a clearly limited and well defined funnel-shaped stoma, occuping the dorsal extremity of the coelomic cleft (79 in fig. 2), and apparently opening directly into the spaces of the early lym- phatic plexus. The remaining sections of this series figured (figs. 4 to 7) confirm this impression. The conditions here described for early embryos of the cat strongly support the views expressed by Marcus 9 in his studies on the lymphatic development of Hypogeophis. The micropho- tographs here given should be compared with his description on pp. 599-601 of the paper quoted, with his text fig. 6, and with his figs. 5 and 6 of plate xvi. The early mesodermal spaces here described and figured are lymphatic in character and form part of the extensive temporary network of lymphatic channels which appears for a time during mammalian ontogenesis and which bears a close resemblance to the corresponding lymphatic organization in amphibia and reptiles. The peri-aortic lymphatic sinuses and the exaggerated subcutaneous lymph channels of the earlier mammalian stages are portions of this evanescent and reminiscent system, which subse- quently in large part retrogrades, and either disappears altogether or is extremely modified to meet the definite permanent condi- tions of mammalian lymphatic organization. Thus the early periaortic spaces become much reduced in course of further devel- opment. They then become associated, in a way presently to be described in detail, with elements of the axial venous plexuses of the mammalian embryo and form the anlages of the main segments of the thoracic ducts. This ontogenetic temporary recall of antecedent phylogenetic types of vascular development appears to be chiefly centered, in the mammalian embryo, in the region around the omphalo-mesenteric artery, where, in the adult, the definite and permanent lymphatic trunks closely resemble in their arrangement the peri-omphalo-mesenteric annular veins of certain reptilian embryos. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 27 In the succeeding stages the mesodermal lymphatic anlages assume, in large part, the Very definite relation to certain embry- onic venous channels, which led McClure and myself to describe them in 1906 in our preliminary account above quoted, 21 as "extraintimal" or/'perivenous" structures. This relationship is of two kinds: A. Total replacement of temporary embryonic veins by extrainti- mal lymphatic channels. In t"he earlier embryonic stages the areas of the future definite venous channels are largely occupied by an antecedent venous or capillary network, out of which, along definite hydrostatic lines, the subsequent veins develop by confluence of the plexus ele- ments occupying these lines. 29 Parts of the early capillary reticu- lum, not thus included in the path of the definitely organized venous trunks, remain, after the latter have become established, as a perivenous plexus. Some of the elements of this secondarily established plexus develop into permanent tributaries of the main veins. Others undergo a process of separation from the permanent functional channels and degenerate. In many cases their blood-cell contents break down and are eliminated, while their endothelial lining appears to revert to the indifferent type of the embryonic mesodermal cell. Thus in embryos between 13.5 and 16 mm. many striking instances of this reversion are to be observed. Thp former vascular channel appears as a collection of clearly differentiated and very highly stained mesodermal cells. Figs. 47, 50, 51, in Part II, show these mesodermal vascular derivatives very clearly. They form the dark masses seen in the field dorsal and dorsolateral to the O3sophagus and in the peritracheal region. In many regions of the mammalian embryo, however, these detached and retrograding venous elements do not attain this condition, but in an earlier stage, constitute lines around which the most active primary lymphatic organization of the mammalian a * H. v. W. Schulte and Fred. Tilney : "Note on the Organization of the Venous Return, with Especial Reference to the Iliac Veins." Anal. Record, vol. iii, no. 11, 1909. 28 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS embryo centers. The intercellular mesodermal clefts above described develop especially along and around these decadent venules and finally envelop them. As the result of this process the lymphatic anlages appear in certain mammalian ontogenetic stages, in large part, as distinct spaces enclosing the remnant of the embryonic vein. The latter may still, for a time, contain a few degenerating red blood cells, but these soon disappear, and then the entire anlage is formed by a collapsed and empty en- dothelial tube, the abandoned channel of the earlier vein, sur- rounded by a second endothelial tube, formed by the confluence of the independently developed extraintimal or peri venous meso- dermal spaces. As these spaces enlarge and join each other their lumen increases, and the limiting cells become flattened and assume typical endothelial characters. The height of this phase of lymphatic development is reached in embryos of the cat in the 14 mm. stage, and numerous demonstrations of the appearance of the structures on section are given in Part II of this paper. The remnant of the embryonic vein bears a relation to the replacing perivenous lymphatic channel which is exactly the same as that of a collapsed inner tube to the enveloping shoe of a pneumatic tire. The inner skin of the shoe and the rim of the wheel repre- sent the lymphatic intimal endothelium. The space between the shoe and the collapsed inner tube is the lumen of the future lymphatic channel. The empty inner tube itself is the decadent embryonic vein upon and around which the secondary lymphatic channel is built. In the course of further development the venous remnant disintegrates and disappears, leaving a clear lumen for the lymphatic vessel, which thus completely replaces the earlier vein and comes to occupy absolutely the topographical position of the latter. Often the replacing lymphatic begins as an extraintimal channel partially surrounding the receding vein. This leads in course of further development to an expansion of the lymphatic channel not concentric with the axial line of the shrinking vein. The remnant of the vein then retires to a point on the intimal surface of the new lymphatic channel, and appears to project into the lumen of the latter. The resulting histological picture RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 29 will depend on the plane of section in reference to the course of the lymphatic and its contained venous remnant. Thus, as shown in the following schema, many sections give the appear- ance indicated in 1. If the plane of section should, however, lie in the line A B, it will divide the shrinking vein (4) and the envel- oping lymphatic (5) in such a way as to produce the picture shown under 2. In other cases the lymphatic spaces unite around the entire circumference of the abandoned venule, and the lumen is then contained for long distances entirely within the lumen of the replacing lymphatic channel. The process just described is remarkably constant and uniform in the critical stages of mammalian lymphatic development. As can be readily seen in following the individual sections in the microphotographs published in Part II, the significance of the conditions here shown is unmistakable. This is not a hap- hazard process, observed only occasionally, in a limited number of embryos, and then only in single sections, or, at most, in a few successive sections. In any average embryo of the proper age the same structures appear regularly in the same situations and in identical relationship to the embryonic environment. It is often possible, as the microphotographs and the corresponding 30 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS reconstructions herewith published clearly show, to trace the forming lymphatic with its atrophied vein kernel for long stretches, and in different embryos of approximately the same crown-rump measure the consistent repetition of identical histological pictures is remarkable. There are, of course, as in the ontogeny of other structures, individual cases of variation in which systemic lymphatic develop- ment is either more advanced or more retarded than is normal for the average of any given stage. But if a large number of embryos of each typical period are examined and compared, the average standard of extraintimal lymphatic development attained by the majority of individuals in each stage is remarkably con- stant and uniform. The earliest stage in which I have encountered this typical replacement of an early embryonic vein by a perivenous extrainti- mal lymphatic is presented by certain 12 mm. cat embryos along the caudal circumference of the azygos-precardinal confluence. In the concavity of the azygos arch on each side, as this vessel turns ventrad to join the precardinal vein, these earliest evidences of the extraintimal replacement of preceding embryonic veins by independent perivenous lymphatic spaces are encountered. Thus, fig. 10 shows a section of this region in a 12 mm. cat embryo (series 217, slide x, section 12). Here the typical pic- ture of the central collapsed core of the earlier vein (Jf), enveloped by the clear extraintimal lymphatic space (5), is plainly to be seen. Fig. 11 shows the same region in another 12 mm. embryo (series 211, slide x, section 15). Both the degenerating venule (4) and the enveloping lymphatic (5) are larger than in the pre- ceding series, and occupy the same position between aorta (7) and left precardinal vein (6). In a 14 mm. embryo (series 127, slide viii, section 12, fig. 12) the section passing just caudad to the junction of left azygos (6) with left precardinal vein (6} shows these early lymphatic spaces (5} and their relation to the contained venous remnant (4) fully developed. Only one of the areas is denoted by leaders in the figure, but two equivalent areas are seen further dorsad and nearer to the ventral aspect of the azygos arch. The lym- RELATION OF LYMPHATIC TO BLOOD- VASCULAR SYSTEM 31 phatic plexus, as development proceeds, from the 12 mm. stage on, approaches the large venous trunks more and more, until the spaces lie in direct apposition with the same, and unite to form the lymphatic trunk eventually destined to replace the left azygos arch and adjacent position of the left precardinal vein. This trunk then constitutes the cephalic end of the broncho- mediastinal duct (37). Thus figs. 13 and 14 show two sections through the same region in a 15 mm. embryo (series 219, slide xiv, sections 19 and 16). Compared with the 14 mm. embryo .the left azygos vein shows a marked reduction. The lym- phatic spaces have enlarged and present a clear lumen on section, the remnant of the earlier vein, around which they developed, having disappeared. The spaces lie between aorta and left precaval vein, in close approximation to the dorso-medial circumference of the latter. The azygos segment of the thoracic duct (86) is seen dorsal to the interval between aorta and oesoph- agus, and ventral to the scant remnants of the earlier interazygos venous anastomosis (15). The reconstruction of a 15 mm. embryo (series 218) shown in fig. 170 in Part II, gives a clear idea of the extent and relations of this lymphatic complex (37) in this stage. B. Partial replacement of portions of the territory of an early embryonic venous pathway by an extraintimal lymphatic vessel, both venous and lymphatic channels either persisting side by side up to later developmental periods, or forming correlated components of the permanent adult vascular organization. The developmental processes just described appear most clearly marked in the earlier stages, and in connection with temporary embryonic channels and plexuses which are destined to undergo rapid degeneration and ultimate complete elimination. In the case of the embryonic veins which are retained for a longer period, or carried over into the permanent adult organization, the his- togenetic stages of lymphatic development are identical in kind, and differ only in degree from those just detailed. In place of complete replacement of the antecedent vein by the lymphatic channel, this replacement is only partial and leads to the typical 32 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS close parallel association of lymphatic and venous vessels so characteristic of the later embryonic stages and of the per- manent adult pathways of both the lymphatic and the venous systems. Figs. 15 and 16 show transverse sections of a 14 mm. embryo (series 222, slide vii, section 26, and slide viii, section 4) in the region of the external jugular vein. The lymphatic spaces (27'} are in full development, and are applied chiefly to the medial aspect of the plexus of the external jugular vein (27}. The lym- phatic endothelium is clearly marked. Fig. 17 shows a transverse section in the upper thoracic region of another 14 mm. embryo (series 37, slide xiii, section 12). A typical lymphatic anlage (53} is applied to the medial wall of the left precaval vein. Fig. 18 shows a transverse section through the mid-thoracic region in a 17 mm. cat embryo (series 258, slide xviii, section 9, X225). The extraintimal anlages of the thoracic ducts (36} , which usually in thisstage have advanced to the production of a continuous and uninterrupted channel system, are seen on each side closely ap- plied to the ventral aspect of the left and right azygos veins. The latter are in the height of their development, forming large and symmetrical longitudinal venous trunks (8, 6} connected by the supra-aortic interazygos anastomosis. In course of further development the left azygos vein and the interazygos anastomosis are destined to undergo progressive reduction until they are eventually entirely lost. Their topographical position is then occupied by the replacing left segment of the thoracic duct com- plex (86}. The right azygos vein, which is carried as a perma- nent vessel into the adult organization, also undergoes consider- able relative reduction, correlated to the corresponding increase in the caliber of the main (right) segment of this portion of the thoracic duct complex. The beginning of this process is seen well in fig. 19, which shows a transection of the same region in a 19 mm. cat embryo (series 253, slide xxiv, section 9, X 225). The change from the preceding RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 33 stage (fig. 18) is marked both in the venous and the lymphatic channels. The former are relatively much reduced, while the latter have correspondingly increased in extent. The azygos is mainly represented by the right channel (3} . The left channel (0) has become small, but is still connected by the transverse inter- azygos anastomosis (15} with the larger and permanent right trunk. The right thoracic duct (86} is likewise large. A dorso- medial extension of the same, which carries into the interval between aorta and interazygos plexus, will, in later stages, replace the latter secondarily, after the complete recession of the left azygos vein. The left thoracic duct (86} is also of large size and fills a considerable part of the area formerly (fig. 18) occupied by the left azygos trunk. Some decadent remnants of the ventro- medial azygos plexus (4) are still seen associated with the left duct, and are in process of replacement by a lymphatic space (5} destined to make connection between the left thoracic duct anlage and the mesenteric lymphatic plexuses (cf. p. 148, figs. 266 to 270,5.7). Figs. 20 to 24 show transverse sections in the region of the devel- oping mesenteric lymphatics and of the ascending lumbar lym- phatic trunks in a 17 mm. embryo (series 258, slide xxiii, sections 34, 33, 32 and 31.) The embryonic veins (74) occupying the root of the mesentery caudal to the subcardinal cross-anastomosis are in the process of being replaced by extraintimal lymphatic spaces (51} which are destined to become confluent and form the anlage of the future mesenteric lymphatic sac. The region in question is not only interesting in reference to the ontogenesis of the abdominal lymphatic sacs and channels, but the arrangement of the periaortic axial venous trunks and their relation to the developing lymphatics is, in combination with the next following stage (20 mm., figs. 25 and 26), of the highest importance in interpreting the phylogenetic relations of the main abdominal veins in mammalia. For these reasons a somewhat more detailed consideration of the sections may properly be in- troduced here. Fig. 20 shows section 34 of slide xxiii of series 258 in a magnifica- tion of 75 diameters and affords a comprehensive picture of the entire 34 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS topographical field involved. The strands of the sympathetic (l) are supra-aortic. On each side of the aorta (7) are the large and sym- metrically developed right and left postcardinal veins (67 and 68} , with the ascending trunks of the ilio-lumbar arteries (A.ilio-lumb. transv. ant.) (not labelled in the figure) applied to their ventro- lateral circumference. The subaortic area shows the cross-sections of four symmetrical vascular channels, two venous and two lym- phatic. Immediately vential to the aorta, and closely applied to its ventral wall and to each other, are two longitudinal parallel axial veins which are connected at intervals by a few short trans- verse anastomoses. These vessels are the temporary and very evanescent homologues in the placental embryo of the channels which McClure 30 has described as the "cardinal collateral trunks" in the embryo of Didelphis marsupialis, and from which he has traced the development of the preaortic postcava characteristic of the Marsupalia. These vessels are derivatives from the earlier preaortic cardinal-subcardinal venous plexus below the cross- anastomosis, but differentiate in Marsupials along separate and distinct axial lines. They are destined, as are the corresponding portions of the subcardinals proper, to be entirely replaced in the typical placental development by the chain of preaortic lymph channels and nodes, but are capable, in aberrant types among the placentalia, of yielding, by further and continuous development, a type of preaortic postcava which in every respect corresponds to that encountered in Marsupials. McClure 31 has described this condition in Tragulus, and his observation has been confirmed in a number of dissections by Beddard and others. The fortunate acquisition recently of a series of Tragulus embryos, through the kindness of the officials of the Smithsonian Institution, has enabled my associate Tilney to trace, in a publication now in preparation for the press, the development of the venous and lymphatic systems in this aberrant ungulate in their mutual interdependence, and to show the correspondence of the venous genetic processes * C. F. W. McClure: "The Anatomy and Development of the Postcava in Didelphis marsupialis." Am. Jour. Anat., vol. v, 1906. 11 C. F. W. McClure: "The Postcava of an Adult Indian Chevrotain (Tragulus meminna, Erxleben). Anat. Am., Band, xxix no. 13 and 14, 1906, pp. 375-377. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 35 with those observed by McClure in Didelphis. We have here, therefore, an undoubted instance in which, in an ungulate placen- tal, the organization of the abdominal venous and lymphatic com- plex dates back, phylogenetically, to the period of a common mam- malian, or even amniote, vascular groundplan (text figure A, p. 41), forming the starting point from which all types of normal and variant postcaval organization, monotreme, marsupial and placental, radiate. This makes it possible, as in the ungulate genus just mentioned, to find in otherwise valid placental forms, as a constant and normal structural character, the postrenal segment of the postcava as a single median preaortic vessel, which receives, usually, both sex veins, and corresponds in every respect to the characteristic marsupial venous type. This channel is the final and permanent product of the fusion of the two embryonal cardinal collateral veins, which, in the typical placental, either do not develop at all, as distinct veins, or else, as in the cat, appear during the ontogenseis only for a very short period, and are sub- sequently entirely replaced by extraintimal lymphatic spaces eventually uniting to form the extensive right and left ascending lumbar lymphatic trunks. These vessels (75} are seen in the prep- aration under discussion on each side, ventro-lateral of the aorta in the space between the postcardinal veins (67 and 68) and the cardinal collateral veins (74). These large lymphatic channels have been developed through the fusion of numerous separate extraintimal spaces surrounding and finally replacing some of the elements of the earlier preaortic venous plexuses. A condensed portion of the latter still persists in this embryo as the cardinal collateral veins (7Jf) above referred to. Eventually these also are destined to undergo, in course of normal development, complete extraintimal replacement by lymphatic channels. These latter thus come to occupy in the typical placental secondarily the ter- ritory which in monotremes, marsupials and atypical placentalia (Tragulus) is filled by the postrenal preaortic postcava (cardinal collateral). It is interesting to note in this connection that a detailed exami- nation of over 900 adult cats failed to reveal a single instance in which the postrenal segment of the postcava was formed through 36 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS persistence and further development of the cardinal collateral veins. In other words, not a single individual in the entire series possessed a marsupial postcava, although every shade in the possi- ble range of variation in the district of the post- and supracardinal lines was represented by numerous examples. Thus with an ab- dominal venous organization of very unstable equilibrium, as shown by the large percentage of cardinal variants, 24 the cat yet keeps entirely within the district of the common genetic ground plan assigned to the placentalia. This phylogenetic consistency is maintained in spite of the fact that, as just demonstrated in series 258, the embryo develops the raw materials, as cardinal collateral channels, out of which a preaortic postcava of the marsupial type could be evolved. In my own estimation cats possessing this form of postcava may exist and may eventually be found. But the failure to encounter them in the relatively large series of adults already examined speaks volumes for the value of vascular organ- ization in interpreting phylogenetic relations. In this light the postcaval development and adult structure of Tragulus, for example, are of the utmost importance and signifi- cance. The unprejudiced observer is often forced to wonder why some exponents of palaeontological research are content to draw far-reaching phylogenetic conclusions from the remnants of the incomplete locomotory apparatus at their disposal, without utilizing the results of modern comparative anatomical and embryo- logical investigation in determining at least the mutual relations of the extant forms, massed by an ironclad taxonomy into more or less questionable groups, whose ancestry and derivation form one of the primary problems of the palaeontologist. The case of Tragulus just alluded to, the parotid complex and alimentary canal of Hyrax, the amniote homologies of the derivatives of the Sulcus buccalis determined by Schulte, the sharp line of lymphatic demarcation recently shown by Silvester to structurally separate absolutely the platyrrhine and catarrhine divisions of the lower primates, these and other facts are only instances in which the inadequacy of a superficial convergence of dental and skeletal characters, for the purpose of establishing valid phylogenetic relations, is revealed by cardinal divergence in the far more stable and important organization of vascular and visceral structure RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 37 The prevalent lack of coordinate deduction between verte- brate morphology and palaeontology is accentuated 'by contrast with such publications as Weber's " Saugethiere " and, more recently, W. K. Gregory's "The Orders of Mammals", based largely on the author's joint work with Osborn, and on the latter's previous classical researches, and published by the American Museum of Natural History (Bulletin, vol. 27, February, 1910). Workers in the general field of vertebrate structure appreciate fully the immense practical value in their own special investi- gations of books of this type, in which, to quote Gregory's words, "the data of systematic mammology, of comparative anatomy and embryology shall ultimately be integrated with the data of pal- aeontology, to the great advantage of these now more or less independent lines of study." In embryo 258 (fig. 21) the space ventral to the cardinal col- lateral veins and the ascending lumbar lymphatic trunks, is oc- cupied by a plexus of mesenteric lymphatics (51} draining into the latter. On each side are seen sections of the ureter (58}, and further laterad of the iliac vessels (61}. Ventral of the intestine (62} are the Wolffian ducts (64) and the cloaca (63} , with the hypogastric arteries (66} laterally. Fig. 21 shows the important central vascular region of the same section in a magnification of 150 diameters. The relation of the cardinal collateral trunks (74) to the ascending lumbar lymphatics (75), and of the latter to the postcardinal veins (67, 68} can be more clearly seen. The lumen of the mesenteric lymphatics (51} still contains in places remnants of the decadent venous plexus around and upon which they developed as replacing ex- traintimal spaces. The three succeeding sections, tracing the structures caudo- cephalad, are shown in figs. 22, 23 and 24, all in a magnification of 150 diameters. The connections of the lymphatic channels developing along the iliac vessels with the ascending lumbar trunks are especially well seen on the left side of the three figures, also the anastomoses between the two cardinal collateral veins in figs. 23 and 24. Figs. 25 and 26 show transverse sections of the hinder end of the body in a 20 mm. embryo (series 241). 38 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS Fig. 25 (series 241, slide xxx, section 4) gives, in a magnifi- cation of 75 diameters, a topographical view of the entire field. This stage, compared with the preceding 17mm. embryo, is marked by the full development of the supracardinal venous line, respon- sible for the production of the greater portion of the typical adult placental postcava below the renal level, and by the cor- related development of the supra- or retro-aortic lymphatic sinuses associated with the same. The periaortic area in fig. 25 gives a clear view of the vascular relations and of the postcardinal and supracardinal axial venous trunks. The former (67, 68,} are seen on each side, between aorta (7) and metanephros (65), receiving the veins from the mesonephroi in whose dorso-medial border they are lodged. The latter (59, 60) lie dorsal to the aorta (7) between this ves- sel and the sympathetic strands (1). The right supracardinal (60) has already gained the ascendency and is in process of establishing the channel of a normal right retro-aortic postcava, which is the typical vein for the cat. The correspondingly reduced left supracardinal (59) occupies the same situation on the left side. Associated with the supracardinal venous channels are the supracardinal lymphatic trunks (76), which form the anlages of the main adult retro-aortic lymphatic plexus. These develop as extraintimal spaces replacing portions of the earlier supracardinal venous reticulum. In accordance with the normal type of development observed in this individual embryo, the large permanent supracardinal (postcaval) vein of the right side is accompanied by a relatively small lymphatic channel (76 right), while on the left side the much reduced left supracardinal (59) is already nearly replaced by the correspond- ing lymphatic vessel (76 left). In course of further normal devel- opment this replacement will become complete and then the area formerly occupied by the left supracardinal vein will be entirely filled by the substituted large left retro-aortic lymphatic. The permanent functional venous channel of the right side (60) on the other hand, developing into the typical placental post- cava, will be accompanied by a relatively small right lymphatic DELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 39 trunk following its dorso-lateral aspect. As will be shown later (PartV.), departures from the normal type of venous develop- ment in this region, and the substitution of other embryonic pathways for the right supracardinal in building up this section of the adult postcava, produce corresponding and correlated changes in the arrangement of the main retro-aortic lymphatic channels. Fig. 26 shows the periaortic region of the same embryo, further caudad, in a magnification of 150 diameters (series 241, slide xxx, section 14.) The section is taken at the level of a pair of dorsal intersegmental arteries which pierce the supracardinal venous (59, 60) and lymphatic (76) complex, and divide the vein of the right side into two components (60, 60). Further ventrad the two postcardinal veins (67, 68) are seen, already consider- ably reduced, ventrolaterad to the ureters (58) and the accom- panying ascending' lumbar lymphatic trunks (75). The connec- tion of the latter with the supracardinal lymphatic channel is especially clear on the right side of the embryo. Taken together, the 17 mm. and 20 mm. embryos just figured and described afford a very clear and comprehensive picture of venous and lymphatic development in their mutual relationship in this region. The schematic text figures A, B and C may help to explain this relationship. Fig. A is based on the joint studies which McClure and I made on the development of the postcava in embryos of the domestic cat. The figure was demonstrated to the 21st Session of the Association of American Anatomists in 1906 at the time of pre- sentation of the communications, although not reproduced in the brief abstracts of the papers subsequently published. 23 ' 25 The figure represents a composite schema of the main peri- aortic venous axial pathways of the abdominal region. These pathways developing along definite and constant axial hydro- static lines out of the periaortic venous reticulum,.have all been determined by us in embryos of the cat. They do not, of course, all coexist at the same time in any embryonic stage, but normally succeed each other in definite sequence. The entire range of 40 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS extensive variations in the domain of the adult postcava of the cat can be clearly interpreted genetically 25 on the basis of this common groundplan, through abnormal persistence of one or more of the embryonal pathways usually destined for complete obliteration, thus producing farreaching modifications in the structure and relations of the resulting atypical postrenal seg- ment of the adult postcava. This periaortic axial venous lattice with connecting trans- verse branches (Fig. A) contains four components on each side , which develop in the following order : 1 . The postcardinal veins (1 ) . 2. The subcardinal veins (2}. 3. As secondary derivatives of these two veins, the preaortic cardinal collateral channels (3). 4. As secondary dorsal derivatives of the postcardinal trunks, the supracardinal veins (4). In the course of normal venous development along the line typical for the great majority of placental mammals the right supracardinal vein (4) obtains the preponderance and furnishes the postrenal segment of the adult postcava, thus freeirg the ureter from its primitive retro-venous position. A part of the early capillary periaortic reticulum, out of which this vessel develops, is secondarily replaced by extraintimal lymphatic spaces, which through their confluence form the rela- tively small retro-aortic lymph channel (4'), folio wing in the adult the dorso-lateral circumference of the postcava (4). (Figs. B and C). The right and left postcardinal veins (1} are in part retained as the terminals of the sex veins, in part replaced by the accom- panying lymphatic trunks (!'}. (Figs. B and C). The left supracardinal vein (4), and both subcardii^, 1 veins (2}, below the cross-anastomosis, as well as both cardinal collat- eral veins (3), retrograde and are entirely replaced secondarily by lymphatic channels. The lymphatic replacing the left supracardinal vein (4 ) forms normally in the adult the main retro-aortic lymphatic sinus. (4', lef t, in figs. Band C). RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 41 "\ Monotreme Ureltei Sex vei Sex vein 42 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS The lymphatic channels replacing the subcardinal and cardinal collateral venous trunks form the extensive system of the adult ascending lumbar and preaortic lymphatic vessels and nodes (2 f and 3' in figs. B and C). Of course it is quite apparent that the adult placental differ- entiation occurs in the district of the post- and supracardinal lines, with a strong predilection for the right supracardinal as the main path of the postrenal segment of the adult postcava. It is equally apparent that in correctly valuing the significance of the departures from the normal type of placental postcava all four of the available components, viz., right and left post- cardinal and right and left supracardinal lines, must be taken into account as potential factors in the development of the atyp- ical placental postcava. The relation of the veins to the ureter will then decide the question of the genetic derivation, as being either the persistent postcardinal or supracardinal channel of either the right or left sides, in the case of single trunks, or of both sides in instances of double bilateral adult channels. Thus all the recorded cases of variant postcaval veins of the cat, and of man, can be clearly interpreted on this basis, as has been done by McClure, Darrach and myself in previous publi- cations. 23 ' 24 ' 25 Furthermore, the placental types in which a normally so-called double postcaval vein occurs, as, e.g. in some of the aquatic carnivores, some insectivores and edentates, are readily led back to persistence of both right and left axial channels with absence or reduction of the iliac anastomosis. Again the position of the ureter in reference to the bilateral trunks will characterize each of them as being either post- or supracardinal in derivation. In the marsupials McClure's researches already quoted, 31 show clearly that the members of this subclass depend upon the continued development of the ventral preaortic venous path- ways (2, 3} of the common vertebrate groundplan (fig. A) for the evolution of their typical ventral preaortic postrenal seg- ment of the postcava, with consequent reduction of the post- cardinal line (.7) to the role of a sex vein terminal, and the complete suppression of the typical placental supracardinal lines (4) in most RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 43 forms, while rudiments of the same are found, as shown by Schulte's and Tilney's observations, in connection with the venous return along the caudal vein only in the Macropodidae. 29 Schulte's work on the venous organization of the Monotre- mes 32 proves, I think, conclusively the phylogenetic value of the periaortic venous reticulum with axial pathways of hydrostatic selection which the detailed study of placental embryos establish. His dissections of both Platypus and Echidna revealed for the first time the persistence in the Monotreme of both dorsal supra- cardinal channels, and of ventral preaortic vessels of subcardinal and cardinal collateral derivation, while, as appears uniformly throughout the mammalian class, the primitive postcardinals retain normally solely the function of venous drainage for the gonad. The facts just stated have been in a large part already pub- lished in outline by McClure, Schulte, Tilney, Darrach and myself. I hope it will be possible to collect the numerous obser- vations, with adequate illustrations, in a publication to be issued in the near future. I have recorded some of the results obtained through these joint investigations in this paper in order to use them for the purpose of clearly outlining on a broad basis the genetic possibilities in the development of the venous system, and the correlated interdependence of the systemic lymphatic vessels. The latter will, in the mammal, take over and further develop territory formerly occupied by transient embryonic venous channels, which they secondarily replace through extra- intimal development. Hence the resulting mirror picture which the lymphatic system of the adult mammal presents in reference to the axial venous trunks, whatever type of central venous organ- ization may obtain in any individual instance. The examples just given could, df course, be indefinitely multi- plied. They all show absolutely congruent, uniform and constant pictures in all parts of the body of the developing lymphatic chan- nels in close association with the adjacent veins, but not connected 32 H. v. W. Schulte: "The Range of Variations in Monotremes and Australian Marsupials." Anat. Rec., no. 3, April 1, 1907. Am. Jour. Anat., vol. vi, no. 3, 1907. 44 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS with the same. Nowhere is there any suggestion of a bud or an outgrowth from the vein as forming the origin of these lymphatic spaces. It now remains to clearly prove the genesis of these spaces, and to trace their growth from their inception up to the stages just pictured in which fully organized lymphatic and ve- nous channels lie side by side in the mutual relation above fig- ured and described. The proof of their origin is furnished by the series of microphotographs of successive sections of the earlier stages given in Part II of this communication, in connection with the individual series described and figured in tracing the develop- ment of the preazygos and azygos portions of the thoracic ducts. The microphotographs, and especially the reduced reproductions figured, are not so clear as the actual preparations, because focal adjustment is required to follow the endothelial lining of the spaces in their entire circumference, and because they lack the differential stain of the sections. Still they are sufficiently dis- tinct to establish definite conclusions. Merely referring, there- fore, at this time to the following detailed illustrations, the general topic of extraintimal replacement of embryonic veins by lym- phatic spaces and the character of the latter deserve some further consideration. The lymphatic anlages, as above stated, if studied under suffi- ciently high power and with some care, are seen to begin as inter- cellular clefts in the periaortic mesoderm, adjacent to the postcar- dinal venous plexus, and chiefly on its ental aspect, between it and the aorta. The individual lymphatic spaces, at first small and separated from each other, enlarge, elongate and become confluent, to form larger continuous channel segments, while innumerable newly formed spaces of the same character appear in the surrounding tissue, join with each other, and with the earlier preformed lym- phatic channels, in exactly the same manner, and with the same appearance of lymph endothelial "budding" or " sprouting" as is observed in haemal vascularization of new areas by the junction of the earlier blood capillary anlages with secondary haemal plexuses. In these later stages the veins are surrounded by a close lymphatic plexus, which, however, does not as yet form a con- RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 45 nected channel system, but is composed of longer and shorter segments still independent of each other. These finally become confluent, to form the main systemic lymphatic collecting trunks, and then only do these establish their final junction with the jugular lymph sacs, through whose interposition, as above stated, they gain in the typical mammal their permanent entry into the venous system. In this ontogenesis of the systemic lymphatic vessels certain relations between them and the venous system deserve further notice. In the early stages the lymphatic mesenchymal spaces form a wide meshed network (cf. series 111, figs. 1 to 7, series 120, figs. 8 and 9). There is thus a marked similarity in the earliest stages of both the haemal and the lymphatic vessels, for the peripheral venous embryonic pathways are in their corresponding stages like- wise still largely in the condition of a capillary reticulum. As the main lines of venous drainage crystallize out of the antecedent plexiform arrangement, the adjacent enlarging lymphatic channels crowd in on the condensing venous line and continue the close re- lationship which the earliest lymphatic anlages maintain to the adjacent veins. Thus the main embryonic venous channels de- velop along certain definite hydrostatic lines by enlargement and confluence of the individual plexiform elements of the indefinite antecedent network occupying these lines. The capillaries out- side of these lines retrograde, so that the area of crosssection of the defined venous channel is less than the cross-cut area of the plexiform network which it replaces. 29 The distinct impression is given that the space thus vacated by the condensation of the plexiform venous network of the ear- lier stages affords to the replacing lymphatic plexus the oppor- tunity for greater growth and expansion, and that subsequently, in repetition of the process previously active in the venous reticu- lum, the lymphatic network condenses in a similar manner into more defined channels along similar hydrostatic drainage lines, so that the newly established main lymphatic vessel now closely follows the main venous channel. It is to be noted, however, that this organization of main vascular channels is usually less 46 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS complete in case of the lymphatic vessels, as compared with the corresponding vein. The lymphatic system retains, much more perfectly than the venous, in many situations the original embry- onic plexiform type. At first the cells limiting the earliest lymphatic spaces are of the usual irregular cuboidal form. As the spaces enlarge, open out and thus become better defined, the limiting mesodermal cells become flattened, and finally assume a typical endothelial char- acter and form. Thus, for example, the endothelial lining of the primitive mesodermal lymphatic spaces (77) is more clearly devel- oped in the 10 mm. embryo, series 120, shown in figs. 8 and 9, than in the corresponding sections of embryo 111, of the same crown-rump measure (figs. 1 to 7). The former embryo is slightly in advance of the latter as regards the development of the parie- tal endothelial lining of the primitive mesodermal intercellular lymphatic spaces. In some instances a few modified mesodermal cells intervene between the cells limiting the lymphatic spaces and the endothelium of the adjacent venous radicle. In others no such intervening cell-layer exists, and the lymphatic space is separated from the venous lumen only by the latter 's endothelial wall. In other words, in the extent of the lymphatic anlage, a single-celled membrane furnishes a part of the venous intima and at the same time contributes to the endothelial definition of the lymphatic space. This relation of vein and lymphatic anlage is shown very clearly in fig. 17. The lymphatic space (53), which is closely applied to the medial wall of the left precardinal vein, is only separated from the lumen of the latter by the endothelial membrane which serves to line both spaces for the area of their mutual contact in this stage. Subsequently, with the regression of the left precardinal vein, this lymphatic anlage will correspond- ingly enlarge to form an extensive lymphatic plexus, which will eventually topographically replace the vein along which it arose. In order to briefly characterize this relation between vein and lymphatic, McClure and I defined in an earlier publication 21 these spaces as the "Extra-intimal" anlages of the systemic lymphatic vessels, with due regard to the relation existing between them and the intimal endothelial lining of the embryonic veins. Themechan- RELATION OF LYMPHATIC TO BLOOD- VASCULAR SYSTEM 47 ical concept involved in this term seems, to judge from a recent publication, 33 to have been difficult to acquire. I am glad to be able to make myself clear by reference to fig. 17, where the mutual relation of the two vascular lumina is evident without further description, and to the numerous detailed illustrations on a larger scale of magnification which accompany the account of the devel- opment of the thoracic ducts in Part II of this communication. By far the larger number of the early lymphatic channels are the product of fusion of these " extra-intimal " spaces, and hence closely follow the veins of their respective regions. Subsequently, with the development of a venous adventitia, this relationship is somewhat altered in case of those veins which are included in the permanent venous organization. The close relation existing, however, throughout life between these veins and the accompany- ing lymphatics is based on this intimate primitive association of their respective anlages. On the other hand, the extra-intimal position of the earliest lymphatic spaces furnishes the explanation of another relation manifested between the developing systemic lymphatic channels and those embryonic veins which in course of normal venous devel- opment are destined to undergo reduction and finally complete suppression, when the primitive bilateral and symmetrical venous system of the earlier embryonic stages shifts to the dextral as- symmetrical type of the main adult axial channels. In these circumstances the systemic lymphatic vessel associated with the temporary embryonic vein experiences, apparently through the shrinkage of the latter, an impetus to its own more extensive de- velopment, so that it comes to occupy in general topographically the space filled by the vein in the earlier stages. Thus the embryonic period which marks the normal ontogenetic swing of the main venous line to the right through the secondary sinistro-dextral iliac, hemiazygos and brachiocephalic cross anas- tomoses, sees the simultaneous increase in the corresponding lym- phatic channels of the left side, which topographically replace the abandoned left embryonic venous pathways of the earlier and sym- 33 Sabin: Anat. Rec., vol. ii, 1908, p. 50. 48 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS metrical stage. This occurrence leads to the well-known relative location of the main axial veins and lymphatics in the normal adult, in which the lymphatic vessels are chiefly situated on the left side and form, so to speak, a mirror-picture of the right sided axial venous channels. Fig. 27 shows the reconstruction of the anterior venous and lymphatic complex in a cat embryo of 18 mm. (series 88) in the ventral view, and fig. 28 of the same preparation in the lateral aspect from the left side. The brachiocephalic cross anastomosis is already well under way, resulting in a marked diminution of the left anterior caval vessel and a corresponding increase in the permanent right anterior cava or right duct of Cuvier. Con- versely, the lymphatic vessel accompanying the diminishing left precaval vein is of large size, while that applied to the massive right precaval is comparatively small. This principle of lymphatico-venous replacement, indicated clearly in the later embryonic stages, is strikingly illustrated in the adult. Thus, for example, the adult cat presents normally the arrangement of the great veins of the head and neck which is so frequently encountered in mammals below Primates, in which the large embryonic internal jugular vein is much reduced or even entirely obliterated, while secondarily the external jugu- lar vein has assumed the function of the main vessel. Under these normal circumstances the lymphatic trunk accompanying the minute internal jugular vein or, in case of its entire default, occupying its position, is well developed and the largest element of the entire cervical lymphatic complex, while the external jugu- lar vein is, on the other hand, accompanied usually by two very slender lymphatic vessels. In instances, however, in which the embryonic proportion between the two jugular veins is retained in the adult, so that the internal jugular appears as a large and functionally important vessel, while the external is correspondingly diminished, the inter- nal jugular lymphatic trunk is reduced, while the double lymphatic vessel along the external jugular is enlarged, and evidently acts in compensation in the cervical lymphatic return. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 49 Again, in the same way, in adult animals with normally placed right postcava, the main supracardinal lymphatic trunk, draining the abdomen and the posterior extremities, follows the left side of the large artery. In the not infrequent instances, however, of left sided postcava or postcardinal vein in the adult the reverse obtains, and the peri- aortic lymphatic channels predominate on the right side and oc- cupy the place usually filled by the large vein in normal venous development. I have encountered in the adult series so far examined no in- stance of persistent left precava replacing functionally the normal right superior cava, but have no doubt that this venous variation would involve a transposition of the proximal end of the thoracic duct to the right side, or at least a marked increase in the size and functional importance of the usually insignificant preazygos seg- ment of the adult right lymphatic duct. The developmental outline just given describes the mutual ontogenetic relations of the venous and lymphatic systems through- out the greater part of the body. Systemic lymphatic development in these regions is, however, by no menas confined to the immediate environment of degener- ating embryonic veins. The same field, which shows the above described histogenetic processes in the development of extrain- timal lymphatic spaces surrounding and replacing a decadent venule, will at the same time contain numerous equivalent lym- phatic mesenchymal clefts and spaces which continue to develop independently of any association with retrograding veins. Nat- urally, these independently developed early lymphatic anlages are less striking than those above described as developing in as- sociation with a receding vein. They are smaller, because they lack the bulk of the contained venous core, and they are more difficult to clearly differentiate against the surrounding mesen- chyme. They are, however, always present and their eventual connection with the larger perivenous lymphatic spaces can be ascertained definitely by following their development through the proper stages. 50 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC "VESSELS In addition, in certain areas, a small group of the earliest lym- phatic anlages appear to develop in the mesenchyme along defi- nite lines, and in distinct patterns, but without any preceding venous reticulum. They impress me, for example, in the area surrounding the omphalomesenteric artery, as systemic lymphatic channels developing ia the placental embryo in regions which are no longer ontogenetically the seat of venous development, al- though occupied by vei.is in other mammalian types. Thus the cardinal collateral line of the marsupials" and the correlated venous area of the monotremes 32 no longer develops as a perma- nent veoous plexus in placentalia 34 but only partially appears in certain forms as a temporary and evanescent component of the abdominal venous complex, as described above for certain stages in the development of the cat (pp. 29 to 33 and figs. 20 to 24). Its place, however, is partially occupied by an early lymphatic plexus de\ eloped in the preaortic mesoderm from the omphalo- mesenteric anlages caudad. Here we are apparently dealing with an instance in which general phylogenetic venous lines have been almost or entirely abandoned in favor of other pathways. Such lines appear, however, to be retained under these con- ditions in the lymphatic organization. Thus, the spaces just referred to, as will be shown subsequently, form the first incep- tion of the extensive network of lymphatic vessels which in the adult cat surrounds the aorta and the origin of the superior mesenteric artery, closely interwoven with the semilunar sym- pathetic and the adrenal plexus, and connecting on the one hand with the portal and intestinal lymphatics, and on the other with the beginning of the thoracic duct. This adult mammalian lymphatic plexus forms a perfect lymphatic shadow-picture of the lacertilian ontogenetic peri-omphalomesenteric venous ring. In conclusion, I wish to give briefly a summary of my reasons for regarding the structures described in this communication as the anlages of the systemic lymphatic vessels. "Except, as recently determined, in Tragulus, in wnich Ungulate the adult postcaval system is of the marsupial type, and in certain embryonic stages of the Cat (15. 5-17 mm.) in which the channels appear as evanescent preaortic vessels, subsequently entirely replaced by lymphatics. (Cf. series 258; figs. 20 to 24.) RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 51 These early lymphatic anlages, whether formed independently in mesoderm, or on the site of phylogenetically abandoned venous lines, or, as is generally the case throughout the body, in close correlation to the embryonic venous pathways, always appear in the same situations and, in the average embryo, at the same developmental period. Their constant character, and regular occurrence and relations, repeated within very narrow limits of individual variation in every embryo of the proper stage, imparts to them a definite morphological character. In every series of the proper age in my collection I find the same spaces in the same place and in identical relationship to the adjacent veins. In some individuals, as above stated, they develop earlier than usual, in others their appearance is retarded, but this applies only to the achievement of the full development typical for the average embryo of a given stage. In the retarded individuals the same structures are always present, only they are less strikingly developed and less numerous when compared with the average normal type charac- teristic of the period under consideration. With sufficient mag- nification it is not difficult to distinguish sharply between the perihaemal lymphatic spaces and the blood-vascular channels proper. With sufficient material every stage of their development can be followed up to the confluence of the entire system and its final entrance into the jugular lymph sacs. These spaces are neither artefacts due to embryonic shrinkage, nor are they the unfilled portions of the blood-vascular capillary network. They are, on the other hand, the well-defined earliest anlages of the systemic lymphatic vessels. The more perfect, as a matter of fact, the embryonic fixation is, the more clearly can these structures be ^identified under the microscope. Their history, as will be shown presently, can be traced with the utmost accuracy, and they can be followed step by step in their de- velopment up to their inclusion in the completed and connected lymphatic channel system. But even in their earliest stages they possess an unmistakable and definite morphological character, quite as distinct as that of the adjacent blood channels. They can be followed closely in 52 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS good serial sections of the proper thickness and fixation, and can be reconstructed in the same manner and with the same accuracy and certainty as the blood channels with which they are for the most part so closely associated, although their lumen connects at no point with that of the vascular channels. These statements are based, not on isolated observations, but on the close and repeated examination of a very large number of embryos of the same form. It seems curious to me that the pres- ence of the first lymphatic anlages, as above described, should be denied, or, as has been recently done, that the isolated appear- ance of these spaces should be ascribed to the " sudden collapsing" of a lymph vessel. At the time at which they make their first appearance there are no " lymph vessels" to " collapse," no more than there are in the homologous haemal ontogenetic stages blood- vessels in the sense of continuous channels. On the contrary, when they reach their period of most striking development (cat, 13-14 mm.) these perivenous lymphatic spaces are, if anything, distended, not only by their fluid contents, but by the remnant of the embryonic atrophying vein which they are in the process of replacing. The only structure showing any sign of "collapse" is the empty endothelial bag of the decadent venule. The spaces become relatively reduced in size in the later stages, after the multiple separate early anlages have fused into a more continu- ous lymphatic channel system. These spaces are always present in embryos of the proper stages in the typical position and in constant relation to the venous channels. By following carefully and with sufficient material their further growth and development in succeeding stages, a clear and consecutive picture of systemic lymphatic genesis is given. It is noteworthy, in view of the incorrect statements published to the contrary, that these primary anlages of the systemic lym- phatic system develop constantly in embryos of the cat before the definite organization of the jugular lymph sacs. These latter structures, in the 10 mm. cat embryo, are still largely in the con- dition of a perivenous capillary plexus, at a time when the first lymphatic anlages can be distinctly recognized in the axial mesoderm. RELATION OF LYMPHATIC TO BLOOD-VASCULAR SYSTEM 53 In view of the fact that these first anlages can be traced step by step as developing into the systemic lympnatic channels of the adult, it becomes apparent that their origin cannot be ascribed to " outgrowths" from the jugular sacs, as yet not formed. A similar lack of correct reasoning appears in portions of the recent anatomical literature in reference to the injection of lymphatics in the early stages. It is manifestly impossible to inject some- thing which does not as yet exist in the form of a continous and injectable system of communicating channels. The spaces which I am describing and which I regard as the first anlages of the systemic lymphatics cannot be injected at the time of their first appearance because they are intercellular spaces, extremely nu- merous, but as yet isolated from each other, and not in communi- cation with veins or with anything else that could be injected, in the sense of filling a lumen artificially with a colored fluid, or air, or any other foreign substance. They multiply rapidly, become confluent to form larger spaces, and these finally unite into continuous channels. But the genetic determination of the origin of the lymphatic system must be based on the close study of embryonic stages long prior to the formation of continuous channels capable of injection. Where lymphatic vessels can once be demonstrated by this method they have passed the formative stage. It then becomes merely a question of topographical ex- tension and readjustment, such as is encountered in all vascular structures alike in the later embryonic periods. The real problem of the genesis of the lymphatics can only be solved by the recognition of tne earliest formative stages, begin- ning with the first appearance of the mesodermal spaces above described. Finally I can sum up my observations, and the views thereon based, as follows: 1 The systemic lymphatic vessels of the entire body arise, in the mammalian embryo, by confluence of mesodermal spaces, developed on the same lines as those governing the formation of the earliest haemal vascular channels, but independent of the same. 54 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS 2 The lymphatic endothelium is an independent modifica- tion of the mesodermai cells lining the first anlages of the lym- phatic spaces, and is not derived from the haemal vascular endothelium. 3 I am obliged to deny the assumption that the mammalian systemic lymphatic vessels arise by the confluence of numerous elements detached in course of development from the embryonic veins. 4 I am obliged to put myself emphatically on record against the assumption that the mammalian systemic lymphatic vessels arise by a progressive sprouting from center to periphery from the endothelium of veins, or from that lining the jugular lymph sacs, or equivalent structures in other regions of the body. The mammalian embrvo offers no evidence of such occurrences. PART I, PLATES FIGURES 1 TO 28 The series here figured and described are in the embryological collection of Columbia University, with the exception of series 34, which belongs to the em- bryological collection of Princeton University. I am greatly indebted to Prof. C. F. W. McClure for the opportunity of studying this series and publishing the eight sections shown in figs. 245 to 251. FIGURE 1 1 Transverse section of 10 mm. cat embryo (series 111, slide VIII, section 4), X 175 Topographical picture of site of earliest lymphatic space development. 1 Sympathetic nerve. 8 Oesophagus. 68 Left postcardinal vein. 69 Lungs. 72 Right dorsal aorta. 73 Left dorsal aorta. 77 Mesenehymal intercellular lymphatic anlages. 78 Coelom. THE SYSTEMIC LYMPHATIC VESSELS PLATE 1 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 2 2 Transverse section of 10 mm. cat embryo (series 111. slide VIII. section 4) X 300 left side of embryo. 68 Left post cardinal vein. 73 Left dorsal aorta. 77 Mesenchymal intercellular lymphatic anlages. 78 Coelom. 79 Coelomic stoma. THE SYSTEMIC LYMPHATIC VESSELS PLATE 2 MEMOIR NO. 1, HUNTINOTON, 1911 FIGURE 3 3 Same. X 600. THE SYSTEMIC LYMPHATIC VESSELS PLATE 3 MEMOIR NO. 1, HCJNTINGTON, 1911 FIGURES 4 AND 5 4 Transverse section of 10 mm. cat embryo (series 111. slide VIII, section 6), X 300. 5 Same, section 7. THE SYSTEMIC LYMPHATIC VESSELS PLATE 4 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 6 AND 7 6 Same, section 8. 7 Same, section 9. THE SYSTEMIC LYMPHATIC VESSELS PLATE o MEMOIR NO. 1, HCNTINGTON, 1911 FIGURES 8 AND 9 8 Transverse section of 10 mm. cat embryo (series 120, slide IX, section 25) X 300. 9 Same, section 26. 1 Sympathetic nerve. 68 Left postcardinal vein. 73 Left dorsal aorta. 77 Mesenchymal intercellular lymphatic anlages. 78 Coelom. THE .SYSTEMIC LYMPHATIC VESSELS PLATE 6 73 '^PH v^*fe^fc 1% . . I i J* V_ . A % g.A^ . 3&$9&R &i;&OT*3S2i5r'^ if* ! P VW ^.' jm "i*f\^K/ ' ^ / t^W^T 's* fii 5 ^ * Z^^ -^ v ? 1?^.^ >-. >^; i -f . i r - * ^ *^ illtt -V * *^ . * *.' % ' MEMOIR NO. 1, HCNT1NGTON, 1911 FIGURES 10 AND 11 10 Transverse section of anterior thoracic region of 12 mm. cat embryo (series 217, slide X, section 12), X 225 showing early extraintimal lymphatic develop- ment. 11 Transverse section of anterior thoracic region in a 12 mm. cat embryo (series 211, slide X, section 15), X 225. 3 Precardinal or precava, resp. azygos of right side. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerating embryonal vein. 6 Precardinal or precava, resp. azygos of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 7 * >*"T^ V v'^' * vVi ^i*^^ 10 11 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 12 12 Transverse section of anterior thoracic region in a 14 mm. embryo (series 127, slide VIII, section 12), X 225. 4 Atrophying embryonal vein, forming kernel in interior of develop- ing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 6 Precardinal or precava, resp. azygos of left side. 6' Left asygos vein, thoracic portion. 7 Aorta. 8 Oesophagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 8 12 MKMOIU NO. 1, HUNTINGTON, 1911 FIGURES 13 AND 14 13 Transverse section of anterior thoracic region in a 15 mm. cat embryo (series 219, slide XIV, section 17), X 225. 14 Same, section 18. 6 Precardinal or precava, resp. azygos of left side. 6' Left azygos vein, thoracic portion. 7 Aorta. 8 Oesophagus. 9 Trachea. 15 Interazygos segment of thoracic duct. 22 Vagus. 36 Azygos segment of thoracic duct. 53 Precaval lymphatics. THE SYSTEMIC LYMPHATIC VESSELS PLATE 9 15 36 13 - .- . v t . C,.* ^- ^-*! ' -'^^^V -- % MEMOIR NO. 1, HUNTINOTON, 1911 FIGURE 15 15 Transverse section through lower cervical region of a 14 mm. cat embryo (series 222, slide VII, section 26), X 150. 11 Jugular lymph sac. 25 Internal jugular vein. 27 External jugular vein. 27' External jugular lymphatics. THE SYSTEMIC LYMPHATIC VESSELS PLATE 10 15 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 16 Hi Same, slide VIII, section 4. 11 Jugular lymph sac. 25 Internal jugular vein. 27 External jugular vein. 27' External jugular lymphatics. THE SYSTEMIC LYMPHATIC VESSELS PLATE 11 11 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 17 17 Transverse section of upper thoracic region of 14 mm. embryo (series 37, slide XIII, section 12), X 150. 1 Sympathetic nerve. 3 Precardinal or precava, resp. azygos of right side. 6 Precardinal or precava, resp. a^ygos of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 22 Vagus. 53 Precaval lymphatics. THE SYSTEMIC LYMPHATIC VESSELS PLATE 12 22 - ji^^^S^^V^^^^isSfKL 'Sl^wiwe^jfiBi-""' : "><-' ' ; " ' $&. V ; sasas g jpE^" "-* " p *a.;&K* - x fr 17 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 18 AXD 19 18 Transverse section of middle thoracic region in a 17 mm. cat embryo (series 258, slide XVIII, section 9), X 225. 19 Transverse section through mid-thoracic region of a 19 mm. cat embryo (series 253, slide XXIV, section 9), X 225. 1 Sympathetic nerve. 2 Intersegmental arteries. 3 Right azygos. 4 Atrophying embryonal vein. 5 Extraintimal or perivenous lymphatic space. 6 Left azygos. 7 Aorta. 15 Interazygos venous plexus. 36 Azygos segment of thoracic duct. THE SYSTEMIC LYMPHATIC VESSELS 2 PLATE 13 -;-. .. ..- &^\ ^V^AXs> i v- * j x , : *~*~$P~ ' ._ ^>. ^.^^^>*/.; MEMOIR NO. 1, HUXTIXGTOX, 1911 FIGURE 20 20 Transverse section of upper abdominal region of a 17 mm. cat embryo, show- ing developing components of ascending lumbar lymphatic trunks and of mesen- teric sac in relation to embryonic veins in the root of the dorsal mesentery and their relation to the cardinal collateral venous channels (series 258, slide XXIII, section 34), X 75. 1 Sympathetic nerve. 7 Aorta. 51 Mesenteric lymphatics. 58 Ureters. 61 Iliac vessels. 62 End gut. 63 Ventral division of cloaca (urinary bladder). 64 Wolffian ducts. 66 Umbilical arteries. 67 Right postcardinal vein. 68 Left postcardinal vein. 74 Cardinal collateral veins. 75 .Ascending lumbar lymphatic trunks. THE SYSTEMIC LYMPHATIC VESSELS PLATE 14 '..jteW^j^i,. r :<*>, -. ' 67 if*; m^Wi^ wjmi^jl? i^lptspift '%i^$^ .SA 75 51 68 - -'%V- % T j2>V* *"*' -" i - r - I>; .--' T r^--- T ' *** "< - 11 - 11 y ..'' ^ > *' ' /wtinr'ivvy.-. raHf4f mX. t 9n ' 'ft '<' u'* J ttfiBfcJi *. t ,/rirfla mil ;ifci ,_ .%_ ff ,/^_ , .* y,. * x ' MkK u r !' . !*~aSHS 75 74 61 T\ : ^' ; ^^\^fe^ML^^ Jafe vJ- 'i 1 '.'-*.'!- 1 ?' 11 : ^AM^T-r^V^ 20 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 21 21 Same section, X 150. 7 Aorta. 51 Mesenteric lymphatics. 58 Ureters. 67 Right postcardinal vein. 68 Left postcardinal vein. 74 Cardinal collateral veins. 75 Ascending lumbar lymphatic trunks. THE SYSTEMIC LYMPHATIC VESSELS PLATE 15 21 Ml..\ (Hi NO. ]. HfNTTNCTf.N, 1011 FIGURE 22 22 Same, section 33. 1 Sympathetic nerve. 7 Aorta. 51 Mesenteric lymphatics. 58 Ureters. 67 Right postcardinal vein. 68 Left postcardinal vein. 74 Cardinal collateral veins. 75 Ascending lumbar lymphatic trunks. THE SYSTEMIC LYMPHATIC VESSELS PLATE 16 22 MBMOITt NO. 1, HtTXTINGTON, 1011 FICURKS 23 AND 24 23 Same, section 32. 24 Same, section 31. 7 Aorta. 51 Mesenteric lymphatics. 67 Right postcardinal vein. 68 Left postcardinal vein. 74 Cardinal collateral veins. 75 Ascending lumbar lymphatic trunks. I 111'; SYSTKMIC LYMPHATIC VKSSKLS PLATE 17 67 23 MEMOIR NO. 1, HUNTIXGTON, 1911 FIGURE 25 25 Transverse section of posterior end of a 20 mm. cat embryo (series 241, slide XXX, section 4), X 75 showing cxtraintimal replacement of left supra car- dinal vein (59) by ascending lumbar retroaortic lymphatic channel (76). 1 Sympathetic nerve. 7 Aorta. 58 Ureters. 59 Left supracardinal vein. 60 Right supracardinal vein. 62 End gut. 63 Ventral division of cloaca (urinary bladder). 64 Wolffian ducts. 65 Metanephros. 66 Umbilical arteries. 67 Right postcardinal vein. 68 Left postcardinal vein. 76 Retroaortic supracardinal lymphatic trunks. THE SYSTEMIC LYMPHATIC VESSELS PLATE 18 M-GMOIR NO. 1, HUNTINGTON, 1911 25 FIGURE 26 26 Same, section 14, X 150. 1 Sympathetic nerve. 58 Ureters . . 59 Left supracardinal vein. 60 Right supracardinal vein. 67 Right postcardinal vein. 68 Left postcardinal vein. 75 Ascending lumbar lymphatic trunks. 76 Retroaortic supracardinal lymphatic trunks. THE SYSTEMIC LYMPHATIC VESSELS PLATE 19 75 26 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 27 27 Reconstruction of anterior vascular complex in an 18 mm. cat embryo (series 88), X 50. Ventral view. THE SYSTEMIC LYMPHATIC VESSELS PLATE 20 LEFT INT. JUG. V. EXT. JUG. V JUGULAR APPROACH THYRO CERVICAL A. LEFT COMMON JUGULAR VEIN EFT SUBCLAVIAN LEFT SUBCLAVIAN V. INT. MAMMARY A. PRECAVAL E LYMPHATIC INNOMINATE A. RIGHT PRECAVA LEFT PRECAVA 27 MBMOIIt NO. 1, HUNTING-TON, 1911 FIGURE 28 28 Same, lateral view of left side. THE SYSTEMIC LYMPHATIC VESSELS PLATE 21 LEFT INT. JUG. V LEFT EXT. JUG. V. JUGULAR APPROACH THYRO-CERVICAL A. LEFT COMMON JUGULAR VEIN LEFT SUBCLAVIAN V. LEFT SUBCLAVIAN A INT. MAMMARY A ANDV. RIGHT PRECAVA INNOMINATE A LEFT CEPHALIC VEIN LEFT SUBCLAVIAN APPROACH JUGULO-SUBCLAVIAN TAP THYRO-CERVICAL A. SUBCLAVIAN A THORACIC DUCT LEFT PRECAVA 28 MEMOIR NO. 1, HUNTINOTON, 1911 PART II THE DEVELOPMENT OF THE PREAZYGOS AND AZY- GOS SEGMENTS OF THE THORACIC DUCTS In certain developmental stages in the cat, the embryonic anlages of the thoracic ducts appear as potentially symmetrical bilateral lymphatic channels. Subsequently, the cephalic portion of the left duct, having secondarily assumed the function of conveying the lymph from the trunk and the posterior extremities, forms, cephalad to the aortic arch level, the main continuation of the thoracic duct proper, emptying, through the intervention of the left jugular lymph sac, into the general venous system at the left common jugular confluence, or at the jugulo-subclavian junction, or at both of these points, according to the type of adult lymphatico- venous connection established in individual cases (26) > (2?; > (34) > (35;> The corresponding cephalic preazygos portion of the right lym- phatic duct then remains as a more or less rudimentary structure, draining the anterior mediastinal region and emptying into the right jugular lymph sac. Caudad of the level of the aortic arch, the adult thoracic duct is a composite of parts of three embryonic lymph channels pre- senting a great range of individual variation. Two of these anlages develop as bilateral extraintimal lymphatic trunks along the ventral aspect of the right and left azygos veins. The third intermediate channel, which is the first to distinctly differentiate in the embryo and is the main element in the typical organiza- tion of the adult duct, develops in the caudal portion from ex- traintimal lymphatic anlages, which replace part of the ventro- medial tributary plexus of the right azygos vein. In ascending 35 C. F. W. McClure and C. F. Silvester: "A comparative study of the Lym- phatic-Venous communications in Adult Mammals," Anat. Rec., Vol. Ill, pp. 634-551. 55 56 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS it increasingly turns to the left and continues to develop cephalad along the ventral surface of the inter-azygos plexus, by extra- intimal replacement of elements belonging to the latter. (Com- pare the series of six reconstructions shown in figs. 188, 189, 190, 191, 192, 193.) The details of the adult organization of the thoracic ducts, and the genesis of the observed adult variations on the basis of the development of the embryonic channels, will be considered subse- quently (Part V). The present communication is confined to the consideration of the development of the preazygos and azygos segments of the ducts. The thoracic ducts develop in cat embryos between 11 and 16 mni. crown rump measure. In the 16 mm. embryo, McClure (15J states that they "for the first time make their appearance as definite, continuous lymph chan- nels, which are disconnected from the venous system, except through the mediation of the jugular lymph sacs." This is the case in the majority of 16 mm. embryos, but in a number of 15, 15.5 and 16 mm. series in my collection the final union of the individual segments has not yet taken place. It is possible, by combining a large number of observations on numerous embryos of the same length, to establish an average developmental condi- tion of the thoracic ducts for each stage in embryos between 12 mm. and 17 mm. crown-rump measure. Individual embryos will, however, show marked differences in development, and many departures from the standard plan are encountered. The same conditions obtain here which McClure and I found in our study of the development of the jugular lymph sacs. Some embryos show, for their length, unusually advanced thoracic duct development in certain areas, while in other parts this development is relatively retarded and more primitive. In other series the entire region of thoracic duct development appears more uniformly ahead of the average condition, normally en- countered in embryos of this period. Finally, in a third group, the appearance of the typical lymphatic development seems to be delayed until a later stage than is normally the case. But in spite of these individual differences, a uniform and consistent ground plan of thoracic duct development can be demonstrated, to DEVELOPMENT OF THE THORACIC DUCT 57 which the majority of embryos conform and into which even the aberrant types fit. As a matter of fact, the regularity with which the thoracic duct anlages first appear in the majority of the individuals of the proper stage, and the uniform and constant character of the resulting histological pictures, is one of the most striking and significant features of the entire process. Prior to the 11 mm. stage no definite anlage of the anterior segments of the future ducts is observable. . I believe that the adult thoracic ducts of the cat are developed by fusion of three distinct and separate regional segments. Each of these segments is in turn formed by confluence of a number of originally discrete and independent anlages, which develop inde- pendently of the venous system as extra-intimal or perivenous mesenchymal spaces in the sense previously defined. (1 2 > 21 > 22) - These spaces are applied to, or surround, the walls of the embryonic veins of the lower cervical and of the dorsal mediastinal region. The three main divisions, thus developed independently of the venous system, unite with each other to form the channels of the left and right thoracic ducts, and these usually gain their point of entrance into the systemic veins by uniting with a process of the jugular lymph sacs (" thoracic duct approach") derived from their dorsal aspect, just cephalad to the common jugular approach. Exceptionally the main (left) thoracic duct, by combining with the ventral mediastinal trunk presently to be described, unites (fig. 98) with the ventral process of the "subclavian approach," prolonged from the jugular lymph sac over the ventral surface of the jugulo-subclavian confluence, and thus establishes its connection with the venous system (cf. pp. 84 89, figs. 89 to 99). The ontogenetic history of the ducts may therefore (fig. 29) be considered under four headings, viz.: 1. THE THORACIC DUCT APPROACH of the jugular lymph sac, forming the terminal of the adult duct on each side (fig. 29, 1, blue). 2. THE PREAZYGOS SEGMENT (fig. 29, II). This includes two distinct and separate channels: A. The broncho-mediastinal or ventral mediastinal lymphatic trunk (fig. 29, 37. green), which drains the ventral mediastinum 58 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS cephalad of the pericardium, viz., the pericardial, tracheal, bronchial, lateral oesophageal, and thymic areas. This lymphatic channel, associated with the embryonic venous plexuses of the pretracheal region, the pulmonary vessels and the vagi, develops through confluence of a large number of sep- arate and independent extra-intimal lymphatic spaces following and surrounding the embryonic venous plexuses of the ventral mediastinum. The chain formed by these spaces eventually unites with the similar chain forming the anlages of the preazygos segment of the thoracic duct. B. The preazygos segment of the main thoracic duct (fig. 29, 35, brown). This portion of the main duct extends from the point of its entrance into the jugular lymph sac, through the thoracic duct approach of the latter, caudad to its intersection with the dorsal surface of the aortic arch. In the adult animal this segment forms the relatively long por- tion which ascends cephalo-sinistrad from the point where the duct parts company with the right azygos vein, to its junction with the jugular lymph sac, under cover of the aortic arch, and the vertical portion of the left subclavian artery, dorsal to the vertebral vein and to the left innominate confluence. In this part of its course the thoracic duct frequently receives the lym- phatic return from the ventral mediastinum through channels which unite it to the ventral mediastinal trunk as just defined. The preazygos segment of the main duct is again formed in the embryo by confluence of independent mesenchymal spaces around and along the prevertebral and dorsal mediastinal venous plexuses of the embryo. 3. THE AZYGOS SEGMENT (fig. 29, III, 36, yellow) comprises the portions of the thoracic ducts caudal to the level of the aortic arch. It develops, again independently, as the result of fusion of a number of extraintimal mesenchymal spaces closely applied to the ventral surface of the azygos veins, or of their ventro- medial tributaries, or surrounding the latter. 4. THE POST- AZYGOS SEGMENT (fig. 29, IV, 52, green), through which the thoracic ducts establish their connection with the Receptaculum and the system of the abdominal lymphatics. DEVELOPMENT OF THE THORACIC DUCT 59 Inasmuch as the development of this segment of both ducts is intimately connected with that of the principal abdominal lymphatic channels, its consideration will be included in that of the latter system (Part III). The present communication deals only with the first three segments above enumerated. Fig. 29 shows in a color schema the genetic composition of the thoracic ducts according to the segments just outlined, and their relation to the embryonic venous system. . In proceeding caudo-cephalad the post-azygos portion (IV) develops in the larger right segment (52) by replacement of the caudal part of the ventro-medial tributary plexus of the right azygos vein. The smaller channel of the left side replaces the caudal part of the left azygos vein. In the azygos area (III) the left channel follows, as a much reduced left thoracic duct, the prevertebral portion of the left azygos vein. The larger right lymphatic channel (36), gradually inclining cephalo-sinistrad, develops along the ventral aspect of the inter-azygos cross-anastomosis and unites with the left duct eventually in a close plexiform inter- lacement perforated by the intersegmental arteries, which con- tinues cephalad to near the level of the aortic arch. Here the plexus condenses into two more or less well defined channels which turn up to the right and left into the preazygos segment (II). The smaller right channel follows, cephalad of the right azygos terminal, the dorso-medial aspect of the right precaval, innominate and common jugular veins, developing by confluence of extraintimal spaces along the course of these vessels (cf . recon- struction of series 88, figs. 27 and 28), and forms the cephalic end of the reduced right thoracic duct of the adult. It terminates by joining the thoracic duct approach (12) of the right jugular lymph sac. The left and larger trunk replaces the extra-peri- cardial portion of the left precava, and ascends dorsal to the left innominate confluence to its junction with the thoracic duct approach (12) of the left jugular lymph sac. 37 in fig. 29 indi- cates the broncho-mediastinal trunk in one of its commonly encountered types, joining caudally the preazygos portion of the thoracic duct (35), and terminating cephalad in the ventral 60 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS process of the subclavian approach (14) of the jugular lymph sac, in company with the main lymphatic channels returning from the anterior extremity. 1. THE THORACIC DUCT APPROACH OF THE JUGULAR LYMPH SACS (FIG. 29, 18) This structure is an integral part of the jugular lymph sacs and forms in the adult animal the portal of entry of the thoracic ducts into the lymph sacs, and through them into the systemic veins. McClure, in the paper quoted Cl5) > has described its general formation and relation to the jugular lymph sac and has given an excellent illustration of its appearance in the dorsal view of a reconstruction of a 14 mm. cat embryo. I have here defined it as the "thoracic duct approach" of the sacs, in con- formity with the nomenclature which McClure and I adopted in describing (26> 27 ) the development of those portions of the jugular sac which, forming the " jugular" and " subclavian" "approaches," project as blind processes from the caudal end of the sac, and are destined to effect subsequently the permanent lymphatico-venous connections at either the common jugular, or the jugulo-subclavian confluence, or at both of these venous an- gles. The term, as here used in describing the similar process of the sac which is to establish definite connections with the thor- acic duct, is intended to emphasize the fact that the ' 'thoracic approach" of the lymph sac is, like the rest of the structure from which it proceeds, venous in origin, while the thoracic ducts proper, with which it secondarily unites, are not derived from the veins, but are, from their beginning, independent of the blood-vascular channels, and develop by confluence of independent extra- or peri- venous mesenchymal spaces. This fundamental difference in origin is shared, in my opinion, by all the other main systemic lymphatics which, in addition to the thoracic ducts, enter the jugular sacs as the subclavian, ex- ternal jugular, cephalic and internal jugular lymphatics through the processes which the sac sends to meet them. The thoracic duct approach of the earlier embryos arises as a short, blunt, curved, conical process from the dorso-medial aspect DEVELOPMENT OF THE THORACIC DUCT 61 of the sac, a short distance cephalad of the jugular approach and of the common jugular venous confluence. It extends meso-caudad across the dorso-lateral and dorsal aspects of the internal and common jugular veins between them and the thyro-cervical artery, towards the vagus-sympathetic strand, dipping ventro-mesad between the main vein and the nerves and ending blindly in this situation. The process makes its first distinct appearance, after the jugular sac has been fully established, in embryos of between 11 mm. and 14 mm. crown-rump measure. It is noteworthy that individual embryos within these limits show considerable difference in the degree of development of the process. In some it appears earlier, in others a little later, so that in individual instances an embryo of slightly shorter measure may have the thoracic duct approach of the jugular lymph sac further developed than one which exceeds the first in total length by the fraction of a millimeter. Thus one of the earliest distinct stages of the process is seen in a 14 mm. embryo (series 210, slide viii, sections 39-42, figs. 30 to 33 inclusive). In this embryo, the thoracic duct approach appears as a short blunt process which projects dorso-mesad from the caudal end of the left jugular lymph sac, a short distance cephalad of the common jugular confluence. Sections 39, 40 and 41 (figs. 30, 31 and 32, 12) show the earliest stage in the development of this process. In them the area dorsal to the main vein, between the sympathetic nerve and the jugular lymph sac, is occupied, in the direction meso-laterad, by the sympathetic nerve (1), the termination of the large dorso-medial venous tributary (left sup. intercostal vein) (16), the thyro-cervical artery (24),. and, laterad to this, by the blind terminal portion of the thoracic duct approach (12), derived from the dorso-medial aspect of the jugular lymph sac (11). In this embryo the process of the sac ends blindly laterad to the thyro-cervical artery (figs. 33, 34, 24). It has not yet extended meso-caudad across the dorsal surface of the main vein toward the interval between the thyro-cervical artery and the sympathetic nerve. The thyro-cervical artery 62 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS (24) lies at a little distance directly dorsad of the internal jugular vein. The dorso-medial angle of the main vein is related to a large compound medial venous tributary, draining the dorso- medial (pre vertebral), (17) and the dorsal somatic areas (16). This compound vessel is the anlage of the future superior inter- costal vein. In fig. 33 (section 42) the common trunk opens into the main vein just dorsal to the sympathetic nerve (/). In this section, the blunt medial extremity of the thoracic duct approach (12) is separated from the portion still connected with the jugular lymph sac, thus expressing the caudo-mesal curve of the entire process. It ends blindly in the following section, and does not reappear in the succeeding sections. The common jugu- lar confluence occurs in section 44 of this slide. We are dealing, therefore, with a short, blunt conical process of the left jugular lymph sac, which curves meso-caudad for two sections, and terminates blindly dorsal to the main vein. In another 14 mm. embryo (series 212) the thoracic duct ap- proach is much more extensively developed and occupies sections 11 to 26 of slide ix. Selected sections of this series are shown in figs. 34 to 42 inclusive. Fig. 34 (series 212, slide ix, sec. 14) shows part of the caudal end of the left jugular lymph sac (11) with the thoracic duct approach (12) arising from its medial aspect and extending mesad across the dorsal surface of the internal jugular vein (25). On the mesal aspect of the vein are seen the sympathetic (1) and the vagus (22). The large common trunk formed by the union of the dorsal somatic and pre vertebral venous tributaries (16, 17) lies dorso- medial to the main jugular channel. It joins the latter, as left- superior intercostal vein, (16, 4-6,) in section 26 of this slide (fig. 41). On the lateral aspect of this vessel is the ascending dorsal branch of the thyro-cervical artery (24). The six succeeding figures (35 to 40 inclusive) are cut down so as to only show the dorsal circumference of the internal jugular vein and the superincumbent thoracic duct approach of the jugular lymph sac. In fig. 35 (section 15) the approach has extended further mesad DEVELOPMENT OF THE THORACIC DUCT 63 across the main vein. In fig. 36 (section 16) a blind terminal protrusion (12} separates from the main lumen of the sac (11}, indicating the frequent bifid character of the thoracic duct ap- proach. The succeeding sections (18 and 20, figs. 37 and 38) hence show an apparent recession of the entire process. In sec- tion 22 (fig. 39) the approach has again extended mesad toward the thyro-cervical artery and the superior intercostal vein. In the succeeding section (fig. 40) the fundus of the process terminates for the second time in a blind pouch-like protrusion (12}. In the following section (24, fig. 41) the approach (12} again appears fully developed and in open connection with the main jugular lymph sac (11}. It maintains the character in section 25, and in section 26 (fig. 42) the terminal of this second protrusion (12} ends blindly. It does not reappear in the succeeding sections. We are dealing here with an instance of very common occur- rence, in which the thoracic duct approach of these earlier stages not only gives off a number of blind terminal diverticula, but in which the entire approach arises by two or three distinct and separate origins from the main jugular lymph sac. This character will again be discussed below (cf. p. 66). in connection with the interpretation of certain adult conditions, but the instance now under discussion shows the early type of this variation very clearly, so that in a graphic reconstruction the thoracic duct approach of this embryo would appear as in figure on page 64. This embryo therefore presents for its age an unusually long and well developed thoracic duct approach, with triple terminal pouches, extending through sixteen sections, while at the same time it remains throughout this extensive course in the early position, viz.: lateral to the thyro-cervical artery (24) and the common trunk of the dorsal and dorso-medial tributaries of the internal jugular vein (16, 17, 46)- The approach has not as yet arched mesad into the neuro-venous recess, and is hence in marked contrast to the succeeding stages presently to be described. In a 13.5 mm. embryo, series 189, the development of the tho- racic duct approach has gone much further, although the embryo, a litter mate of series 210, measures half a millimeter less. The thoracic duct approach of the left jugular lymph sac 64 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS occupies sections 10 to 23 of slide vii in this embryo, and is shown in the dorsal view of the reconstruction, fig. 89, to which the individual sections following are referred for orientation. The process arises in the usual situation from the dorso-medial aspect of the left jugular lymph sac and trends dorso-mesad f 12 Sec. 16 Fit 36 / >n / Sec. 18-20 22 Fig. 37 38-39 \^S / k 7 Sec 23 FiJ. 40 I SJ / across the internal jugular vein (series 189, slide vii, section 10, fig. 43). In reference to the level of the common jugular conflu- ence, which occurs in slide vii, section 26, the thoracic duct approach in series 189 arises more anteriorly from the jugular lymph sac than in the preceding series 210. Its extent and course DEVELOPMENT OF THE THORACIC DUCT 65 can be followed accurately in the successive sections shown in figs. 43 to 55 inclusive (series 189, slide vii, sections 10 to 22 inclusive). In section 11 (fig. 44) the thoracic duct approach is separated by a thin endothelial partition from the lumen of the main jugular lymph sac, which it rejoins in the following section (section 12, fig. 45), only to again separate in the next section (section 13, fig. 46). In the meanwhile, the thoracic duct approach has arched from the lateral to the medial side of the internal jugular vein (25), closely applied to its dorsal surface, between the same and the thyro-cervical artery (24), and its blind terminal is beginning to bend ventro-mesad into the neuro-venous recess between the vagus-sympathetic strand (1, 22) and the internal jugular vein (25). In the next three sections (sections 14, 15, 16, figs. 47, 48 and 49) the thoracic duct approach becomes slightly dilated and diverticular (sections 15 and 16, figs. 48 and 49), but remains detached from the jugular sac. In section 15 (fig. 48) it appears subdivided, by endothelial partition, into two, and in section 16 (fig. 49) into three components. The protru- sion which the common jugular confluence sends towards the jugular approach of the lymph sac and which forms the point of the subsequent invagination of the venous angle by the lym- phatic tap, begins to appear in section 16 (fig. 49), and can be traced through the succeeding sections as a derivative of the lateral aspect of the main internal jugular vein, between it and the median surface of the jugular lymphatic sac. In the six following sections, 17 to 22 inclusive, (figs. 50 to 55 inclusive) the proximal end of the process, connected with the sac, gradually recedes, while the blind distal end (12') continues to develop in the ventro-medial direction between the main vein (25, in fig. 55) and the dorsal and dorso-medial venous tributaries (16 and 17, fig. 55), dipping ventrad and approaching the interval between the internal jugular vein (25) and the sympathetic nerve (1), the site of its subsequent junction with the independ- ently developed perivenous lymphatics of the ventral medias- tinal and dorsal prevertebral regions, which together constitute the anlages of this cephalic portion of the future thoracic duct, (cf. series 143, slides x and xi, figs. 172 to 183). The thoracic duct approach ends blindly in section 26 of slide vii. The 66 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS course and the relations to surrounding structures of the thoracic duct approach of this embryo are shown in the dorsal view of the reconstruction (fig. 89), containing the 50 sections of slide vii in a magnification of 225 diameters. This view of the reconstruction presents from the lateral to the mesal border the following structures. 1. Caudal end of left jugular lymph sac, with the jugular (13) and subclavian (14) approaches, between which passes the ventral branch of the thyro-cervical artery (24). The fifth spinal nerve (54) penetrates the lymph sac, while the sixth nerve lies on its lateral aspect with the primitive ulnar veno-lymphatic arching over it to enter the lymph sac. 2. The dorsal aspect of the internal jugular vein (25). 3. The main trunk and branches of the thryo-cervical artery (24) . 4. The left superior intercostal vein (46), with the dorsal (16) and dorso-medial or prevertebral (17) tributaries, enters thedorso- medial angle of the internal jugular vein between the thyro- cervical artery (24) and 5. the double strand of the vagus (22) and sympathetic (l) nerves. 6. The thymus (21). 7.' The aortic arch (7), with the left carotid artery (23) and the origin of the innominate artery (45). The thoracic duct approach (12) arises by two branches from the dorso-medial aspect of the sac, turns caudo-mesad between the internal jugular vein and the ascending branch of the thyro- cervical artery, in front of the left superior intercostal vein, and approaches the recess formed between the main vein and the bundle of the nerves. The reduplication of the origin of the thoracic duct approach from the jugular lymph sac present in this embryo is not unusual, and leads to the type of divided termination of the thoracic duct in the adult shown in fig. 56. The left jugular lymph sac (3) connects with the venous system at the common jugular angle (2) by a larger ventral and smaller dorsal jugular approach (18). It also taps at the jugulo-subclavian angle (19). The thoracic duct enters the dorsal aspect of the sac by a divided terminal, the lower branch connecting with the subclavian approach. DEVELOPMENT OF THE THORACIC DUCT 67 Finally an intermediate condition is seen in another 14 mm. embryo, series 122, of which selected sections are appended, from slides ix and x (figs. 57 to 62). Section 21 of slide ix (fig. 57) is taken at the level where the jugular lymph sac has divided for the passage of the thyro-cervical artery (24) into its two main terminal components, viz., the dorsally placed subclavian ap- proach (14), and the ventrally situated jugular approach (13). The latter appears in the interval between the internal jugular vein (25) and the jugulo-cephalic trunk (29). These veins are approaching each other to form the common jugular confluence (fig. 39, '26), at which point the jugular approach usually makes the secondary permanent lymphatico-venous tap. The thoracic duct approach (12) arises from the dorsal aspect of the jugular approach by a relatively narrow pedicle and then hooks, as an. elongated and expanded curved pouch, ventro-mesad over the dorsal aspect of the internal jugular vein towards the neuro- venous recess. In section 28 of the same slide (fig. 58) the internal jugular and jugulo-cephalic veins have met in the common jugular con- fluence (26) which includes the jugular approach. Consequently the latter structure only appears in its ventral prolongation which is seen occupying the ventral angle of the common jugular junc- tion (not labelled in fig. 58). The subclavian approach (14) has moved relatively nearer to the dorsal surface of the main vein channel. The blind terminal end of the thoracic duct approach (12) is seen dorsal to the common jugular vein, between a large dorso- medial tributary of the latter (16) mesally, and the thyro-cervical artery (24) laterally. Its pointed extremity is directed further ventro-mesad toward the interval between the sympathetic nerve (1) and the medial surface of the common jugular vein (26). In section 7 of slide x (fig. 59) the terminal of the thoracic duct approach (12) has relatively receded and lies in the narrow interval between thyro-cervical artefy (24) and common jugular vein (26). In fig. 60 (section 9 of slide x) a large dorso-medial venous tributary (16, left superior intercostal vein) empties into the 68 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS dorso-medial angle of the common jugular trunk. The thoracic duct approach (12), occupying the same relative position as in the preceding figure, lies between this tributary and the thyro-cer- vical artery, dorsal to the main vein. Section 12 of slide x (fig. 61) shows the distal reduced end of the approach in this same situation. The subclavian approach of the jugular lymph sac (14) is reduced to a narrow longitudinal cleft on the dorso-lateral aspect of the main vein. This reduction is coincident with the derivation from the dorsal aspect of the subclavian approach of the primitive ulnar veno-lymphatic (31). This channel arches dorsad over the Vlth spinal nerve (30), and appears in this sec- tion in its typical relation to this nerve. Finally in the succeed- ing section, 13 of slide ix, (fig. 62), the thoracic duct approach ends blindly. The structures shown in the section dorsal to the common jugular trunk are now encountered in the following order: mesally the terminal of one of the branches of the dorso- medial venous plexus (16)} next the thyro-cervical artery (24)} then the Vlth segmental nerve (30), and, dorsal of this, the primitive ulnar veno-lymphatic (31). The series of sections of this embryo shown here are interest- ing not only in respect to the behavior of the thoracic duct approach, but also in reference to the mutual relations of this structure and the terminals of the dorso-medial tributary venous plexus. These relations are to be discussed subsequently in con- sidering, as whole, the evidence upon which some investigators have based their conception of direct venous origin of the lym- phatic system by fusion of a number of separate elements de- tached from the main venous trunks (cf. infra p. 153.) The variable appearance of the thoracic duct approach in the younger stages is perhaps best illustrated by the exaggerated development presented in a 13 mm. embryo (series 92), in which the process occupies the last thirteen sections of slide x, and the first fourteen sections of slide ix. Selected microphotographs of this series are shown in figs. 63 to 78. The embryo presents the largest and most dilated thoracic duct approach of any in my collection. Fig. 63 gives a topographical view of the region involved. The succeeding sections are cut down and confined DEVELOPMENT OF THE THORACIC DUCT 69 to the immediate neighborhood of the thoracic duct approach. Fig. 63 shows the sections of the internal jugular vein (25) and jugulo-cephalic trunk (29), approaching their point of union in the common jugular confluence. A large process of the jugular lymph sac (13) extends, as the jugular approach, into the interval between the main veins. Along the mesal aspect of the internal jugular vein (25). are placed the sympathetic (1), the vagus (22), the left common carotid artery (28), and, further ventrad, the thymus (21 ) . The dorsal portion of the sac is about to form the subclavian approach (11, 14). In fig. 64, the jugular approach (13) has entered more deeply into the narrowing interval between the internal jugular (25) and jugulo-cephalic trunks (29), and the sharply marked con- striction between it and the remainder of the sac (subclavian approach, 14) foreshadows its coming separation from the latter. A protrusion of the dorso-medial angle of the jugular approach, adjacent to the internal jugular vein, indicates the beginning derivation of the thoracic duct approach (12). In the succeeding section (fig. 65) the jugular approach (13) has separated from the subclavian approach (14) of the jugular lymph sac and thus cleared the path for the passage of the thyro- cervical artery (24) between the two cardinal divisions of the lymph sac, whose caudal terminations are represented by the jugular and subclavian approaches. The origin of the thoracic duct approach (12) from the dorso-medial angle of the jugular approach is now defined. In the succeeding section (fig. 66) the triple termination of the jugular lymph sac, in the subclavian (14), jugular (13) and tho- racic duct approaches (12), is accomplished, and the latter is begin- ning to extend mesad across the dorsal surface of the internal jugular vein. The main thyro-cervical artery is approaching the junction with its secondary dorsal branches. In fig. 67 the jugular approach (13) has divided into the inter- mediate part between the internal (25) and jugulo-cephalic (29) veins, which is to establish the tap at the common jugular angle, and a ventral process which extends from the tap caudad over the ventral surface of the common jugular angle and vein. 70 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS The thoracic duct approach has enlarged, is partially subdivided (cf. p. 65, figs. 48 and 49, series 189) and its blind terminal extremity is reaching ventro-mesad into the angle between internal jugular vein and sympathetic (7). The thyro-cervical artery (24) lies between the subclavian and thoracic duct approaches, dorsal to the jugulo-cephalic trunk (29). In fig. 68 the thoracic duct approach has further enlarged and is still subdivided into two compartments the terminal of which abuts directly against the sympathetic nerve. The jugular approach (13) communicates in this section with the internal jugular vein, while a narrower slit-like division continues further caudad toward the common jugular angle. In fig. 69 the internal (25) and jugulo-cephalic (29) trunks are only separated from each other by a narrow septum containing the lumen of the slit-like prolongation of the jugular approach. Both in this and in the preceding section the ventral process of the jugular approach lies in the angle between the two veins, closely applied to their ventral surface. The thoracic duct approach is a large sausage-shaped structure, curved over the dorsal surface of the internal jugular vein. The septum in its interior has disappeared and the process presents a clear lumen. In fig. 70 (slide ix, section 2), the confluence of internal jugular and jugulo-cephalic trunks has taken place (26). The thoracic duct approach (12) has extended both ventro-mesad into an area of indifferently defined mesodermal spaces, mesad to the sym- pathetic nerve (1), and laterad toward the subclavian approach of the jugular sac (14), passing between common jugular vein (26) and thyro-cervical artery (24)> In section 5 of slide ix (fig. 71) the thoracic duct approach (12) appears as a long curved multilocular channel, which laterad reaches the subclavian approach of the jugular lymph sac, between the thryo-cervical artery (24) and common jugular vein (26), while its blind mesal extremity terminates just dorsal to the sym- pathetic-vagus strand. In the following section (fig. 72) the thoracic duct approach (12) connects with the subclavian approach (14) of the jugular lymph sac affording an instance of the gene- DEVELOPMENT OF THE THORACIC DUCT 71 sis of the reduplicated or multiple thoracic duct terminations oc- casionally encountered in the adult as above described (cf.p. 66, fig. 56, series 212, slide ix, figs. 34 to 42). In fig. 73 (series 92, slide ix, section 9) the subclavian approach (14) is much reduced, and lies along the lateral aspect of the com- mon jugular vein '(26). Its connection with the thoracic duct approach (12) is plainly visible, passing between the main vein (26) and the thyro-cervical artery (24). Dorsal somatic and prevertebral venous tributaries (16), approaching the dorso- medial angle of the common jugular vein (26), intersect the course of the thoracic duct approach, and begin to divide the same into a medial and a lateral element. This division is completed in the following section (fig. 74) in which the united dorsal and pre- vertebral tributaries form a venous trunk of considerable size (16) which is approaching its confluence with the common jugular vein by passing between the lateral and medial components of the thoracic duct approach. In other words, the appearance of the thoracic duct approach in this and in the following sections is the expression of the ventro-mesal arched course of this structure across the fork or angle between the common jugular vein (26) and the terminal of its ventro-medial tributary plexus (16). The recon- structions shown in figs. 89, 90 and 91 indicate this topographical relationship clearly. The blind mesal end of the arched process (12) now lies mesal to the tributary vein (16) and dorsal to the sympathetic nerve (1). In the following section (section 11 of slide ix, fig. 75), the dorso-medial venous tributary (16) is elong- ated and close to the main vessel, while the reduced sections of the thoracic duct approach lie on either sid of the same, as indi- cated by the forked leader (12). The subclavian approach (14), still prominent and connected with the medial section of the thoracic duct approach in the preceding section 10 (fig. 74), appears now in section 1 1 reduced to a detached small blind end (not labelled in the figure), close to the lateral wall of the main vein and ventral to the thyro-cervical arterj^ (24)- In the suc- ceeding section 12 (fig. 76) the dorso-medial venous tributary (superior intercostal vein) enters the common jugular trunk. The lateral portion of the thoracic duct approach (not labelled 72 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS in the figure) is seen in the deep recess between this entrance, the main vein (26} and the thyro-cervical artery (24} . The mesal end of the thoracic duct approach terminates blindly in this section just dorsal to the sympathetic nerve (!}. In section 13 (fig. 77) the lateral terminal of the thoracic duct approach (12} can still be clearly seen in the interval between the wide dorsal opening of the combined somatic and prevertebral (17} tributary stem into the main vein (26} and the thyro-cervical artery (24}> It appears very much reduced in the same situation in section 14 (fig. 78), and ends blindly in the following section. The last three sections of this series (figs. 76, 77 and 78) show the begin- ning of the jugulo-subclavian confluence. In the later stages the thoracic duct approach has extended further ventro-mesad into the recess between the common jugular vein and the sympathetic nerve, and in this situation it eventually makes its connection with the two independently developed lym- phatic channels of the anterior mediastinal region, viz., the preazygos segment of the thoracic duct dorsally, and the broncho- mediastinal trunk ventrally. The mode of this union is dis- cussed below under a separate heading (cf. infra, p. 111). Figs. 79 to 82 show transverse sections of the left side of the lower cervical region in a 15 mm. cat embryo (series 245, slide xi, sections 24 to 27, X 225), and fig. 90 gives the dorsal view of the reconstruction on the same scale of the thoracic duct approach and the related structures in the same embryo (slide xi, sections 1-44). In fig. 79 (slide xi, section 24) the terminal extremity of the thoracic duct approach (12} is seen between the common jugular vein (26} and the sympathetic nerve (!}, having reached this situation by extending ventro-mesad across the dorsal tribu- tary 16 (left superior intercostal vein), between this vessel and the main venous trunk. The section is taken below the common jugular confluence, consequently the jugular approach is no longer seen, and the blind end of the thoracic duct approach appears isolated. The only other portion of the jugular lymph sac carried caudad to this level is the subclavian approach (14), seen on the lateral aspect of the main vein. The sequence of structures now successively DEVELOPMENT OF THE THOKACIC DUCT 73 encountered along the mesal and dorsal aspects of the main com- mon jugular vein (26} are the following, proceeding meso-laterad : 1. Thymus (21), with sub- and suprathymic venous tributaries entering the main vein. 2. Left common carotid artery (23). 3. Vagus nerve (22). 4. Sympathetic nerve (1). 5. Thoracic duct approach (12). 6. Dorsal somatic tributary of main vein (16). (Anlage of left superior intercostal vein). 7. Thyro-cervical artery (24). 8. Subclavian approach of jugular lymph sac (14), on the dorso-lateral aspect of the common jugular vein. This order and relationship is maintained uniformly and clearly in the three succeeding sections shown in f gs. 80, 81 and 82. In the last of these the dorsal venous tributary (16) enters the main common jugular channel. The sections can be easily followed and oriented by reference to the dorsal view of the reconstruction shown in fig. 90. The same is composed of the plates of all 44 sections of slide ix of series 245. In this embryo the thoracic duct approach has already established its definite and permanent connection with the inde- pendently developed channel of the preazygos segment of the thoracic duct (35, in fig. 90), as described in detail below (cf. p. 111). The point of this junction is indicated in the reconstruc- tion by the narrowing of the two spindles 12 and 35 (thoracic duct approach and preazygos segment of thoracic duct) to form an isthmus situated just cephalad of the leader marked 16-46. The reconstruction includes the terminal of the jugular ap- proach (13), with the origin from the same of the thoracic duct approach (12), and the entire subclavian approach (14) of the jugular lymph sac. The ventral division of the thyro cervical artery (24) passes forward between the jugular and subclavian approaches of the lymph sac, accompanied by a vein (not labelled in the figure) which is constantly found in this situation in the adult, draining from the prevertebral area into the common jugular trunk. 74 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS The dorsal and dorso-medial venous tributary plexus has united into the stem of the left superior intercostal vein (16-4-6}. The reconstruction shows well .the caudo-mesal turn of the thoracic duct approach (12) in the fork between the main vein (26) and the superior intercostal (16-46), accompanied by the mesal branch of the thyro-cervical artery. The terminal of the thor- acic duct approach (12) meets and unites with the preazygos segment of the thoracic duct (35) on the mesal shoulder of the superior intercostal vein terminal, between it and the sympathetic nerve (1). This is one of the two types of topographical relation- ship encountered (cf. infra pp. 75, 76). Further laterad follow the vagus (22) and the sympathetic strand (1), intersecting at an acute angle. Then the left common carotid artery (23) and the thymus (21) with the perithymic venous plexuses. Figs. 83 to 8,8 show the caudal end of the left jugular lymph sac and the thoracic duct approach in a selected series of sections from a 17 mm. cat embryo (series 142), all magnified 225 diame- ters, and fig. 91 gives the dorsal view of the reconstruction of this region from the same embryo. These preparations are introduced here for the purpose of illustrating, in comparison with the preceding series (245) the second type of relationship which the thoracic duct approach may bear to the combined vagus-sympathetic strand. Fig. 83 (series 142, slide x, section 13) shows the transection of the caudal end of the left jugular lymph sac in the interval between the internal jugular (25) and jugulo-cephalic (29) veins which are approaching their confluence to form the common jugular trunk. The form of the jugular lymph sac already indicates its coming division into the three terminal " approaches," the jugular ap- proach (14), the subclavian approach (13) and the thoracic duct approach (12). The latter covers the entire dorsal circumference of the internal jugular vein (25) and its termination points to the angular recess between the sympathetic (1) and vagus (22). Ventrad to the latter the field includes the thymus (21). The thyro-cervical artery is cut in several of its divisions, two of which are indicated by the forked leader 24- DEVELOPMENT OF THE THORACIC DUCT 75 In fig. 84 (series 142, slide x, section 16), the subclavian ap- proach (14) has separated from the jugular approach (13) and the thyro-cervical artery (24) is passing meso-laterad between these two divisions. The origin of the thoracic duct approach (12) from the dorso-medial circumference of the jugular approach (13) is well shown in the section. The concavity of the jugulo- cephalic arch is expressed by the two cross-sections labelled 29. In fig. 85 (series 142, slide x, section 8) the thoracic duct approach (12) has separated from the jugular approach (18). The curve of the process, with the concavity directed meso- caudad, gives the two cross sections of its lumen included in the forked leader 12. The termination of the thoracic duct approach has still further invaded the recess between sympathetic (1) and vagus (22) and a slight interval has developed between the two nerves. The lateral circumference of the vagus (22) covers practically the entire mesal aspect of the internal jugular vein (25). In the following figure (fig. 86, series 142, slide x, section 20) the termination of the thoracic duct process (12) is seen slipping through the interval between sympathetic (1 ) dorsally and vagus (22) ventrally. Instead 6f lying therefore dorso-lateral to the sympathetic as in the preceding series 245 the end of the thoracic duct approach in- the present instance passes mesad on the ventral aspect of this nerve, between the same and the vagus. This relationship is also seen in the two succeeding figures, 87 and 88 (series 142, slide x, sections 22 and 24.) We are dealing therefore apparently with two potential paths which the blind terminal of the thoracic duct approach may take in its passage from the jugular lymph sac mesad beyond the vagus- sympathetic line to the upper mediastinal region, where it estab- lishes eventually its secondary connections with the preazygos segment of the thoracic duct and with the broncho-mediastinal trunk. In one type (illustrated by the sections and reconstruc- tions of series 245 (figs. 79 to 82, fig. 90) the process dips, after passing between the left superior intercostal and main jugular vein at an acute angle, caudad on the lateral side of the sympa- thetic strand (fig. 90). 76 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS In the other type, illustrated by series 142 (figs. 83 to 88; fig. 91) the thoracic duct approach, maintaining the identical relation to left superior intercostal and common jugular veins, gains the ventro-medial aspect of the sympathetic cord, by passing ob- liquely between the same and the vagus. In schematic cross section these two conditions may be illus- trated by the accompanying text figures: In regard to the remaining structures fig. 87 (series 142, slide X, section 22) shows, by comparison with the preceding figure 86, the tap of the jugular approach into the internal jugular vein (25} just above the confluence with the jugulo-cephalic trunk (29], and the prolongation of its ventral process (13) over the ventral aspect of the common jugular angle. Fig. 88 (series 142, slide X, section 24) shows the confluence of the internal jugular (25) and jugulo-cephalic (29) veins fully established, and gives a very clear picture of the relation which the ventral prolongation (13) from the jugular approach of the lymph sac bears to the common jugular angle. In both figs. 87 and 88 the dorsal venous tributary (sup. intercostal vein anlage) is indicated by the leader 16. It enters the main vein in section 3 of the following slide (xi). Fig. 91 gives the dorsal view of the reconstruction of the vascular and nervous structures of the left side in slides DEVELOPMENT OF THE THORACIC DUCT 77 x and xi of this embryo, carried caudad to the level of the verte- bral vessels (41, 4)- Comparison with fig. 90 will show the correspondence of the two preparations in all essential points and will accentuate the above described difference in the course of the thoracic duct approach (12} and the preazygos segment of the thoracic duct (85} in relation to the vagus-sympathetic strand. It is readily seen, in comparing the reconstructions of the two older stages just described (figs. 90 and 91) with that of the earlier embryo shown in fig. 89, that the latter, in the construction of the terminal of its thoracic duct approach, combines poten- tially the ability of developing into either one of the two more advanced conditions just described. The process which the thoracic duct approach turns nearly horizontally mesad into the interval between vagus and sympathetic is present in all three reconstructions. In many embryos, as will appear presently in describing the critical stages in detail, the thoracic duct approach appears to make its connection with the independently developed anlages of the preazygos portion of the thoracic duct in this situa- tion, and continued development will lead to the condition seen in series 142 (fig. 91). If, on the other hand, the caudal extremity of the thoracic duct approach of the 13.5 mm. embryo (fig. 89) joins the independently developed lymphatic anlages at the root of the left superior intercostal vein and along the common jugular lateral to the sympathetic line, the resulting relation will be as seen in series 245 (fig. 90). These observations are further supported by comparison of the preparations just described with the corresponding portions of the two reconstructions shown in figs. 170 and 171. Fig. 170 (series 218) shows the stage in a 15 mm. embryo just prior to the junction of the thoracic duct approach with the inde- pendently formed preazygos segment of the thoracic duct (35}. The latter is in its greater part already a distinct and considerable segment of the future continuous duct channel. Between its blind cephalic extremity and the blind caudal end of the thoracic duct approach (12} are a number of scattered and still separate lymphatic anlages along the common jugular and innominate veins on each side of the sympathetic line. By continued exten- 78 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS sion of these as yet isolated lymphatic anlages the still lacking intermediate bondpiece will be produced, which, by union of its cephalic extremity with the thoracic duct approach of the jugular lymph sac and junction of its caudal end with the cephalh ternrnnl of the preazygos segment of the thoracic duct (35), will produce the continuous and uninterrupted lymph channel of the succeed- ing stage (se:ies 143, 15.5 mm. embryo, fig. 171). A study of the reconstruction of series 218 makes it evident that the opportunity still exists of developing this final junction either on the dorso- lateral or ventro-medial side of the sympathetic cord, for the lymphatic anlages exist in both situations. The end of the thoracic duct approach of this embryo rides with a well marked lateral and medial prolongation in the fork formed by the en trance of the left superior intercostal vein (16-4-6} into the common jugular. The lateral process, carrying the leader 12 in the figure, is in position to join the line of separate lymphatic anlages extend- ing caudad on the lateral side of the sympathetic, and a prominent cephalic process of the already established segment of the preazy- gos lymph channel extends cephalo-laterad across the dorsal surface of the nerve, as if to meet the lymphatic line just described. If development had proceeded along this line in this embryo, the condition seen i.n series 245 might have resulted, with the modifi- cation of swinging the thoracic duct approach to the lateral instead of to the medial side of the left superior intercostal vein, but carrying the thoracic duct along the dorso-lateral aspect of the sympathetic nerve. In this case the large separate lym- phatic anlage labelled 35' in fig. 170 would take an important share in building up the resulting lymph channel. On the other hand, the definite connection with the preazygos segment of the thoracic duct might have been established through the ventro- medial terminal prolongation of the thoracic duct approach. A piece of the sympathetic nerve has been removed in the recon- struction in order to show the position of this structure, in the interval between sympathetic nerve (1) and vagus (22). The long slender prolongation extending forward between vagus (22) and common carotid artery (23) is the path of junction for the ventral or broncho-mediastinal lymphatic trunk. I am under DEVELOPMENT OF THE THORACIC DUCT 79 the strong impression that the observed variations in the relation of the embryonic thoracic duct approach to the sympathetic nerve hinge largely for their production upon the share which the broncho-mediastinal duct is to take in individual cases in the establishment of the preazygos segment of the thoracic duct channel. (Compare relations of 35 and 37 in the schematic cross sections shown above in the text figures.) If now this ventro-medial end piece of the thoracic duct approach had effected its junction with the preazygos segment of the thoracic duct along the line of the lymphatic anlages on the medial side of the sympathetic and in the interval between this strand and the vagus, then the individual condition already de- scribed in series 142 (fig. 91) would have resulted. This same topographical relationship of sympathetic nerve and lymph channel is also seen in the 15.5 mm. embryo, series 143, shown in reconstruction from the right side in Fig. 171. The thoracic duct approach (12) intersects the sympathetic nerve (1) on its ventro-medial aspect at an acute angle to establish its con- nection with the cephalic end of the preazygos segment of the thoracic duct (35). This type appears the prevalent one in the embryos in my collection. I am strongly inclined, on the evidence of the embryos of my collection, to regard the development of the thoracic duct approach of the jugular lymph sac as occupying three distinct chronological stages : 1. Early stage. Embryos of 13-13.5 mm. Series 9213. mm. Series 18913.5 mm. In this period the thoracic duct approach appears large, at times multilocular, and shares the redundancy and capacious lumen which characterizes the entire jugular lymph sac in the earlier genetic period. 2. Intermediate stage. Embryos of 14 mm. Series 210 14 mm. Series 21214 mm. Series 12214 mm. The sac, as a whole, and the thoracic duct approach in particu- 80 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS lar, condenses and becomes more clearly defined against the sur- rounding mesoderm. The multilocular character due to rem- nants of the capillary walls in the interior, disappears and the approach obtains a clear and clean-cut lumen. The entire struc- ture appears to contract somewhat on itself and develops more distinct walls. The recession, during which the thoracic duct approach occupies a relatively smaller territory is succeeded by 3. Final stage. Embryos of 15-17 mm. Series 245 15 mm. Series 218 15 mm. Series 143 15.5 mm. Series 142 17 mm. In this period the sac, through its digitate processes, establishes the definite connections with the independently formed systemic extraintimal lymphatic channels. Specifically the thoracic d uct approach of the sac unites secondarily with the preazygos seg- ment of the thoracic duct and with the broncho-mediastinal lymphatic trunk. While I do not insist on the above outlined details in the chronological order of development of the thoracic duct approach, the embryos of my collection certainly tend to support the view expressed. Thus the 13 mm. embryo (series 92) offers the highest degree of expansion of the thoracic duct approach. The structure recedes and diminishes somewhat in extent in the 13.5 mm. stage (series 189), appears uniformly further reduced and shortened in the 14 mm. embryos (series 210, 212, 122), and finally, from the 15 and 15.5 mm. stages on, makes its definite connections with the preformed independently developed chan- nels of the thoracic and broncho-mediastinal ducts. In this entire genetic process there is not the slightest indica- tion of an extension of lymphatic trunks from the jugular sac or from its thoracic duct approach toward the periphery. There is absolutely no suggestion of so-called " budding" or " sprout- ing," or of any other form of centrifugal extension of lymphatic vessels. On the contrary, the sac as a whole, and the processes therewith connected, recede actually for a time, and only then establish a secondary junction with the independently developed DEVELOPMENT OF THE THORACIC DUCT 81 general systemic lymphatic channels. These have in the mean- while formed along and around the temporary embryonic veins, as the result of the confluence of large numbers of extraintimal perivenous spaces in the sense previously defined in detail. Their line of further growth, extension and union with each other is altogether centripetal, from the outlying regions toward the jugular lymph sacs, and their union with the latter's processes is the last and final step in the definite organization of the lym- phatic system. PART II, PLATES FIGURES 29 TO 91 FIGURE 29 29 Color schema, showing the four genetic segments of the thoracic ducts (I, II, III, IV) in the cat, and their relations to the embryonal veins, the broncho- mediastinal trunk (37), and the jugular lymph sacs (II). 3' Precardinal or precava of right side. 3 Right azygos vein, thoracic portion. 6 Precardinal or precava of left side. 6' Left azygos vein, thoracic portion. 7 Aorta. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 24 Thyrocervical artery. 25 Internal jugular vein. 26 Common jugular vein. 27 External jugular vein. 28 Cephalic vein. 33 Subclavian artery. 33' Subclavian vein. 35 Thoracic duct, preazygos segment. 36 Thoracic duct, azygos segment. 37 Broncho-mediastinal duct. 40 Innominate vein. 46 Left superior intercostal vein. 52 Postazygos segment of thoracic duct. THE SYSTEMIC LYMPHATIC VESSELS PLATE 22 27 28 IV. MEMOIR NO. 1, HUNTINGTON, 1911 29 FIGURES 30 AND 31 30 Transverse section through left side of lower cervical region in a 14 nun. cat embryo (series 210, slide VIII, section 39) , X 225. 31 Same, section 40. 1 Sympathetic nerve. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 18 Peritracheal venous plexus. 19 Suprathymic venous plexus. 20 Subthymic venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. THK SYSTEMIC LYMPHATIC VESSELS PLATE 23 16 18 25 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 32, 33 AND 34 32 Same, section 41. 33 Same, section 42. 34 Transverse section through left side of lower cervical region in a 14 mm. cat embryo (series 212, slide IX, section 24), X 200. 1 Sympathetic nerve. 8 Oesophagus. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 18 Peritracheal venous plexus. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 24 > ;V-v']r:.- > .-"f*>//^a^ *$. * .*. . .*^^*. *. * ^A " ** * - ^. , " 1 *'%& ^'i ' ,- v.'.' ' -vV/' 33 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 35 TO 41 35 36 Same, section 15. Same, section 16. 37 38 Same, section 17. Same, section 20. 39 Same, section 22. 40 Same, section 23. 41 Same, section 24. 1 Sympathetic nerve. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach 16 Dorso-medial somatic venous tributaries. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular vein. 46 Left superior intercostal vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 2.1 35 , 37 38 / V =-11 ** > 24 11 12 25 39 40 24 22 ^m s ^, 11 12 ^ 25 41 MEMOIR NO. 1, HUNTINOTON, 1911 FIGURES 42 AND 43 42 Same, section 26. 43 Transverse section through left side of lower cervical region of a 13.5 mm. cat embryo (series 189, slide VII, section 10), X 150. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 46 Left superior intercostal vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 26 ;.V :<$||&p P?^5S^S?^C ' . ' 't 'Jtf 1 ,*'.* *,*' , * ' ' f * * ^ M 16 k . " , " ' S&' /; V v '/'* i*''* v -*.*v V' 'x 46 ,'''. ' '.y f " ; '\ * ^. 'x^^ * **"-.''' *y^ 11 ' *'* N***' "'- "* NS - vf ; .f ^-r* ^ '+ c * *O^ 'M ', '. */ Ji i^-L,* "J % , '* '*"*.' : * : i. 'X I 1 ^, y, t '-^i^'?!' x * 1 1 *< "' v;r.^-y^ - *' V v V J V * *w ^ *^' 5/> % \^i ^va^j 1 'ta'V?* > /^ ^'} -' XK\ * ' >\- >*^'4 'V^, \A %4 99 ^^"4 &im %: * -I/. :'f *\ * > f > 1 n ii >V.-v V^ V ; ./^l Mttmft: 42 MEMOIR NO. 1. HUNTINGTON, 1911 FIGURES 44 TO 49 44 Same, section 11. 45 Same, section 12. 46 Same, section 13 47 Same, section 14. 48 Same, section 15. 49 Same, section 16. 1 Sympathetic nerve. 8 Oesophagus 9 Trachea. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk THE SYSTEMIC LYMPHATIC VESSELS PLATE 27 12 17 im 16 24 11 1 22 25 23 44 25 24 11 12 25 25 22 24 11 25 47 48 40IR NO. 1, HUNTINGTON, 1911 FIGURES 50 TO 55 50 Same, section 17. 51 Same, section 18. 52 Same, section 19. 53 Same, section 20. 54 Same, section 21. 55 Same, section 22. 1 Sympathetic nerve. 8 Oesophagus. 11 Jugular lymph sac. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 28 29 23 24 11 25 12 22 %>' - *y > * :^ >;.->-. <. v "> . v**-:'^-^ --^ *^ '; -5 V "-^^>^,:>/5^^^ 24 11 25 51 52 |M4 3w fr*fl,lt ."> v ft ..^*KiP ^ fy 12 y^* r ^''.^VY'*-'^'' ' J ' <>T /S^' *" " vi?^^(^ . gj^'^> "^ * - ' ^ - j^l ' -.* \'-'"* ^^Ssiv -^Myp 22 !^r^^e ^^^1^ 24 11 25 54 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 56 56 Dissection of left cervical region of adult cat, showing reduplication of thoracic duct approach of left jugular lymph sac (Adult No. 83). 1 External jugular vein and lymphatic-.-. 2 Common jugular vein. 3 Jugular lymph sac. 5 Subclavian vein. 9 Thyrocervical artery. 11 Vertebral vein. 13 Left subclavian artery. 18 Jugular approach and tap. 19 Subclavian approach and tap. 20 Innominate artery. 24 Vena cava superior. 25 Left carotid artery. 26 Aorta, THE SYSTEMIC LYMPHATIC VESSELS PLATE 29 18 20 26 13 MfiMOIR NO. 1, HUNTING-TON, 1911 FIGURES 57 AND 58 57 Transverse section through left lower cervical region of a 14 mm. cat embryo (series 122, slide IX, section 21) , X 150. 58 Same, section 28. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 26 Common jugular vein. 29 Jugulo-subclavian trunk. PLATE 30 THE SYSTEMIC LYMPHATIC VESSELS 1% '.**,. I..." > .**'* A ifeull FIGURES 59, 60 AND 61 59 Same, slide X, section 7. 60 Same, section 9. 61 Same, section 12. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 26 Common jugular vein. 30 Sixth spinal nerve. 31 Primitive ulnar veno-lymphatic. THE SYSTEMIC LYMPHATIC VESSELS PLATE 31 22 23 24 12 26 t I T i i' *i <" ' i AJT ,'&' *. ' w .. ~* ^*.'T?^% ^on* SI^^^^IFii ^^' : 0y^^ 60 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 62 AND 63 62 Same, section 13. 63 Transverse section of left lower cervical region in a 13 mm. cat embryo (series 92, slide X, section 27), X 225. 1 Sympathetic nerve. 8 Oesophagus. 11 Jugular lymph sac. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 26 Common jugular vein. 29 Jugulo-subclavian trunk. 30 Sixth spinal nerve. 31 Primitive ulnar veno-lymphatic. THE SYSTEMIC LYMPHATIC VESSELS PLATE 32 ' - : ! ,, 9& ifr.X 22 0.,'i0^-'.-wp , , fW.S^lSr -" V -J^ &\ V ^ - ; ; >Y * * jt .^^ *.J'';V-' -.,^f ^fe-3 lr*^A>'w v;J- 4 ^r^ * fTr*^^ % Jt 4-->V. ffi ^e * * 3 55jJU* ** . ^-4 Cl Z'&X' , .V ^,/-' *^* r f ' J'*/, P^;S\^>< IB^sS?! Wtt$*< f &Jr%, 3 Jl. ->?' IT-' ; i* #'' v -.- . ; -? ? " r . ./*^ t^ ft^'"* ""* 13 29 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 64, 65 AND 66 64 Same, section 31. 65 Same, section 32. 66 Same, section 33. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 33 29 64 29 14 24 13 *29 66 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 67, 68 AND 69 67 Same, section 35. 68 Same, section 36. 69 Same, section 38. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 34 JEMOIK NO. 1, HUNTINGTON, 1911 FIGURES 70, 71 AND 72 70 Same, slide IX, section 2. 71 Same, slide IX, section 5. 72 Same, slide IX, section 6. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 14 Jugular lymph sac, subclavian approach. 22 Vagus. 24 Thyrocervical artery. 26 Common jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 35 *?^imT STf aTMskJ-JfJ %*>,* 24 14 26 71 $ J^W* 3r^> ** 26 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 73, 74 AND 75 73 Same, slide IX, section 9. 74 Same, slide IX, section 10. 75 Same, slide IX, section 11. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 22 Vagus. 24 Thyrocervical artery. 26 Common jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 36 MEMOIR NO. 1. HUNTINGTON, 1911 FIGURES 76, 77 AND 78 76 Same, slide IX, section 12. 77 Same, slide IX, section 13. 78 Same, slide IX, section 14. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 22 Vagus. 24 Thyrocervical artery. 26 Common jugular vein. PLATE 37 ' * < ^k.^r*' * . * * * -* r-V.-,vJ^ \^v J^fln. MI *tfl^*9fe^. ,_""^i> *i 26 MEMOIR NO. 1, HCNTINGTON, 1911 EXPLANATION OF FIGURE 79 79 Transverse section of left lower cervical region in a 15 mm. cat embryo (series 245, slide XI, section 24), X 225. 1 Sympathetic nerve. 8 Oesophagus. 12 Jugular lymph sac, thoracic duct approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 26 Common jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 38 79 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 80, 81 AND 82 80 Same, section 25. 81 Same, section 26. 82 Same, section 27. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 26 Common jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 39 * * ^hstF^w^** - WJSi^ ^Ift A* ^^r^Jjf^yif^sfi^fo. "H^ /V 26 80 . ..v: v --*Sfr,Y .r*J fjS x **; -vr .;/;?-& .>'>:v:^ . /,"; 81 12 16 24 -i MKMOIR XO. 1, HUNTINGTON, 1911 FIGURES 83 AND 84 83 Transverse section of left lower cervical region in a 17 mm. cat embryo (series 142, slide X, section 13), X 225. 84 Same, section 16. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 21 Thymus. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 40 29 2J MEMOIR NO. 1, HUNTINCiTON, 1911 FIGURES 85 AND 86 85 Same, section 18. 86 Same, section 20. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 21 Thymus. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular vein. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 41 21 13 MKMOIlt NO. 1, HUNTINGTON, 1911 FIGURES 87 AND 88 87 Same, section 22. 88 Same, section 24. 1 Sympathetic nerve. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 14 Jugular lymph sac, subclavian approach. 16 Dorso-medial somatic venous tributaries. 21 Thymus. 22 Vagus. 24 Thyrocervical artery. 25 Internal jugular lymphatics. 29 Jugulo-subclavian trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 42 24 14 29 13 24 21 MKMOIH NO. 1. HUNTINCJTON, Iflll DEVELOPMENT OF THE THORACIC DUCT 83 2. THE PREAZYGOS SEGMENT This portion of the thoracic duct develops, again by confluence of numerous extraintimal perivenous anlages, in the cephalic and lateral mediastinal space, between the termination of the thoracic duct approach of the jugular sac and the level of the intersection of the left duct with the aortic arch. Its ontogenesis offers the most striking and definite evidence of the development of the systemic lymphatic channels in this mammalian embryo as the result of the confluence of a large number of originally separate and independent "extra-intimal" or " perivenous" mesodermal spaces, following and enveloping the early embryonal veins, or completely surrounding them, and finally entirely replac- ing them. The spaces which I am describing as the anlages of the systemic lymphatic channels in the mammalian embryo under consideration (Felis domesticd) are surely easy enough to see. If they are followed in successive stages they fully and completely reveal the histogenesis of the systemic lymphatic vessels in this mammalian form, and conclusively establish the relation of the first mammalian lymphatic anlages to the embryonic veins, which they are destined to replace, as previously outlined in the prelim- inary publications quoted (1, 2, 21, 22). The structures which form the first anlages of the systemic lymphatic channels in the embryo of the cat cannot be injected, because at the time of their first definite appearance they represent the still disjointed links of a chain which is only subsequently to be assembled into a continuous whole. It is possible to inject in earlier periods the veins which are later replaced by these extraintimal lymphat- ics, and it is possible to inject in later stages the lymphatics when they have united into a connected system of channels. But in either case injection methods merely prove topographi- 84 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS cal conditions for any given stage. They afford absolutely no proof of the histogenetic origin of vascular structure. The mere fact that a lymphatic plexus occupies topographically the place of an antecedent venous plexus is absolutely no proof of the histogenetic identity of the channel system involved. However, to return to the matter in hand, while the region under discussion serves no better than any other part of the embryo for this pur- pose, it fully and completely illustrates the relative condition of embryonic veins and the replacing extraintimal lymphatics in the genesis of the adult systemic lymphatic channels. The ontogenetic history of the preazygos segment of the thoracic duct must be considered in reference to three regions: A. The development of a broncho-mediastinal or ventral mediastinal lymphatic channel, draining the ventral and lateral mediastinal areas, cephalad of the pericardium, and including the lymphatic return from the thymic, tracheal, bronchial, and oesophageal regions. B. The development of the proximal portion of the thoracic duct proper, between the termination of the thoracic duct ap- proach of the jugular lymph sac, and the beginning of the azygos segment of the thoracic ducts, caudal to the level of the aortic arch. C. The junction of these two independently formed lymphatic trunks with each other and with the thoracic duct approach of the jugular lymph sac. Adult Conditions Before taking up the embryological consideration of the develop- ment of these channels and the establishment of their subsequent connections, it seems advisable, on account of the complicated topographical relations, to establish these clearly by reference to a small selected series of adult animals with successful injection of the lymphatic structures concerned. This becomes all the more necessary because the adult variations observed in the arrangement of the lymphatics of this region have an important DEVELOPMENT OF THE THORACIC DUCT 85 bearing on the observed embryological conditions. An under- standing of the former is necessary for the correct interpretation of the latter, and vice versa. Hence, while the details of the adult anatomy of the cat's lymphatic system in its relation to the development are reserved for Part V of the second publica- tion of this series, a limited number of adult illustrations are here introduced in order to prepare the ground for the consideration of the ontogenesis of the broncho-mediastinal trunk and of the preazygos segment of the thoracic duct. In a certain proportion of adults the ventral or broncho- mediastinal trunk has no direct connection with the preazygos segment of the thoracic duct, but drains solely cephalad into a ventral prolongation from the subclavian approach of the jugular sac, which process also receives the main lymph vessels from the axilla and the forelimb. Figs. 92 and 93 (adults, nos. 67 and 13) give good illustrations of this condition. In many other adults, however, the left ventral or broncho- mediastinal trunk is connected caudally, with or without the interposition of one or more lymph nodes common to both, with the preazygos segment of the thoracic duct. This junction com- monly occurs about at the level at which the innominate artery divides. The cephalic extremity of the main broncho-mediastinal trunk then usually still drains into the jugular lymph sac by establishing a connection with the ventral prolongation which the subclavian approach of the sac sends caudad over the ventral aspect of the jugulo-subclavian confluence. This process also receives the main lymphatic trunks from the anterior extremity and from the axillary area of the thoracic wall. Occasionally the cephalic end of the broncho-mediastinal trunk also connects with the similar ventral prolongation which the jugular approach of the lymph sac sends down over the ventral aspect of the common jugular angle of confluence. The broncho-mediastinal trunk is therefore usually represented v by a ventral lymph channel of some size, whose caudal extremity joins the preazygos segment of the thoracic duct, while its cephalic extremity empties, in common with other axial and appendicular 86 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS lymph paths, into the caudal end of the jugular sac, usually into the ventral process of the subclavian approach caudal to the jugulo-subclavian lymphatico- venous tap. In this course the broncho-mediastinal trunk receives afferents from the medias- tinal structures. Fig. 94 shows this general arrangement on the left side in a typical adult (No. 56), in which the left jugular lymph sac connects with the veins only through the single jugulo- subclavian tap. A double ventral process from the subclavian approach effects the cephalic connection of the broncho-medias- tinal trunk with the jugular sac, while caudally two lymph vessels unite it to the preazygos segment of the thoracic duct. It is evident in following the thoracic duct caudo-cephalad that this arrangement enables the flow of lymph in the main canal to take, near the level of the innominate bifurcation, one of two equally available pathways in order to reach the jugular sac and through it enter the venous stream: a. It may, in large part or entirely, continue cephalad in the dorsal channel of the preazygos segment of the thoracic duct proper, or 6. It may ascend in the channel of the broncho-mediastinal trunk, and empty into the venous system through the ventral prolongation of the subclavian approach of the jugular lymph sac. This arrangement, which furnishes the keynote for the main variations of the thoracic duct in this region, and for the interpre- tation of certain important embryological stages, is well illustrated by the dissection of the left side of neck and thorax in the adult cat shown in fig. 95 (adult, no. 22). The azygos segment of the thoracic duct is seen caudad of the level of azygos arch (21) as a plexiform channel whose meshes are perforated by the inter- costal arteries. Cephalad of the azygos arch the preazygos seg- ment of the thoracic duct continues as an undivided canal of larger caliber to the level of division of the innominate artery (20). At this point the broncho-mediastinal trunk (8) diverges from the main canal and ascends ventral to the left subclavian artery, receiving tributaries from the ventral mediastinal nodes. It meets the subclavian lymphatic and a large ventral mediastinal DEVELOPMENT OF THE THORACIC DUCT 87 trunk accompanying the internal mammary vessels (7) in front of the jugulo-subclavian angle in a plexiform junction, which empties into a ventral process prolonged down from both the jugular (18) and subclavian (19) approaches of the jugular lymph sac. The preazygos segment of the thoracic duct, cephalad of its connection with the broncho-mediastinal trunk, takes the usual course upward under cover of the subclavian artery and behind the vertebral vein (11) and opens into the dorsal aspect of the jugular sac a little in front of the common jugular tap. The jugular lymph sac is well developed in its ventral division and opens into the venous system at the two typical points, viz., both at the common jugular and jugulo-subclavian taps. It receives the internal and external jugular and the cephalic lym- phatics in the usual way. The subclavian vein (5) is divided close to the jugulo-subclavian confluence and turned forward to expose the vertebral vein and the duct. The individual affords a good example of the full development of the normal preazygos segment of the thoracic duct in combina- tion with a well marked broncho-mediastinal trunk and plexus. The former is still the main lymphatic channel, although the broncho-mediastinal vessel is fully developed and capable of draining the mediastinal lymphatics either cephalad into the lymph sac or caudad into the thoracic duct. Fig. 96 (adult, no. 29) shows an instance in which the preazy- gos segment of the thoracic duct (12) and the broncho-mediastinal trunk (8) are of nearly equal size, the former receiving the inter- nal mammary (6) and inferior thyroid (4) lymphatics and drain- ing into the jugular sac through the ventral prolongation of its subclavian approach. The preparation shows well the main relations of the lymphatic complex to the vertical portion of the left subclavian artery (13) and its branches, with the broncho- mediastinal trunk (8) lying ventral to the artery, while the tho- racic duct (2) ascends dorsal to the same, and further passes be- neath the vertebral vessels (11) and the thyro-cervical artery (9) in order to reach its connection with the lymph sac. The latter has in this instance a single tap into the veins through the subclavian approach at the jugulo-subclavian angle. The subclavian vein (5) 88 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS has again been cut short and turned mesad to expose the deeper structures. The azygos segment of the thoracic duct (16) shows the same plexiform character and relation to the perforating intercostal arteries as in the preceding dissection (fig. 95). Fig. 97 (adult no. 131) shows an instance in which the main thoracic duct turns ventrad into the broncho-mediastinal trunk, while a small channel follows the usual dorsal course of the normal duct. The left jugular lymph sac has only a single tap into the venous system at the jugulo-subclavian angle. The sac receives at its cephalic end two lymphatic trunks paral- leling the internal jugular vein (17), which is larger than usual, and the combined trunk formed by the external jugular and cephalic lymphatics. The dorsal half of the thyro-cervical periarterial ring is smaller than the ventral, but the foramen is complete. The jugulo-subclavian approach and tap is formed by a large process of the sac which sends a ventral prolongation over the jugulo-subclavian angle. This receives laterally the subciavian lymphatic, ascending along the subciavian vein (5), and from below the broncho-mediastinal trunk (8), which represents the main cephalic continuation of the thoracic duct. The duct at the aortic arch level (26) splits into a smaller dorsal (12) and larger ventral division (8). The former ascends behind the vertical part of the subciavian artery, crosses behind the vertebral vein, and turning ventro-caudad around the lateral aspect of the thyro- cervical artery (9), enters the dorsal aspect of the jugular lymph sac, just in front of the jugulo-subclavian tap. This smaller dorsal division represents the usual preazygos portion of the thoracic duct of the adult. The second larger division of the duct takes the path of the broncho-mediastinal channel (8). It ascends first on the outer side and then in front of the subciavian artery, subdivides into two branches, which reunite and receive a small lymphatic from the left innominate plexus. The upper end of the duct crosses in. front of the junction of the vertical and arched portions of the DEVELOPMENT OF THE THORACIC DUCT 89 subclavian artery and in front of the subclavian vein to terminate, in common with the subclavian lymphatic, in the ventral process of the subclavian approach of the lymph sac. More rarely the broncho-mediastinal channel forms the only terminal path of the thoracic duct. In these instances the preaortic segment of the thoracic duct abandons its normal dorsal position and turns ventrad into the path of the broncho-medias- tinal channel. Thus in fig. 98 (adult, no. 168), in which the common jugular tap (18) is the only lymphatico-venous con- nection of the left jugular lymph sac, the ventral process of the jugular approach is free and prolonged caudo-ventrad over the jugulo-subclavian angle as a large trunk which receives the entire drainage of the thoracic duct. 'The latter (8), after passing the aortic arch, swings ventrad of the subclavian artery into the pathway of the normal broncho-mediastinal trunk and ascends to meet the ventral prolongation of the jugular approach of the lymph sac. It thus becomes apparent that in dealing with the ontogenesis of the cephalic portion of the cat's thoracic duct it is necessary to reckon with the ventral mediastinal and broncho-mediastinal channels as well as with the preazygos segment of the thoracic duct proper, since the former, as just seen, may take over, par- tially or entirely, the drainage of the duct into the jugular lymph sac. A composite schema of the adult preazygos portion of the tho- racic duct in relation to the broncho-mediastinal and ventral mediastinal trunks is given in fig. 99 based on the individual con- ditions encountered in the adult series examined. The plan rep- resents every recorded development of the lymphatic pathways in this region, but is rarely seen in its full development in the adult (cf. adult, no. 29, fig. 96). Usually one or more of these po- tential segments and connections default and thus a great variety of distinct adult conditions is produced. These lymphatic path- ways are, however, all present in the embryo and account for the observed adult variants. PART II, PLATES FIGURES 92 TO 99 FIGURES 92 AND 93 92 Dissection of the veins and lymphatics at the root of 1 he nrck in an adult cat, (No. 67). 93 Same., No. 13. (Figs. 92, 93, and 94 are republished from Huntington and McClure's " Anatomy and Development of the Jugular Lymph Sacs in the Domestic Cat," Am. Jour. Anat., Vol. X, No. 2.) TIIK SYSTEMIC LYMPHATIC VESSELS INT. JUGULAR VHN LEFT EXT. JUGULAR VEIN THYRO-CERVIC COMMON JUGULA COMMON JUGU SUBCLAVIAN APPROAC INNOMINATE VE 92 INT JUGULAR VEIN RIGHT EXT. JUG. V THVRO-CERVICAL A JUGULO-CEPHALIC TRUN JUGULO SUBCLAVIAN T A p INNOMINATE VEIN HYRO-CERVICAL A UGULAR TAP AN APPROACH SUBCLAVIAN V. THORACIC DUCT LEFT EXT. JUG V EPHALIC VEIN THYRO-CERVICAL A. OMMON JUGULAR TAP SUBCLAVIAN V. COMMON JUGULAR V. THORACIC DUCT 93 MEMOIR NO. 1, HUNTINGTON, Iflll FIGURE 94 94 Same, No. 56. THE SYSTEMIC LYMPHATIC VESSELS PLATE 45 INT. JUGULAR VEIN LEFT EXT. JUG. V. RIGHT EXT. JUG. V CEPHALIC VEIN \ THYRO-CERV1CAL A JUGULO- CEPHALIC TRUN COMMON JUGULAR V JUGULO ,mm / SUBCLAVIAN SUBCLAVIAN V SUBCLAVIAN APPROACH INNOMINATE VEIN CEPHALIC VEIN THYRO-CERVICAL A. JUGULO-CEPHALIC TRUNK SUBCLAVIAN APPROACH SUBCLAVIAN V JUGULO-SUBCLAVIAN TAP INNOMINATE VEIN 94 MEMOIR NO. 1, HUNTINGTOK, 1911 FIGURE {)> 95 Dissection of main lymphatic vessels of neck and thorax in an adult cat (No. 22). 1 External jugular vein and lymphatics. 5 Subclavian vein. 7 Internal mammary artery and lymphatics. S Broncho-mediastinal or ventral mediastinal lymphatic trunk. 10 Cephalic vein and lymphatics. 11 Vertebral vein. 12 Thoracic duct, preazygos segment. 15 Oesophagus. 17 Internal jugular vein and lymphatics. 18 Jugular approach and tap. 19 Subclavian approach and tap. 20 Innominate artery. 21 Azygos vein. TIIK SYSTEMIC LYMPHATIC VESSELS PLATE* 46 21 17 12 15 MEMOIK NO. 1, HUNTINGTON, 1911 FIGURE 90 96 Dissection of main lymphatic vessels of neck and thorax in an adult cat (No. 29). 1 External jugular vein and lymphatics. 2 Common jugular vein. 3 Jugular lymph sac. 4 Inferior thyroid lymphatics. 5 Subclavian vein. 6 Internal mammary lymphatics. 7 Internal mammary artery. 8 Broncho-mediastinal or ventral rnediastinal lymphatic trunk 9 Thyrocervical artery. 10 Cephalic vein and lymphatics. 11 Vertebral vein. 12 Thoracic duct, preazygos segment. 13 Left subclavian artery. 14 Junction of preazygos and azygos segments of thoracic duct. 15 Oesophagus. 16 Azygos segment of thoracic duct. THE .SYSTEMIC LYMPHATIC VESSELS PLATE 47 MEMOIR NO. 1, HONTINGTON, 1911 FIGURE 97 97 Dissection of main lymphatic vessels of neck and thorax in an adult cat (No. 131). 1 External jugular vein and lymphatics. 2 Common jugular vein. 5 Subclavian vein. 7 Internal mammary artery. 8 Broncho-mediastinal or ventral mediastinal lymphatic trunk. 9 Thyrocervical artery. 12 Thoracic duct, preazygos segment. 13 Left subclavian artery. 15 Oesophagus. 16 Azygos segment of thoracic duct. 17 Internal jugular vein and lymphatics. 20 Innominate artery. 23 Innominate vein. 26 Aorta. 27 Vagus nerve. 28 Inferior thyroid vein. 29 Jugulo-cephalic venous trunk. THE SYSTEMIC LYMPHATIC VESSELS PLATE 48 97 MKMOIH NO. 1, HUNTINGTON, 1911 FIGURE <)S 98 Dissection of left jugular lymph sac and thoracic duct in an adult cat (No 168). 2 Common jugular vein. 5 Subclavian vein. 8 Broncho-mediastinal or ventral mediastinal lymphatic trunk. 9 Thyro-cervical artery. 13 Left subclavian artery. 15 Oesophagus. 17 Internal jugular vein and lymphatics. 18 Jugular approach and tap. 21 Azygos vein. 22 Trachea. 23 Innominate vein, 24 Vena cava superior 25 Left carotid artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 49 22 17 98 MEMOIR NO. 1, HUNTINGTON, 1911 FKHIRE 09 09 Schema showing composite picture of left jugular lymph sac, broncho me diastinal trunk and preazygos segment of thoracic duct. 1 External jugular vein and lymphatics. 3 Jugular lymph sac. 5 Subclavian vein. 6 Internal mammary lymphatics. 8 J3roncho-mediastinal or ventral mediastinal lymphatic trunk. 9 Thyro-cervical artery. 10 Cephalic vein and lymphatics. 11 Vertebral vein. 12 Thoracic duct, preazygos segment. 13 Left subclavian artery. 16 Azygos segment of thoracic duct. 17 Internal jugular vein and lymphatics. 18 Jugular approach and tap. 19 Subclavian approach and tap. 20 Innominate artery. 23 Innominate vein. 24 Vena cava superior. 2i3 Left carotid artery. 26 Aorta. THE SYSTEMIC LYMPHATIC VESSELS PLATE 5(1 17 10 MEMOIR NO. 1, HUNTINGTON, 1911 DEVELOPMENT OF THE THORACIC DUCT 91 The Development of the Broncho-Mediastinal Trunk and of the Preazygos Segment of the Thoracic Duct The area in which the development of the broncho-mediastinal and preazygos segments of the thoracic duct occurs is shown in fig. 100, which gives a topographical view of a transection of the upper thoracic region in a 12 mm. embryo (series 78, slide IX, section 5, X 50). The section includes the entire coelom ventrally, and the structures between it and the notochord dorsally. The right (3) and left (6) precaval veins occupy the lateral limits of the dorsal region. In the interval between them are seen the oesophagus (8), the trachea (9), the vagi (22) and the dorsal aortic arches (7). Dorsal to the main veins on each side are the strands of the brachial plexus (not labelled in the figure), and dorsal to the nerves is the primitive ulnar veno-lymphatic (31). Further mesad follow the subclavian arteries (33), the entrance of the dor so-medial venous tributaries (16) into the main veins, and the sympathetic strands (1). The letter Y in the figure, between the notochord and the dorsal mediastinal structures named, indicates the center of the area in which the development of the preazygos segment of the thoracic duct will proceed in the suc- ceeding stages. The ventral portion of the field contains the coelom, with right (48) and left (49) auricles, right ventricle (50), ascending aorta (7) and ascending trunk of main pulmonary artery (10). Dorsal to this are seen the cross-sections of the right and left pulmonary arteries (10) descending to the lungs. Between them ventrally, and trachea (9), aortae (7), vagi (22) and praecaval veins (3, 6) dorsally, is the area marked by the letter X, in which the develop- ment of the broncho-mediastinal trunk will proceed. A . The development of the broncho-mediastinal or ventral medias- tinal lymphatic channels, which drain subsequently caudo-sinistrad into the left thoracic duct, or cephalo-sinistrad into the subclavian approach of the jugular lymph sac, or into both. In the earlier stages (embryos of 11-14 mm.), an extensive ventral venous capillary network develops between the main 92 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS venous trunks of the right and left sides, involving the distal part of the internal jugular, the common jugular and the innomi- nate veins. Out of this general antecedent sinistro-dextral capillary reticu- lum the definite retro- or suprathymic innominate anastomosis develops subsequently in normal individuals, but in the earlier stages this plexus extends both ventral and dorsal to the thymus anlage. The venous network develops, in addition to the prevertebral and dorsal somatic components, chiefly on the lateral aspect of the oesophagus, in the peri-tracheal area, and around the thymus and the common carotid artery. It trends cephalo-laterad towards the main venous lino (jugulo-innominate), into whose mesal sur- face the terminals of the plexus discharge. In so doing they encounter, and partially surround, the sym- pathetic and vagus nerves, and the thymus body. Consequently the terminal branches of this ventral mediastinal plexus enter the main vein both ventral and dorsal to these three structures and through the intervals between them. Thus the composite plan of this venous area appears as shown in the schematic fig. 101, which represents the combination of a number of transverse sections of the anterior thoracic region. The terminal connections of the ventral mediastinal plexus may be grouped in reference to their derivation and their relation to adjacent fields of venous drainage into the following five sets, (fig. 101): 1. Dorso-medial somatic tributaries of internal and common jugular and of innominate veins (fig. 101, 16} entering the main venous channel dorsal to the sympathetic nerve, and frequently combined with 2. Terminal of prevertebral venous plexus anterior to azygos veins (fig. 101, 17). 3. Terminal of peritracheal plexus passing laterad between sympathetic (1) and vagus (22) nerves (fig. 101, 18). 4. Ventral mediastinal branch, draining, as " suprathymic plexus," the area between the thymus (21), carotid (23) and vagus (22). This is the line of the future typical permanent left innominate anastomosis dorsal to the thymus (fig. 101, 19). DEVELOPMENT OF THE THORACIC DUCT 93 5. Subthymic plexus, passing ventral to the thymus anlage (fig. 101, 20). This forms, if retained and further developed in the adult, the occasionally observed "prethymic" left innominate vein (36) Any number of good examples of all these conditions are afforded by every 13 or 14 mm. embryo. Thus in a 14 mm. embryo (series 210, slide viii, section 41, X 225), fig. 102, direct com- parison of the schema shown in fig. 101 with the section will, without further description, make the arrangement of the ele- ments composing this plexus (16, 17, 18, 19, 20) clear. (Com- pare also fig. 43 [series 189, slide vii, section 10, 13.5 mm.] and fig. 30 [series 210, slide viii, section 39, 14 mm.]) The ventral divisions of this plexus extend in the earlier em- bryonic stages (11-12 mm.) as a well developed capillary reticu- lum into the upper thoracic region around and between oesopha- gus, trachea, precardinal veins, vagi and pulmonary arteries. Beginning in some 12 mm. embryos, but more constantly and characteristically in the 13 mm. stage, certain components of this ventral mediastinal venous plexus appear partly surrounded by independently developed extraintimal lymphatic spaces, which form the anlages of the future broncho-mediastinal lymphatic trunk. Thus, if the ventral portion of this venous plexus is followed caudad into the upper thoracic region, into the area of the develop- ing broncho-mediastinal trunk marked X in fig. 100, the following observations can be made in stages of the proper length, and adequately fixed and stained: 1. In embryos between 11 and 12 mm. only venous capillaries are found. Thus figures 103 and 104 show transverse sections of this part of the upper thoracic region in a 12 mm. embryo (series 217, slide' viii, sections 21 and 29, X 200). The field, caudal to oesophagus and aorta (7), contains in the middle line the trachea (9), the vagi (22), and further ventrad, the pulmonary 38 Geo. S. Huntington : ' 'Contribution to the topographical anatomy of the thorax in the foetus at term and the new-born child." Rep. Soc. N. Y. Lying-in- Hospital, 1897, p. 343, fig. xxxi. 94 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS arteries (10). The branches of the peritracheal venous plexus (18) are well developed. There is as yet no indication of lym- phatic development. 2. In 13 mm. embryos certain of the venous radicles entering into this plexus, are partly surrounded and enveloped by ex- traintimal lymphatic spaces, which arise independently of the venous system, as the first anlages of the future ventral medias- tinal lymphatic channel. Figures 105 and 106 show transverse sections of this region in a 13 mm. cat embryo (series 92, slide vii, sections 30 and 33) in a magnification of 225 diameters. Ventral to trachea (9), aorta (7) and left vagus (22) are seen sec- tions of a number of venous radicles (4) of the ventral mediastinal plexus, some of which are partially surrounded by very small extraintimal perivenous lymphatic anlages (5). Only a few of the most prominent of these are indicated in the figures by leaders, but they exist in many other parts of the field. A similar section from another 13 mm. embryo (series 107, slide ix, section 40, X 225) is shown in figure 107. Dorsal to the pulmonary arteries (10), between them and the trachea (9) and aorta (7), the section has cut a prominent extraintimal lym- phatic space (5) in typical relation to one of the radicles (4) of the ventral mediastinal venous network. 3. In the 13.5 mm. embryo the full and convincing proof of the extraintimal derivation of this channel is given. In some embryos of the period the process is still seen in its earlier stages. Thus figs. 108 and 109 (13.5 mm. embryo, series 223, slide x, sections 11 and 12, X 225) show on each side, ventral to the trachea (9) and the vagi (22), a number of radicles (4) of the ven- tral mediastinal capillary plexus. One of these small vessels, symmetrically disposed on each side in front of the respective vagus, is in process of being surrounded and replaced by an ex- traintimal lymphatic space (5). Other 13.5 mm. embryos show a rapid advance in the peri- venous lymphatic development. Fig. 110 shows a transverse section of the upper thoracic region of another 13.5 mm. embryo (series 189, slide viii, section 35, X 225). Just ventro-mesad of the left vagus nerve and its DEVELOPMENT OF THE THORACIC DUCT 95 encircling vein is a venous radicle (4) almost completely sur- rounded by an extraintimal lymphatic space (5) in the process of replacing the atrophying vein with which it is so closely asso- ciated. The same structures (4, 5} are seen still more clearly defined in the corresponding position on the right side of the embryo. Comparison with the corresponding sections just given in figures 103 to 109 show that the extraintimal anlage of the earlier stages, developed along the identical venous radicle, has increased in the 13.5 mm. stage, so as to nearly envelop the vein, and that the latter, if followed cephalad and caudad, is now separated from the functional venous channels of this region and is in the process of further recession and degeneration as the lymphatic perivenous space enlarges and more and more replaces the antecedent venous channel. The vein in question in the older embryo (series 189) appears collapsed and shrunken, and contains only a few degen- erating red blood cells. In the same situation on the right side of fig. 110 (series 189, slide viii, section 35), the section has cut the corresponding vein and the enveloping extraintirr al space at right angles, so that the central kernel of the shrinking vein (4), still containing a few degenerating red blood cells, is nearly com- pletely surrounded by the replacing extraintimal lymphatic (5), Fig. Ill shows the following section of the same embryo (series 189, slide viii, section 36), magnified 225 diameters, in a larger field which gives the topographical relations and may serve in in the orientation of the succeeding plates. The ventral circum- ference of the oesophagus (8) and the aorta (7) are seen dorsally. The trachea (9], with the vagi (22} on either side, occupies the central area. Ventrally are the two pulmonary arteries (10). The section is a most important one in interpreting the early histogenetic stages of the broncho-mediastinal lymphatic trunk and its relation to the embryonic mediastinal venous plexus. The same venous radicle (4), ventral to the vagus (22), already noted in the preceeding section (fig. 110), is seen on each side. On the left side it has been cut obliquely by the plane of the section, and hence shows a segment of considerable length, as a practically empty and partially collapsed endothelial bag (4), 96 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS surrounded by the extraintimal lymphatic space (). On the right side of the embryo the corresponding venous radicle (4) and the perivenous lymphatic anlage (5} have been again cut more at right angles, so that the envelopment of the former by the latter can be clearly traced nearly throughout the entire cir- cumference. Fig. 11 1 A shows the decadent vein and the replacing extrainti- mal lymphatic on the left side of this section in a higher magnifi- cation, ( X 300). The structure of the collapsed venous endothelial tube, and the beginning development of endothelial character in the mesodermal cells limiting the perivenous extraintimal lym- phatic anlage, can be here clearly made out. Figs. 112 to 116 show the five succeeding sections of the same embryo (series 189, slide viii, sections 37 to 41 inclusive) in a magnification of 225 diameters. I publish this series of plates in order to prove on the one hand that the extraintimal lymphatic anlages can be followed with certainty and accuracy in their relation to the contained decadent venule for longer or shorter distances, and that, on the other hand, they are, in these earlier stages, still isolated and not yet continuous with similar anlages in adjacent areas. They form, in other words, in these earlier periods, detached links of a lym- phatic chain, not yet united to each other to form the continuous channel of the later stages. In figs. 112, 113 and 114 (sections 37, 38 and 39) beautiful and clear pictures of the atrophying central venule (4) and the peri- venous extra-intimal lymphatic space (5) are found on both sides in the typical situation noted in the preceding sections, with which they of course can easily be coordinated by simple com- parison. Sections 38 and 39 (figs. 113 and 114) are particularly instruc- tive inasmuch as the detached central vein (4) on the left side still contains a few degenerating red blood cells, while its former connection with the valid and functional venous plexus around the left vagus nerve can still be faintly traced. In section 40 (fig. 115) the atrophying vein (4) and the sur- rounding lymphatic (5) of the left side appear divided into two DEVELOPMENT OF THE THORACIC DUCT 97 components. On the right side the corresponding structures end blindly. In the following section (41, fig. 116) the left venule and lym- phatic also terminate almost entirely in the indifferent mesoder- mal tissue, only a small remnant (4, 5) persisting between trachea and left vagus. This disappears entirely in the following section. The succeeding stage (14 mm.) shows the same lymphatic anlages surrounding and replacing the atrophying venules of the ventral mediastinal plexus in still higher development. It ap- pears from the study of numerous embryos of this length, that the average 14 mm. stage represents the point in the lymphatic ontogenesis of this region at which the pictures are most strik- ing. The central kernel of the decadent vein is still large and, in places, as yet incompletely separated from the definite venous channels. At the same time the perivenous lymphatic space has acquired an increased lumen and hence the two combined struc- tures occupy a relatively large field. Thus figs. 117 to 122 show transverse sections of the ventral thoracic region in a 14 mm. cat embryo (series 212, slide x, sec- tions 4, 5, 6, 7, 10 and 11, X 225). In figs. 117 and 118 the developing lymphatic anlages (5) of the broncho-mediastinal trunk can fce readily recognized and present the same relation to the atrophying venules, which they surround, as in the sections of the preceding series just described and figured. In contra- distinction to these structures the components of the permanent functional venous plexus of this region (32) form well-differen- tiated vascular channels. In fig. 119 (series 212, slide x, section 6) the scattered lym- phatic anlages occupying in the two preceding sections the space between the left pulmonary artery (10) and the left vagus (22), with its accompanying ventral mediastinal vein (32), are gathered together into a more extensive elongated space (5). In the interior of this is seen the endothelial bag of the empty and atrophying venule (4) which the lymphatic is about to replace. The vein remnant is almost completely surrounded by the lym- phatic space and only connected to the latter's endothelial wall by a bridge at one point. Other areas of similar lymphatic devel- 98 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS opment are seen on the opposite side (4-6) and (not labelled) in the center of the field, but the structures on the left side are, as in the earlier stages (cf. series 189) better developed and larger. The succeeding figures, 120, 121 and 122 (sections 7, 10 and 11) show the same structures, which can be distinctly followed for some distance caudad of the last section figured. In fig. 122 the original connection of the large upper venous kernel of the left side (4) with one of the components of the peritracheal venous plexus (18} is still suggested by the arrangement of the adjacent cells, although the degenerating vein has been cut off from all actual communication with the permanently valid channels of the mediastinal venous plexus. The 14 mm. stage is so important, and the results gained by the examination of a large number of embryos so conclusive in respect to the true ontogenesis of the systemic lymphatic channels, that I have selected a well marked example and publish here eighteen consecutive, or nearly consecutive, sections from the same individual. I do this in order to show the extent to which extraintimal perivenous lymphatic development has proceeded in this stage, in respect to the length of the resulting segments of the future lymphatic channel, still separated from each other, and to accentuate, by a large number of successive illustrations from the same embryo, the histogenesis of the lymphatic anlages in their relation to the decadent embryonic venules which they eventually entirely replace. The conception involved by the theory of extraintimal systemic lymphatic development has proved apparently, to judge from published comments, difficult to understand in some laboratories from the histo-mechanical standpoint. The following eighteen reproductions of sections of a 14 mm. embryo (series 214, slide xiii) surely explain, with- out detailed description, the principles underlying lymphatic ontogenesis in the mammalian embryo. I very gladly welcome the opportunity, which this series of Wistar publications offers, for presenting them. The plates comprised in this group are shown in order cephalo- caudad in figures 123 to 141 inclusive. In all of them the ex- PEVELOPMENT OF THE THORACIC DUCT 99 traintimal lymphatic anlages of the broncho-mediastinal trunk are indicated by the leader 5, while the included remnant of the atrophying embryonic vein is labelled 4, as in the preceding and following figures. Figs. 123, 124, 125 and 126 (sections 4, 5, 6 and 7) show well the relation of the venule (4), undergoing atrophy and replacement by the extraintimal lymphatic space (5), to the permanent venous channels of the mediastinal plexus. Thus in fig. 123 the dorsal large lighter area marked 4 is the blind cephalic end of a venule which is nearly detached from the permanent peritracheal (18} and ventral mediastinal (32) venous channels, and which in the succeeding sections is seen to be partially surrounded by the extraintimal lymphatic space (5) eventually designed to replace it entirely. In figure 124 (section 5) this peri venous space (5) has developed. The included vein (4) is still partially connected with the func- tional peritracheal plexus (18). Further ventrad is another por- tion of this same plexus. The venous kernel here still contains a number of red blood-cells. These characteristic relations of developing lymphatic and receding vein can be followed clearly throughout the succeeding sections. The contrast between the well filled permanent veins (e.g., 18, 32 in figs. 123, 124, 125 and 126), with walls fitted closely into the surrounding mesen- chyme, and the atrophying venule surrounded by the extraintimal lymphatic anlage, is striking throughout the series. Some of the sections show in certain regions good pictures of nearly complete reduction of the decadent vein to an insignificant central kernel of connective tissue surrounded by the extraintimal lymphatic space; thus, e.g., figs. 137 to 141 in the region ventro-mesad to the left vagus (22). The 15 mm. stage shows the developing lymphatic plexus of the broncho-mediastinal trunk further advanced. Figs. 142 to 145 show four successive transverse sections through the ventral part of the upper thoracic region in a 15 mm. cat embryo (series 216, slide x, sections 32, 33, 34 and 35, X 225). The central decadent venule in the interior of the developing perivenous lymphatic space has in many places disappeared entirely, leav- 100 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS ing only the clear lymphatic lumen (5). In other places remnants of the replaced vein (4) can stiU be distinctly determined. This clearance has resulted in a relative diminution of the actual area of a cross section of the meshes of the lymphatic plexus, since now the entire lumen is given over to the newly formed lymphatic channel, whereas in the earlier stages its central area was still largely filled by the degenerating veil. The lymphatic spaces in these sections have increased considerably in number. (Only a few of the more prominent vessels are indicated in the figures by leaders.) A transverse section of this region in a 15.5 mm. cat embryo (series 215, slide xiv, section 31, X 225) is shown in a fig.146. The preceding more diffuse lymphatic plexus of the earlier stages is condensed to a much smaller mesenchymal network. In a few of the larger meshes a remnant of the former central vein (4) can still be traced as a solid plug of mesoderm surrounded by the lymphatic channel (5}. In general the 15 and 15.5 mm. embryos of the cat represent the stages in the development of the broncho-mediastinal trunk in which the histological pictures of the lymph channels are least distinct and striking. This is due to the fact that on the one hand they have lost the central venous kernel around which they were originally developed through confluence of extraintimal spaces, and which gave to the combined structures their area on section and their distinctive appearance in the earlier (13.5 and 14 mm.) stages, while on the other hand the independent growth and enlargement of the lymph channels proper begins in the 16 mm. stage. In some embryos, measuring 16 mm., the appearances characteristic of the 15 mm. and 15.5 mm. stages are still encountered. Thus fig. 147 shows a transverse section through the ventral area of the upper thoracic region in a 16 mm. embryo (series 230, slide xii, section 25, X '225). The corre- spondence of the histological picture to that presented by the 15 mm. or 15.5 mm. embryo is seen by comparison with series 215 and 216 (figs. 142 to 146). The majority of 16 mm. embryos show, however, a marked advance in the independent growth of the lymphatic channels. DEVELOPMENT OF THE THORACIC DUCT 101 which have become fully established. Their lumen now appears for the most part entirely cleared of the remnants of the decadent venules originally contained therein. In only a few instances does the original vein kernel appear reduced to an insignificant rudiment. The resulting plexus of the broncho-mediastinal lym- phatic duct fills the typical position occupied in the earlier stages by the extensive preceding perivenous lymphatic reticulum. Thus figs. 148-153 give transverse sections through the ventral part of the upper thoracic region in another 16 mm. cat embryo (series 222, slide X, sections 1, 2, 3, 10, 15 and 16, X 225). This embryo affords remarkably clear pictures of the fully devel- oped lymphatic plexus of the ventral or broncho-mediastinal trunk. Figs. 148 and 153 (sections 1 and 16) give the topographical pictures of the entire region involved. The broncho-mediastinal lymphatic plexus (37) is fully developed in the typical area between the trachea (8) and vagi (22} dorsally and the pulmonary arteries (10) ventrally. The lymphatic spaces possess a clear lumen and distinct endothelial walls. The remnants of the degenerating ventral mediastinal venous plexus which they have replaced have almost entirely disappeared. Only slight traces of them are still to be seen here and there. In the main the lymphatic chan- nels are clear of venous kernels and form a rich interlacing lym- phatic plexus. The central field contains the oesophagus (8), aorta (7) and part of the medial circumference of the left praecava (6). Dorsally on the right side the right azygos vein (3) and sympathetic nerve (1) appear, while on the left side the section includes the cephalic part of the azygos segment of the left thoracic duct (36). Sections 2, 3, 10 and 15 of the same slide are shown in their ventral portions in figs. 149 to 152. They all exhibit the extent and continuity of the broncho-mediastinal lymphatic channel complex in this stage. The figures are not leadered, but the individual structures can be readily identified by reference to the topographical figures 148 and 153. The succeeding stages are characterized by a condensation of the extensive lymphatic plexus of the typical 16 mm. and 17 mm. 102 embryos into the more definite channel of the broncho-mediastinal trunk with larger caliber. Figs. 154 to 158 show the region of the broncho-mediastinal channel in four successive sections in a 20 mm. cat embryo (series 241, slide xix, sections 19, 20, 21 and 22, X 225). Fig. 154 again gives a topographical view of the entire field, while the same is cut down to the ventral portion in the remaining plates. In all the multi-meshed lymphatic plexus of the preceding stage is replaced by the more definite and larger channel of the broncho- mediastinal trunk proper (37), in the classical situation between the pulmonary arteries (10) ventrally and the trachea (9), vagi (22) and permanent ventral mediastinal veins (18) dorsally. This stage marks the practical attainment of adult conditions for the lymphatic channel of the broncho-mediastinal trunk, and from this period on its further development is merely a question of continued growth and adaptation to the environment in es- tablishing secondary connections with the tributary channels, developed independently, the preazygos segment of the thoracic duct, and (usually) the ventral prolongation of the subclavian approach of the jugular lymph sac. As will be seen from the foregoing analysis of the development of the broncho-mediastinal lymphatic trunk in embryos of the cat, the principles of mammalian systemic lymphatic ontogenesis defined in Part I of this memoir are consistently sustained by the actual evidence afforded. The development of the perive- nous extraintimal anlages which form the basis for the subse- quent evolution of the systemic lymphatic channels, is a uniform and remarkably consistent process. The existence of these spaces in this and other regions of the embryo has been so often denied by recent contributors to the subject, or, if admitted, explained in every possible way except through the correct interpretation, that I publish in this paper a series of micro-photographs of a considerable number of sec- tions taken from embryos of all the important stages. Of course these observations could be indefinitely multiplied. The larger the embryological collection, and the better the individual prep- arations are fixed and stained, the more conclusive will be the DEVELOPMENT OF THE THORACIC DUCT 103 results of a careful examination of lymphatic development along the lines here indicated. No competent and impartial observer can mistake the signifi- cance of the conditions here shown. Every stage of the process can be followed in detail. The behavior of the decadent embry- onic vein, and its relation to the enveloping extraintimal lymphatic channel, are absolutely demonstrated. The endothelium of the shrinking vein has no share in furnishing the independent lym- phatic endothelium of the replacing mesenchymal space, and nowhere, in the entire process, is there the faintest suggestion of an " out-bud" or of a " splitting off" from the circumference of an otherwise valid embryonic vein of "lymphatic" or a veno- lymphatic" anlages. It is evident in comparing a number of embryos between 13. and 16 mm., crown-rump measure, that the extensive area occu- pied by the lymphatic anlages in the earlier stages (13 mm., 13.5 mm. and 14 mm.) is due to the relatively large ske of the central decadent vein. In the earlier stages (e.g., figs. Ill, 112, 113, 114, 119, 121, 125, 126, 127, series 189, 212 and 214) the evacu- ated and abandoned vein appears as a wide loose and partially collapsed endothelial bag, surrounded by a narrow extraintimal lymphatic space, with but as yet indifferently defined endothelial lining, continuous at one or more points with the external surface of the endothelium of the degenerating vein. As development proceeds the actual perivenous lymphatic spaces increase but they concentrate and narrow down on the contained venous kernel, and, as the latter continues to decrease in size and finally become completely eliminated, the resulting clear lymphatic channel oc- cupies relatively less space than that filled in the earlier stages by the decadent venule and the early surrounding lymphatic anlage combined. Thus in the 15, 15.5 and in some 16 mm. embryos (figs. 142, 143, 144, 145, 146, and 147, series 216, 215 and 230), while the actual lumen of the lymphatic channel has increased that of the enclosed and abandoned vein has correspondingly receded to an insignificant remnant, which projects into the lymphatic space from a narrow pedicle attached to the latter's endothelial lining. 104 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS Consequently, in the pictures of the earlier stages (13, 13.5 and 14 nun. embryos) the area occupied by the still capacious aban- doned venous channel, combined with that of the relatively narrow enveloping lymphatic space, appears larger and more prominent than in the later periods (15, 15.5 mm.) in which the permanent lymphatic channel and its reduced venous kernel occupies less histological territory. It should also be noted that in the earlier stages (13, 13.5, 14 mm.) the perivenous extraintimal lymphatic network is more diffuse and redundant than in the later stages (15, 15.5, 16 mm.). This early plexiform lymphatic reticulum is subsequently con- densed into more limited and better defined lymph channels, in exact conformation to the processes observed in the haemal vascular development, in which the definite blood channels crystallize along main lines out of an antecedent indefinite and more diffuse plexus. The result of this genetic process is strik- ing in tracing developing lymphatics through successive stages. Thus in the region here under discussion the pictures offered by the 13, 13.5 and 14 mm. embryos are more conclusive than those furnished by the 15 and 15.5 mm. stages in reference to the gene- sis of the lymphatic channels. In the former, the lymphatic anlages and the decadent venules around which they develop are taken together, relatively larger and more numerous, and hence easier to recognize. In the latter, this early redundant perive- nous lymphatic plexus has been replaced by a relatively smaller, but much more definite and distinct channel, occupying, however, always a part of the area filled in the earlier stages by the less circumscribed antecedent lymphatic plexus with its contained remnants of abandoned embryonic veins. I lay special stress on this fact, because it is absolutely necessary to take it into account in comparing stages between 13.5 and 15.5 mm. In the latter stage the final lymphatic anlages have definitely formed and have reached the relation to adjacent main venous lines which they henceforth are to occupy. The further growth and enlargement of these channels begins in the 16 mm. embryo, and proceeds from this point on, in many regions at the expense of the adjacent embryonic vein undergoing atrophy. DEVELOPMENT OF THE THORACIC DUCT 105 The only conclusion which seems to me to be warranted by actual observation is that certain embryonic veins form, during the process of their atrophy and final elimination from the definite venous organization, the supporting lines along which certain of the peri venous extraintimal lymphatic anlages first develop. The fact that numerous early embryonic venous channels, large and small, atrophy and disappear during the normal course of subsequent development, appears to afford a more favorable field for the greater development of the adjacent mesenchymal in- intercellular spaces, so that these enlarge more rapidly, as the corre- lated vein recedes. This relationship appears, however, to be based exclusively on the physical and mechanical ad vantages which the abandoned and shrinking primary venous line affords to the adjacent mesenchymal spaces for more rapid enlargement, in the sense of replacing the disappearing vein and occupying second- arily the space formerly filled by the haemal channel. This is evidently an important factor in determining the size and extent of the final lymphatic channel resulting from the confluence of the originally separate and independent perivenous anlages. Consequently, in the adult, the largest and best defined systemic lymphatic vessels either accompany reduced adult remnants of a relatively larger embryonic venous channel, or, in case of the latter's entire default, topographically replace the same. Now, while this relation manifests itself strikingly in many parts of the body, it is quite evident that the development of lymphatic channels occurs in other parts independently of preceding veins, by the confluence of independent intercellular mesenchymal spaces. The cat may, and does, differ in its details of lymphatic develop- ment, and in its adult lymphatic organization, from the condi- tions obtaining in many other mammalian forms. In fact, even in the present very incomplete state of our knowledge of the com- parative anatomy of the mammalian lymphatic system, we realize that what we regard as the mammalian ground plan of lymphatic organization may in certain genera undergo great and far-reaching modifications, as for example the conditions deter- 106 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS mined by Silvester 37 for the whole group of the Cebidae, and those found by myself in Macropus rufus. And yet these differences, established and maintained within the natural limits of the mammalian class, cannot, in whatever degree they actually exist, in individual forms be basic. In any given mammalian type, the systemic lymphatic vessels, whatever their adult relation to and connection with the venous system may be, must develop in accordance with a genetic ground plan common to all mammalia. 37 C. F. Silvester. "On the Presence of Permanent Lymphatico-Venous Communications at the Level of the Renal Veins in Adult South American Monkeys." Anat. Am., Erganzungsh. z. 37. Bd., Vevh. Anat. Ges., 24 Vevs., Brussels, 1910, S. 111-114. PART II PLATES FIGURES 100 TO 158 FIGURE 100 100 Transverse section of upper thoracic region in a 12 mm. cat embryo (series 78f slide IX, section 5), X 50. 1 Sympathetic nerve. 3 Right azygos vein, thoracic portion. 6 Precardinal or precava, resp. azygos of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 16 Dorso-medial somatic venous tributaries. 22 Vagus. 31 Primitive ulnar veno-lymphatic. 33 Subclavian artery. 48 Right auricle. 49 Left auricle. 50 Right ventricle THE SYSTEMIC LYMPHATIC VESSELS PLATE at 50 100 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 101 101 Schema of venous plexuses of left side in lower cervical region of cat embryo between 13 mm. and 14 mm. crown-rump measure in transverse section. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 16 Dorso medial somatic venous tributaries. 17 Prevertebral venous plexus, 18 Feritracheal venous plexus. 19 Suprathymic venous plexus. 20 Subthymic venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 20 Common jugular vein. THK SYSTEMIC LYMPHATIC VESSELS PLATE 52 101 MEMOIR NO. 1, HUNIINQTON, 1911 FIGURE 102 102 Transverse section of lower cervical region in a 14 mm. cat embryo (series 210, slide VIII, section 41), X 225. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 18 Peritracheal venous plexus. 19 Suprathymic venous plexus. 20 Subthymic venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 24 Thyrocervical artery. 25 Internal jugular vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 53 W&S&F .-k^ Ai-*** * t 9,0* , fcyiB !i /js %W>"^ jF JSJi^&.tr JV.* _*!', .v, * : >',;^^r / 4'^-v* s_-v> r ^% E*^; ;,X? ' ,.-V *.-H iWaw -."' -rV, -''/:: ,.>*^s '. -. *-V,*ji*1^0l! 102 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 103 AND 104 103 Transverse section of ventral part of upper thoracic region in a 12 mm. cat embryo (series 217, slide VIII, section 21) , X 200. 104 Same, section 29. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 18 Peritracheal venous plexus. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 51 104 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 105 AND 106 105 Transverse section of ventral part of upper thoracic region in a 13 mm. cat embryo (series 92, slide VII, section 30), X 225. 106 Same, section 33. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerating embryonal vein. 7 Aorta. 8 Oesophagus. 9 Trachea. 22 Vagus. TIIK SYSTEMIC LYMPHATIC VESSELS PLATE 55 106 MKMOIK NO. 1, HttNTINGTON, 1911 FIGURE 107 107 Transverse section of ventral part of upper thoracic region in a 13 mm. cat embryo (series 107, slide IX, section 40) , X 225. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerating embryonal vein. 7 Aorta. 9 Trachea. 10 Pulmonary artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 56 107 MEMOIR XO. 1, HUNTINGTON, 1911 FIGURES 108 AND 109 108 Transverse section of ventral part of upper thoracic region in a 13.5 mm. cat embryo (series 223, slide X, section 11), X 225. 109 Same, section 12. 4 - Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerating embryonal vein. 7 Aorta. 9 Trachea. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 57 MEMOIR NO. 1, HONTINGTON, 1911 FIGURES 110 AND 111A 110 Transverse section through ventral part of upper thoracic region in a 13.5 mm. cat embryo (series 189, slide VIII, section 35) , X 225. 111 A Left part of section 36 of same embryo, shown in the following figure 111. X300. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal orperivenous lymphatic space surrounding degenerating embryonal vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 58 FIGURE 111 111 Transverse section through ventral part of upper thoracic region in a 13.5 mm. cat embryo (series 189, slide VIII, section 36), X 225. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerating embryonal vein. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. ' . 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS, PLATE 59 Rw9L*ftSr*?7 *il MKMOlIi XO. 1, HUNTINGTOX, 1911 1 11 FIGURES 112 AND 113 112 Same, section 37, X 225. 113 Same, section 38, X 225. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degener- ating embryonal vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 113 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 116 AND 117 116 Same, section 41, X 225. 117 Transverse section through ventral part of upper thoracic region in a 14 mm. cat embryo (series 212, slide X, section 4.) ,. X 22.5. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 32 Ventral mediastinal venous plexus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 62 io -*r 10 117 MKMOIIt NO. 1, HUNTIXGTON, 1911 FIGURES 118 AND 119 118 Same, section 5. 119 Same, section 6. Right precaval vein. Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 32 Ventral mediastinal venous plexus. Till: SYSTEMIC LYMPHATIC VESSELS PLATE 63 119 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 120 AND 121 120 Same, section 7. 121 Same, section 10. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 7 Aorta 9 Trachea. 10 Pulmonary artery. 22 Vagus. 32 Ventral mediastinal venous plexus. TIIK SYSTEMIC LYMPHATIC VESSELS //& i, v V <*J* 's&F *' , > RtftZftiy V- ^//- v '? '/; ?&/&& '?'. ' f'*flSK- \&$?^j ^'** *^ffi- w . 3 ^Xf^' - j^ir:' jfc&v&S^*> fc y^--5r-7**9^ ; 10 10 121 MKMOIH NO. 1, HUNT1NOTON, Hill FIGURES 122 AND 123 122 Same, section 11. 123 Transverse section through ventral part of upper thoracic region in a 14 mm. cat embryo (series 214, slide XIII, section 4) X 225. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 18 Peritracheal venous plexus. 22 Vagus. 32 Ventral mediastinal venous plexus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 65 *8lw^v ^ > '>, f* ** " _< f .* *^-^ri- < l 129 MEMOIR NO. 1, HUNTINGTOX, 1911 FIGURES 130 AND 131 130 Same, section 14. 131 Same, section 15. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 6 Precardinal or precava, resp. azygos of left side. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC YKSSKLS PLATE 69 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 132 AND 133 132 Same, section 16. 133 Same, section 17. 3 Right precaval vein, thoracic portion. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 70 133 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 134 AND 135 134 Same, section 18. 135 Same, section 19. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS *;^j^|i^^ '+ ~i f %." ' * *>*\ ' }'*. " p ' . ' 135 MBMOIR NO. 1, HUNTINGTON, 191! FIGURES 138 AND 139 138 Same, section 22. 139 Same, section 23. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. o Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 6 Precardinal or precava of left side. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 22 Vagus. Tlfi: SYSTEMIC LYMPHATIC VESSELS PLATE 73 ?/* *fL'J\**-WJ it. V V< /> '* f tV . i Al -cL*- v/5 / ^> .^i' V'.Wi^ ',! 10 138 r^-* 1 \ % - ^ %1-^*S t -i' ' **" s "/X 10 10 139 MEMOIR NO. 1, HUNTINGTON, 1911 140 Same, section 24. 141 Same, section 25. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 6 Precardinal or precava of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 74 141 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 142 AND 143 142 Transverse section through ventral part of upper thoracic region in a 15 mm. cat embryo (series 216, slide X, section 32), X 225. 143 Same, section 33. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 75 22 * >k/* v *v- >~ * .">*- -/^H^,/r.^#- .;-, MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 144 and 145 144 Same, section 34. 145 Same, section 35. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior ot devel- oping and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 6 Precardinal or precava of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 18 Peritracheal venous plexus. 22 Vagus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 76 10 145 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 146 AND 147 146 Transverse section through ventral part of upper thoracic region in a 15.5 mm. cat embryo (series 215, slide XIV, section 31), X 225. 147 Transverse section through ventral portion of upper thoracic region in a 16 mm. cat embryo (series 230, slide XII, section 25), X 225. 3 Right precaval vein. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 6 Precardinal or precava of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 32 Ventral mediastinal venous plexus. THE SYSTEMIC LYMPHATIC VESSELS PLATE 77 . wrJ3*7fr 'UKSr* % F> 4n F$Kri& .%d 147 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 148 148 Transverse section through ventral portion of upper thoracic; region a 16 mm. eat embryo (series 222, slide X, section 2), X 225. 1 Sympathetic nerve. 3 Right Azygos vein, thoracic portion. 6 Precardinal or precava of left side. 7 Aorta. 8 Oesophagus. 9 Trachea, 10 Pulmonary artery. 22 Vagus. 36 Thoracic duct, azygos segment. 37 Broncho-mediastinal duct. THE SYSTEMIC LYMPHATIC VESSELS PLATE 78 "'* '% u * Z+Jjam-^s' \ -- * f I -r-^"" -, . &%$ ,.-;-'".' '>.: '. - I- J **'-H k " , - _ * ' WS .- *C6^ ^ ^'O.A *- ^ . - k j i$3 3$ ,-\^.:^,S^: -' *" 148 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 149 AND 150 149 Same, section 2. 150 Same, section 3. THE SYSTEMIC LYMPHATIC VESSELS PLATE 79 I Vv- - 3e - -.?~!55 ^^ * *> - ' '^'^^ 149 v * ' ' - " y >,-- 150 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 151 AND 152 151 Same, section 10. 152 Same, section 15. THE SYSTEMIC LYMPHATIC VESSELS PLATE 80 ^milifei"; V V;$L ;: C " ' Vfy^Svw -': --'' Ji% *%<.' - : >%k"' ^'**''J$i-'s' ft 'iJ!Kt' Hff? >& i - -%C!-^^" ?-^i jR' : / ^ . is v /- ^Z^eta-^&Z' '.- ? , '. 8P *' >^i'/ .- . Clf^i^^^ s* v A *^-V < r'f ' s * r '','>. ' : 's, '', ^* ? ** y *> ^ ^fc.; ;^, m'* -*^ * -. v ' ' ** ^kk* * *> V-- --v'"-*"^ v;**^ " v "" : ^ V-/ i* . v -:^> v ' $$& r'^ : x . >- > <->" ^^^- >i ^^ 3; "'<=?55^:^ ^ *-'""" "^^ ^"' ' ^^ ^ ^r\-^r^u^ ^TQ^II t i, c, ii - -. ^. -. ^v-*. SpV- . ^ < Mr*>J' . .-^^or* < ***^ r *yf i* r *">r'-. ~* ^x s. 2t? ; .^^a^ a^i 152 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 153 153 Same, section 16. 1 Sympathetic nerve. 3 Right azygos vein, thoracic portion. 6 Precardinal or precava, resp. azygos of left side. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 36 Thoracic duct, azygos segment. 37 Broncho-mediastinal duct! THE SYSTEMIC LYMPHATIC VESSELS PLATE 81 153 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 154 154 Transverse section through ventral portion of upper thoracic region in a 20 mm. cat embryo (series 241, slide, XIX, section 18), X 225. 3 Right precaval vein. 7 Aorta. 8 Oesophagus. 9 Trachea. 10 Pulmonary artery. 18 Peritracheal venous plexus. 22 Vagus. 37 Broncho-mediastinal duct. THE SYSTEMIC LYMPHATIC VESSELS 10 154 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 155 AND 156 155 Same, section 19. 156 Same, section 20. 3 Right precaval vein. 7 Aorta. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 37 Broncho-mediastinal duct. THE SYSTEMIC LYMPHATIC VESSELS PLATE 83 37 7 22 I* ^ m? r + ^ ' Y * 'vVv-HV-T &?fil: <:-v : /;--: : :^ -T^' s^|^?ds ^ fVr>i: . i^ilaj % 10 155 9 156 MBMOIR NO. 1, HUXTIN'GTON, 1911 FIGURES 157 AND 158 157 Same, section 21. 158 Same, section 22. 9 Trachea. 10 Pulmonary artery. 22 Vagus. 37 Broncho-mediastinal duct. THE SYSTEMIC LYMPHATIC VESSELS PLATE 84 157 10 \rm W$? 22 %&iV'^||;i2L X , * 37 j%* 4^r v. la > f^l '*^&\* A^ fiL^* ^ t %* *i ' i* 1 -. V <* A '*' * *.r_ 158 MEMOIR NO. 1, HUNTINGTON, 1911 DEVELOPMENT OF THE THORACIC DUCT 107 B. The development of the proximal portion of the thoracic duct proper, between the termination of thoracic duct approach of the jugular lymph sac and the beginning of the azygos segment of the thoracic ducts, caudal to the leuel of the aortic arch, viz. the preazy- gos segment of the thoracic duct. In the earlier purely venous stages a venous plexus between oesophagus and vertebral column (17, fig. 101), drains caudo- laterad into the mesal surface of the main jugular and innomi- nate trunks. This plexus continues the supracardinal venous line cephalad beyond the level of the azygos-Cuvierian junc- tion. The terminals of this plexus are frequently joined by dorsal somatic venous tributaries (16, fig. 101), near their entrance into the main vein. These vessels have been previously discussed in their relation to the general tributary system of the jugulo- innominate line (cf. supra p. (92), and their disposition is sche- matically shown in fig. 101. In a 14 mm. embryo (series 210, slide viii, section 41, X 225) the dorso-medial or pre vertebral tributary (17) and the dorsal somatic branch (16) are shown in fig. 102, just prior to their union into the common large trunk emptying into the dorso- medial angle of the internal jugular vein (25), dorsal to the sym- pathetic nerve (1). The same plexus elements (16 and 17) are seen in another 14 mm. embryo (series 212, slide vii, section 23, X 150) in fig. 159. In this section their point of confluence and the entrance of the resulting combined trunk into the internal jugular vein takes place ventral to the sympathetic nerve (1), between the same and the left subclavian artery (33), by utilizing the pathway afforded by the peritracheal plexus (18, in fig. 102). Figs. 160 and 161 show these tributaries (16 and 17) in a 13.5 mm. cat embryo (series 189, slide vii, sections 20 and 21) in a magnification of 225 diameters, as they are approaching their confluence and point of entrance into the internal jugular vein (25). In addition these sections afford admirable examples of two other conditions, previously noted (vide supra pp. 49) : a. The development of independent mesenchymal intercellular lymphatic spaces, with well defined endothelial limiting walls. 108 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS Such spaces (77) are seen in the two sections lateral to the oesophagus. They apparently develop from the beginning as intercellular clefts, without relation to embryonic veins. The preazygos region presents these spaces in almost all 13 and 14 mm. embryos. Their development in the mammalian embryo is of the utmost phylogenetic significance and affords important evi- dence in the interpretation of systemic lymphatic ontogenesis in the sauropsida, and especially in reptiles. The investigation of avian and reptilian lymphatic development has been carried on for the past year in the anatomical laboratory of Columbia Uni- versity. These researches are now nearly completed, and some of the main results were presented at the 27th session of the American Association of Anatomists at Cornell University in December of last year. They establish a common genetic ground plan for the development and adult organization of the amniote lymphatic system. The avian and reptilian type only differs from that encountered in the mammal in respect to the higher degree of development of the jugular lymph sac, as an anterior or cervical veno-lymphatic heart, in the ontogenetic appearance of other areas of equal phylogenetic significance, as remnants of the multiple ancestral series of segmental veno-lymphatic hearts, and in the preponderance of peripheral lymphatic development by confluence of independent intercellular mesodermal spaces, not associated topographically, as in the mammal, with tem- porary embryonic venous channels. In other words, the extra- intimal peri venous development of the majority of peripheral lymphatic channels described in this communication appears as a caenogenetic process in the mammalian embryo, whereas in the sauropsid amniotes most of the systemic lymphatic channels develop along more primitive phylogenetic lines, by the direct confluence of numerous intercellular mesodermal spaces. These relative conditions have been outlined in a previous publication on the phylogenetic relations of the lymphatic and blood- vascular systems in vertebrates. 38 The results of the more 38 Geo. S. Huntington, "The phylogenetic relations of the lymphatic and blood-vascular systems in vertebrates." Anat. Record, vol. IV, no. 1, January 1910. DEVELOPMENT OF THE THORACIC DUCT 109 recent detailed investigations of lymphatic development in birds and reptiles have fully sustained the interpretation given in the publication quoted. b. The darkly stained mass of cells seen in the interval between 16 and 17 in both sections is the result of obliteration of early embryonic vascular channels, which have lost their connection with the vessels, and whose endothelial walls are apparently in the process of reverting to indifferent mesodermal cells. Many of these degenerated vascular aggregations are formed in embryos between 13 and 15 mm. crown-rump measure. Some of the elements of this early embryonic prevertebral venous plexus are secondarily replaced by perivenous or extrainti- mal lymphatic spaces in exactly the same way as is observed in the development of the ventral mediastinal duct. The resulting, originally separate, extraintimal lymphatic anlages, having re- placed the venule along and around which- they primarily devel- oped, unite with each other and form the preazygos segment of the thoracic duct, between the thoracic duct approach of the jugular lymph sac and the level of the aortic arch, at which the azygos portion of the thoracic duct begins. Fig. 162 shows a transverse section of the left side of the upper thoracic region in a 14 mm. cat embryo (series 210, slide IX, section 26, X 200). The prevertebral tributary plexus (17) turns mesad of the sympathetic nerve (1) to enter the dorso-medial angle of the left innominate vein (40). Dorsad of the main vein is the ascending portion of the left subclavian artery (33). Be- tween aorta (7) and innominate vein (40) are 'other elements of the mediastinal venous plexus, in the interval between the sympa- thetic nerve (1) and the vagus (22), with ventral branches (32) in front of the latter. Between the oesophagus and the prevertebral plexus (17) is an area in which venous elements (4), formerly connected with the same, are undergoing replacement by the extraintimal lym- phatic anlages (5) of the preazygos segment of the thoracic duct. These structures cannot be distinctly made out in fig. 162, but they are shown in a higher magnification in detail in fig. 163, which includes the region of the same section dorsal to the oesoph- 110 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS agus magnified 225 diameters. The prevertebral venous plexus (17) extends in an arch concentric with the dorsal circumference of the oesophagus (8), in front of the vertebral centers. Between it and the oesophagus, mesal to the sympathetics (1), are seen on each side former elements of the venous plexus (4) surrounded by the extraintimal lymphatic anlages (5) of the preazygos seg- ment of the thoracic duct. Figs. 164, 165 and 166 show the same structures in the next three sections (25, 24 and 23), proceeding cephalad. Fig. 166 is especially clear. In it the original line of connection of the atrophying venous kernel (4) with the prevertebral plexus (17) can still be made out as a band of condensed mesoderm, and the endothelium of the perivenous lymphatic space (5) is clearly marked. Figs. 167 and 168 show this area in two successive sections of the same embryo further caudad (series 210, slide ix, sections 43 and 44, X 150), and fig. 169 gives the last section magnified 200 diameters. The empty central venous remnant (4), with the enveloping extraintimal lymphatic anlage (5), and the still functional elements of the prevertebral plexus (17), can be seen in all of the sections. They also show, nearer to the dorso- lateral wall of the oesophagus, one of the collections of darkly stained mesodermal cells above referred to as representing rem- nants of formerly patent blood-vessels. The series of microphotographs shown in figs. 160 to 169 inclusive again conclusively demonstrates the origin of this portion of the lymphatic channel by confluence of independent extraintimal mesenchymal spaces surrounding and replacing degenerating embryonic venules. The reproductions only give relatively faint and indistinct views of the actual conditions ob- served under the microscope. The general mesenchymal tissue in this region is extremely loose and the area is very difficult to photograph sharply and distinctly. The actual sections, however, offer histological pictures identical with those observed as above described in the development of the ventral mediastinal lymphatic duct. The extraintimal lymphatic spaces are either closely applied to part of the circumference of the wall of the DEVELOPMENT OF THE THORACIC DUCT 111 shrinking venule, or they occasionally completely surround the same. In the former case the contained venous remnant is connected at part of its periphery with the wall of the envel- oping lymphatic space. In the latter and more exceptional case, the free ventral nucleus of the lymphatic space is formed by the atrophied vein. In either case the lumen of the latter may con- tain a few degenerated red blood cells. These pictures are again constant in embryos of the appropriate stages. The lymphatic anlage can be accurately traced from its indefinite beginning among the perivenous mesenchymal intercellular clefts through a number of successive sections to its similar distal terminations in the same intercellular plexus. Following the sections from this point caudad through a varying intervening area in which no distinct lymphatic channel appears, the same line will sooner or later reveal the repetition of the same process, and the formation of another link in the still disjointed chain of primitive lymphatic anlages. It is again to be noted here that in the earlier stages (13, 13.5, 14 mm.), the individual spaces, plus their kernel of atrophied vein, are relatively larger than in the later (15-15.5 mm.) em- bryos. In the earlier stages (13 mm., 13.5 mm. and 14 mm. embryos) these separate and still isolated segments of the future continuous lymphatic channel begin and end blindly. In the 15 and 15.5 mm. stages they have increased in number and ad- jacent links have united to form longer segments. Finally, in the 16 mm. cat embryo, the preazygos portion of the thoracic duct is usually established as a nearly continuous channel through the further confluence of the originally separate and distinct components. Every stage of the extraintimal lymphatic develop- ment can be observed with absolute certainty and accuracy in every series within the proper length limits. PART II, PLATES FIGURES 159 TO 169 FIGURE 159 159 Transection of upper thoracic region in a 14 mm. cat embryo (series 212, slide VII, section 23), X 150. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 11 Jugular lymph sac. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 20 Subthymic venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 25 Internal jugular vein. 33 Left subclavian artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 85 16 20- 159 MKMOIR NC). 1, HUNTINOTON, 1911 FIGURES 160 AND 161 160 Transverse section of dorsal portion of upper thoracic region in a 13.5 cat embryo (series 189, slide VII, section 20), X 225. 161 Same, section 21. 1 Sympathetic nerve. 16 Dorso-medial somatic venous tributaries. 17 Prevertebral venous plexus. 22. Vagus. 25 Internal jugular vein. 77 Mesenchymal intercellular lymphatic anlages. THE SYSTEMIC LYMPHATIC VESSELS PLATE 17 W^^&^KI m^r&^m^r 22 161 MEMOIR NO. 1, HUNTINGTON, 1911 162 Transverse section of upper thoracic region in a 14 mm. cat embryo (series 210, slide IX, section 26), X 200. 163 Dorsal portion of same section, X 225. 1 Sympathetic nerve. 4 Atrophying embryonal vein, forming kernel in interior of develop- ing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 7 Aorta. 17 Prevertebral venous plexus. 22 Vagus. 32 Ventral mediastinal venous plexus. 33 Left subclavian artery. 40 Innominate vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE S7 32 162 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 164, 165 AND 166 164 Same, section 25. 165 Same, section 24. 166 Same, section 23. . 1 Sympathetic nerve. 4 Atrophying embryonal vein, forming kernel in interior of develop- ing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 8 Oesophagus. 17 Prevertebral venous plexus. ]7 THE SYSTEMIC LYMPHATIC VESSELS PLATE ^^^SSK^^F^TT:] l^^p^^ r;fv^;i:> .:- , ' ' * {%'"?"' ' '* ' - - -"' , < 'i - .,'^:%\vfv' * r'* ' * " * ' * -** . ,-rJA^y^. '.*,**' ' '. *-**.'-; * ". - v *'*-' \^*-' ^ra^j ^^>^^|c^ ^ ffW 1 :n > . *\ '/>> yffigg x A| .' ^ '-* tf^J&< .^."\ ^''v//"^v^SP%>'i^wv "'^ ". r; '- ; v^CSr"^" i. *v . vJT/ .\-r aP^ ^JELrjih-K^r ?* ' .*.*. 166 MKMOIH NO. 1, HUNTINGTON, 1911 FIGURES 167 AND 1H8 167 Same, section 43 X 150. 168 Same, section 44 X 150. 1 Sympathetic nerve. 4 Atrophying embryonal vein, forming kernel in interior of develop- ing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degen- erating embryonal vein. 17 Prevertebral venous plexus. THE SYSTEMIC LYMPHATIC VESSELS PLATE S**/'* ' * " '"***" ''^^y'^l^''^""^^^^ * VV'^K *' ' v v ^ 168 MEMOin NO. 1, HUNTINGTON, 1911 FIGURE 169 169 Same, section 44. X 200. 1 Sympathetic nerve. 4 Atrophying embryonal vein, forming kernel in interior of develop- ing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degener- ating embryonal vein. 8 Oesophagus. 40 Innominate vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE M> 169 MK.MOIH NO. 1, HUNTINGTON, 1911 DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS 113 C. The junction of the two independently formed lymphatic chan- nels above described, viz., the preazygos segment of the thoracic duct and the ventral mediastinal trunk, with each other and with the tho- racic duct approach of the jugular lymph sac. In the 15 and 15.5 mm. embryo of the cat the thoracic duct approach of the jugular lymph sac has extended mesad between the common jugular and innominate veins and the thyro-cervical or subclavian arteries to the interval between the main vein and the sympathetic nerve. In this region the process makes second- ary connections with the independently developed ventral medias- tinal lymphatic trunk and with the equally independently formed preaortic segment of the thoracic duct. The order of this union, as previously explained, appears to have an important influence on the subsequent topographical position of this portion of the adult thoracic duct. If the jugular sac approach first meets and unites with the ventral mediastinal lymphatic channel, then the entire duct appears to pass ventrad to the sympathetic strand, between it and the adjacent dorso-medial surface of the main vein, to turn subsequently dorsad towards the pre vertebral region on the mesal aspect of the nerve (reconstruc- tion, fig. 171, series 143, slides x and xi, figs. 172 to 183). If, on the other hand, the thoracic duct approach first joins the preazygos segment of the thoracic duct, developed by the confluence of the dorsal perivenous lymphatic anlages accom- panying and surrounding the prevertebral venous plexus cephalad of the aortic arch level, then the resulting thoracic duct passes dorso-laterad to the sympathetic nerve, after having received the ventral mediastinal trunk through the interval between this nerve and the main vein (text figs., p. (72). Fig. 170 shows in a dorsal view the reconstruction of the left vascular complex of the lower cervical and upper thoracic regions in a 15 mm. cat embryo (series 218, slides VIII, sections 25^44; ix, sections 1-48; x, sections 1-48; xi, sections 1-48; xii, sections 1-48.) The reconstruction should be followed in the direction caudo- cephalad. The caudal part of the reconstruction includes 114 DEVELOPMENT OF THE THORACIC DUCT the cephalic portion of the azygos system. The left (6") and the right (3 f ) azygos veins are seen arching caudo-cephalad to empty into the left (6) and right (5) precavae. They are joined by the broad plexiform plate of the interazygos anastomosis, whose meshes are perforated by the intersegmental arteries. The sympathetic nerve strands (1) lie between the arterial branches and the azygos trunks, in the a gle between them and their dorsal somatic tributaries. Ventro-mesad of the right precava (3) is seen the right vagus (22). The distal end of the right pul- monary artery (10) is seen below the right azygos-caval arch (3'-3), surrounded by the channel of the broncho-mediastinal trunk (37). In the azygos-caval arch, and along the right pre- cava (3) are the precaval lymphatic anlages (53} previously referred to (cf. supra pp. (30) as being the first extraintimal lymphatic spaces to develop in this region, as early as the 12 mm. stage (cf. figs. TO and 11. The cephalic portion of the azygos segment of the thoracic duct, well developed in this stage, is hidden in this view of the reconstruction by the dorsal interazygos venous plexus, on the ventral face of which the lymphatic network is situated. Only a few scattered and isolated lymphatic anlages are seen projecting cephalad of this anastomosis, and not yet joined to the preazygos segment of the thoracic duct (35). Cephalad of the area just described are the large arterial ves- sels, viz., the Ductus arteriosus Botalli (56), the aortic arch (7), the right subclavian artery (57), the innominate continuation of the carotid trunk (45), and the left subclavian artery (33), with the left vertebral (41) and thyro-cervical (24) arteries arising from it. The ventrolateral continuation of the left subclavian cannot be seen in this view of the reconstruction. Ventrad of the thoracic aorta is seen the ventral mediastinal venous plexus with the accompanying cephalic part of the broncho-mediastinal lymphatic complex (upper 37 in fig. 170). The preazygos segment of the thoracic duct (35) appears as a well developed channel on the dorso-medial aspect "of the aortic arch, ascending behind the left subclavian artery (33). A number of scattered lymphatic anlages, at the level of leader 35, offer DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS 115 the genetic opportunity of establishing, in later stages, a communi- cation between the preazygos thoracic duct (35} and the broncho- mediastinal channel (37), which would correspond to the distal connection noted above (cf. p. (85) in the adult between these two lymphatic channels (cf. figs. 94, 95, 96 and 97). The cephalic end of the preazygos segment of the thoracic duct still ends blindly in this stage in two longer prolongations, the connection with the thoracic duct approach of the jugular lymph sac not having been as yet established. The latter rides in the fork between the main jugular vein and the entrance of the left superior intercostal vein (16-46). One process of the approach (12) hooks caudad on the dorso-lateral side of this vein, between it and the thyrocervical artery (24) and ends blindly lateral to the sympathetic nerve (/). Another blunt terminal of the thoracic duct approach turns ventro-mesad, between vagus (22) and sym- pathetic nerve (1), and is exposed in the reconstruction by the removal of a segment of the latter nerve which otherwise would hide it in the dorsal view here given. A long slender process passes from this portion of the thoracic duct approach ventro-mesad between the vagus (22) and carotid (22). This process, which seems to be constantly present, probably effects in later stages a connection with the cephalic part of the ventral mediastinal lym- phatic plexus. It is evident, in considering the detached and isolated lym- phatic anlages intervening here between the thoracic duct approach and the blind cephalic end if the preazygos segment of the thoracic duct already formed (35} that in course of further development the union between the latter and the jugular sac could have been es- tablished either on the dorso-lateral or ventro-medial side of the sympathetic nerve, thus leading to one or the other of the topo- graphical conditions above discussed (cf . p. 75 and text figures) . The fifth spinal nerve (54) still penetrates the caudo-lateral part of the lymph sac. Below the same are seen the sixth, seventh and eight segmental nerves of the brachial plexus (55) . The thyro- cervical artery (24) sends its ventral branch forward between the jugular and subclavian approaches of the lymph sac (just caudal to vein 16-46), and then continues cephalad on the doral aspect of the sac. 116 DEVELOPMENT OF THE THORACIC DUCT Fig. 171 shows a reconstruction of the left vascular complex of the lower cervical and upper thoracic regions in a 15.5 mm. cat embryo (series 143, slides viii, ix, x, xi, and xii, 225) viewed from the right side and from behind, and figs. 172 to 183 give transverse sections of the left upper thoracic region in the same embryo, selected from slides xi and x. The figure of the reconstruction gives by means of side lines and roman numerals the level of the slides entering into its composition, so that the individual sections shown in the microphotographs can be referred to and oriented by the reconstruction. The reconstruction as well as the photographs of the sections should be followed from below upwards. Description of reconstruction, fig. 171. The reconstruction is carried caudad to the upper azygos region. The right azygos arch (3'} is seen entering the dorsal circumference of the right precava (3). Azygos (6') and precava (6) of the left side cannot be distinctly seen in this view of the model. Mesal to the right praecava (3) is the right vagus (22). Projecting cephalad above the azygos arches and the interazy- gos anastomosis are seen the cephalic elements of the azygos por- tion of the thoracic ducts (36), which extend thence caudad along the ventral surface of the azygos complex. The cephalic end of the azygos segment of the ducts has not yet joined the caudal extremity of the preazygos segment (35), although they have approached each other closely, and adjacent detached lymphatic elements between them foreshadow their coming union. The preazygos segment of the thoracic duct (35) is a well devel- oped plexiform channel, which receives a considerable branch from the lymphatic complex along the ventral aspect of the pre- vertebral venous plexus (.77). The latter is now fully formed and differentiated clearly from the adjacent lymphatics, draining as a chain of venous spaces caudad into the azygos arches. This central part of the preazygos duct (35) intersects with the medial aspect of the large vertebral vessels (41, 4%) an d the underlying left subclavian artery and receives just caudal to this point a lymphatic channel connecting it with the broncho-mediastinal duct. This lymphatic vessel cannot be seen distinctly in this DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS 117 view of the reconstruction, because it is largely under cover of the extensive venous plexus encrusting the medial surface of the left innominate vein on which it lies. It can, however, be clearly traced in the corresponding sections of slides xi and x shown in figs 172 to 183 (cf. infra). The preazygos segment of the thoracic duct then continues on the mesal side of the sympathetic (1} cephalad, swings laterad on the ventrolateral aspect of the nerve, between it and the common jugular vein, mesal to the thyrocervical artery (24), and unites in this situation with the thoracic duct approach of the jugular lymph sac (12). The latter extends as a well defined prolongation from the dorsal aspect of the jugular approach (13). The aortic arch (7) with the innominate trunk (45), the right subclavian origin (57) and the two carotid arteries (23, 32'), occupies the centre of the model. The left subclavian artery cannot be followed in this view, but the left vertebral artery (41) is shown. Further ventrad is the section of the left innominate vein (40). The thymus (21} is nearly buried in the extensive perithymic venous anastomosis (19, 20}, which also hides from view the most of broncho-mediastinal lymphatic plexus. The thyrocervical artery (24) perforates between the two divi- sions of the lymph sac, lateral to the jugular approach (13). The fifth spinal nerve (54) and a companion vein still penetrate the caudo-lateral part of the sac. Further caudad and laterad are the sixth and seventh segmental nerves (55). The jugular approach sends a caudal prolongation along the dorse-medial aspect of the main vein, between it and the sympathetic nerve, which parallels the thoracic duct approach, and may, together with a large de- tached lymphatic island further caudad, represent an incomplete form of double or bifid thoracic duct approach as previously described (cf. supra, p. 63, text figure.) Union of broncho-mediastinal duct with preaortic portion of thor- acic duct and junction of both with thoracic duct approach of the jugular lymph sac. The achievement of this stage is excellently illustrated in the microphotographs of this 15.5 mm. embryo (series 143) shown in figs. 172 to 183. In fig. 172 (series 143, slide xi, section 16), a section through the left side of the upper thoracic 118 DEVELOPMENT OF THE THORACIC DUCT region, the cephalic end of the preaortic segment of the thoracic duct (fig. 172, 35} is seen lateral to the oesophagus (8], between the sympathetic nerve (1) dorsad and the subclavian artery (33) and left innominate vein (40) ventrad. The cephalic end of the broncho-mediastinal channel (37) lies in the angle between sub- clavian artery and innominate vein, lateral to and at some dis- tance from the trachea (9). The two lymphatic channels are traced in the succeeding sections cephalad to their union and their continuation with the thoracic duct approach of the jugular lymph sac. In fig. 173 (series 143, slide xi, section 14) the lymphatic ducts have approached each other and a third lymphatic cross-section appears mesal to the subclavian artery in the interval between the thoracic duct and the broncho-mediastinal channel, foreshadow- ing their approaching fusion (38). In fig. 174 (series 143, slide xi, section 12) this fusion has occurred (38 in fig. 174 representing union of 35 and 37) In fig. 175 (series 143, slide xi, section 11) the connection is interrupted, and in the succeeding sections (figs. 176, 177, 178 and 179) the two components of the area of confluence (38), viz, the preaortic thoracic duct segment and the broncho-mediastinal trunk, remain separated from each other. The latter assumes more and more the typical character of the thoracic duct approach of the jugular lymph sac, as previously described (cf. p. 60), occupying the typical neuro-venous recess between sympathetic nerve and innominate vein, as shown well in figs. 178 and 179 (series 143, slide xi, section 2, and slide x, section 32). In fig. 180 (series 143, slide q, section 28) the two lymphatic components have again united (38), and they continue from this point on as a single channel following the typical course of the thoracic duct approach of the jugular lymph sac cephalad to union with that structure. We have therefore in this stage a gradual approach of the cepha- lic ends of the preaortic segment of the thoracic duct and of the ventral broncho-mediastinal trunk, a preliminary anastomosis between them (figs. 173, 174), followed further cephalad by a final confluence and union with the thoracic duct approach of the jugular lymph sac. DEVELOPMENT OF THE SYSTEMIC LYMPHATIC VESSELS 119 A comparison of the series just studied (series 143) with another 15.5 mm. embryo (series 141) shows identical conditions. Thus figs. 184 and 185 show transverse sections through the upper part of the left thoracic region of the embryo. In fig. 184 the preazy- gos segment of the thoracic duct and the broncho-mediastinal trunk (38) are approaching each other in the interval between the innominate vein (50) and the subclavian artery (33) laterad, and trachea (9) and oesophagus (8) mesad. This quite agrees with the conditions seen in series 143, in figures 172 amd 173 (sections 16 and 14 of slide xi). In fig. 185 (slide x, section 21 of series 141) the two lymphatic anlages have joined (38), repeating practi- cally conditions which series 143 shows in sections 11 and 12 of slide xi, figs. 174 and 175. The permanent functional peri- tracheal venous plexus (18) appears clearly differentiated from the lymphatic channels in both sections. The earlier stages in the development of this junctional area just studied in the 15.5 mm. stages (series 143 and 141) are found in a 13.5 mm. cat embryo, series 223. Two sections of the left upper thoracic region are shown in figs. 186 and 187 (series 223, slide viii, sections 39 and 40, X 225). The sections pass through the left innominate vein (40), the left subclavian artery (33), at the level of origin of the vertebral (41) and the costocervical arteries (43), and the innominate artery (45). Both sections show the radicles of the peritracheal mediastinal venous plexus, and the beginning replacement of certain elements of the same (4) by extraintimal perivenous lymphatic spaces (5), which form the anlage of the future junctional channel between the broncho-mediastinal lymphatic trunk and the preazygos segment of the thoracic duct. These spaces are situated dorsad to the aorta and the innominate artery (45), in the interval between the tra- chea (9) mesad, the oesophagus (8) dorso-mesad, and the inno- minate vein (40) and sympathetic nerve (1) laterad. The developing perivenous lymphatic spaces seen here in this earlier 13.5 mm. embryo correspond to the finished lymphatic channel segments of the later 15.5 mm. stages, in which the pre- azygos segment of the thoracic duct meets and joins witht he inde- pendently developed broncho-mediastinal trunk, as just seen in series 143 and 141 (figs. 172 to 185). PART II, PLATES FIGURES 170 TO 187 FIGURE 170 170 Reconstruction of left vascular complex of lower cervical and upper thoracic regions of a 15 mm. cat embryo (series 218, slide VIII, sections 25 to 44; slide IX, sections 1 to 48; slide X, sections 1 to 48; slide XI, sections 1 to 48; slide XII, sections 1 to 48, X 225), dorsal view. 1 Sympathetic nerve. 3' Right azygos vein. 3 Right precava. 6 Left precava. 6' Left azygos vein, thoracic portion. 7 Aorta. 10 Pulmonary artery. 12 Jugular lymph sac, thoracic duct approach. 16 Dorso-medial somatic venous tributaries. 22 Vagus. 24 Thyrocervical artery. 33 Left subclavian artery. 35 Thoracic duct, preazygos segment. 35' Detached lymphatic space in path of preazygos channel. 37 Broncho-mediastinal duct. 41 Vertebral artery. 42 Vertebral vein. 45 Innominate artery. 46 Left superior intercostal vein. 53 Precaval lymphatics. 54 Fifth spinal nerve. 55 Sixth, seventh and eighth spinal nerve 56 Ductus arteriosus Botalli. 57 Right subclavian artery. THE SYSTEMIC LYMPHATIC VESSELS 16 46 PLATE 91 MEMOIR NO. 1, HUNTINGTON, 1911 10 170 FIGURE 171 171 Reconstruction of left vascular complex of lower cervical and upper thoracic regions in a 15.5 mm. cat embryo, (series 143, slides VIII, IX, X, and XI), X 225 View from behind and from right side. 1 Sympathetic nerve. 3' Right azygos arch. 3 Right precava. 6 Left precava. 6' Left azygos vein, thoracic portion. 7 Aorta. 12 Jugular lymph sac, thoracic duct approach. 13 Jugular lymph sac, jugular approach. 17 Prevertebral venous plexus. 19 Suprathymic venous plexus. 20 Subthymic venous plexus. 21 Thymus. 22 Vagus. 23 Left carotid artery. 23' Right carotid artery. 24 Thyrocervical artery. 29 Jugulo-subclavian trunk. 35 Thoracic duct, preazygos segment. 36 Thoracic duct, azygos segment . 40 Innominate vein. 41 Vertebral artery. 42 Vertebral vein. 45 Innominate artery. 54 Fifth spinal nerve. 55 Sixth, seventh and eighth spinal nerve 57 Right subclavian artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 92 VIII. XII 171 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 172 AND 173 172 Transverse section of left side of upper thoracic region of a 15 mm. cat embryo, (series 143, slide XI, section 16), X 200. 173 Same, section 14. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 17 Prevertebral venous plexus. 33 Left subclavian artery. 35 Thoracic duct, preazygos segment. 37 Broncho-mediastinal duct. 38 Confluence of broncho-mediastinal duct and preazygos segment of thoracic duct. 40 Innominate vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 93 * * V- '. * * V s ( T ?<^4- v w'-fclti ' -^i ' -i>.fc<^^*^v-* ^-V-i 5 *-^ '-* 172 .-*** -*- -"'- ' 173 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 174, 175 AND 176 174 Same, section 12. 175 Same, section 11. 176 Same, section 6. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 33 Left subclavian artery. 35 Thoracic duct, preazygos segment. 37 Broncho-mediastinal duct. 38 Confluence of broncho-mediastinal duct and preazygos segment of thoracic duct. 40 Innominate vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 94 40 175 wW < r\\ ->*>* 176 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURES 177, 178 AND 179 177 Same, section 5. 178 Same, section 2. 179 Same, slide IX, section 32. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 35 Thoracic duct, preazygos segment. 37 Broncho-mediastinal duct. 38 Confluence of broncho-mediastinal duct and preazygos segment of thoracic duct. 40 Innominate vein. 41 Vertebral artery. 44 Costo-cervical vein. THE SYSTEMIC LYMPHATIC VESSELS PLATE 95 177 178 40 179 MEMOIR NO. I, HUNTINGTON 1911 FIGURES 180, 181 AND 182 180 Same, slide X, section 28. 181 Same, slide X, section 26. 182 Same, slide X, section 23. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 33 Left subclavian artery. 38 Confluence of broncho-mediastinal duct and preazygos segment of thoracic duct. 40 Innominate vein. 41 Vertebral artery. 43 Costo-cervical artery. 44 Costo-cervical vein. THE SYSTEMIC LYMPHATIC: VESSELS PLATE !)0 40 180 181 182 MEMOIR NO. 1, HUNTINGTON, 1911 FIGURE 183 183 Same, slide X, section 22. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 12 Jugular lymph sac, thoracic duct approach. 22 Vagus. 23 Left carotid artery. 33 Left subclavian artery. 35 Thoracic duct, preazygos segment. 40 Innominate vein. 41 Vertebral artery. 43 Costo-cervical artery. 45 Innominate artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 97 jfr..Vfri fojr^'j - F^v^rTTrrvTpv al*VrvJf T j ** 4' x .- 7.v. :**/! -"M lf^. * - - *, '*. i^ I '- *&i t v>/* - *y"J ** ; V\i 183 MEMOIR NO. I, HUNTINQTON, 1911 FIGURES 184 AND 185 184 Transverse section through upper part of left thoracic region in a 15.5 mm. cat embryo, (series 141, slide X, section 18), X 200. 185 Same, section 21. 1 Sympathetic nerve. 8 Oesophagus. 9 Trachea. 16 Dorso-medial somatic venous tributaries. 18 Peritracheal venous plexus. 22 Vagus. 23 Left carotid artery. 33 Left subclavian artery. 38 Confluence of broncho-mediastinal duct and preazygos segment of thoracic duct. 40 Innominate vein. TI1K SYSTKMIC LYMPHATIC V10SSKLS PLATK 98 38 18 184 ^ " " " n :&jjj$$PK: % i gp5 - ; . f^'/ : 185 MEMOIR NO. 1, HUNTTNGTON, 1911 FIGURES 186 AND 187 186 Transverse section of left upper thoracic region in a 13.5 mm. cat embryo, (series 223, slide VIII, section 39), X 225. 187 Same, section 40. 1 Sympathetic nerve. 4 Atrophying embryonal vein, forming kernel in interior of developing and replacing lymphatic space. 5 Extraintimal or perivenous lymphatic space surrounding degenerat- ing embryonal vein. 8 Oesophagus. 9 Trachea. 22 Vagus. 33 Left subclavian artery. 40 Innominate vein. 41 Vertebral artery. 43 Costo-cervical artery. 45 Innominate artery. THE SYSTEMIC LYMPHATIC VESSELS PLATE 90 *%fcjfiy i<: *F '''' *** V'A ^^m^j^^$l] ^^P^ ; r^|^ ?^mv-^^v?-?^ '&$&' ^ilii ~, f.m/^