THE DEVELOPMENT OF THE INFRA-UMBILICAL PORTION OF THE ABDOMINAL WALL, WITH REMARKS ON THE AETIOLOGY OF ECTOPIA VESICAE By GEORGE M. WYBURN, M.B., CH.B., F.R.F.P.S.G. Senior Demonstrator ofAnatomy in the University ofGlasgow CONTENTS PAGE Introduction .201 Theinfra-umbilical abdominalwall 203 Embryo McIntyre I . . . . . 203 Sourceofthemesoderm ofthe bodystalk 206 Divisionofthe cloacal membrane . . . . 209 Description ofembryos from 2-4to40mm. 213 Discussion ..222 Ectopia vesicae andepispadias ...224 Summary ..227 References ..228 Explanation ofPlates I-VIII ..230 INTRODUCTION THE umbilical region of the human embryo has at all times attracted much interest as a prolific source ofcongenital defects and abnormalities, in spite of which, perhaps because ofwhich, there still exists some gaps in our knowledge regarding the method of closure and completion of the belly wall. An examination of the sections of a human embryo of 4-5 mm. in length (McIntyre II), with a favourable history, and showing the yolk sac with con- necting duct, seemed to offer a fruitful field for study. The impossibility of limiting observation to the umbilicus alone was soon apparent, and it became obvious that satisfactory results could only be procured by a more compre- hensive survey embracing the whole question of the development of the an- terior abdominal wall. The initialPabsence of a ventral wall to the embryo is found in all verte- brates with the exception of some ofthe Amphibia. In thefrogwithits small, yolked, holoblastic egggastrulationiscomplete- the blastopore forms at the equator with dorsal, lateral and ventral lips and wholly enclosestheyolkcells ofthevegetative pole. Theyolkis thus takeninto the body of the embryo and lies in the floor of the gut. IntheElasmobranchs thelargeamountofyolkimpedes celldivision sothat segmentation is meroblastic. The resulting blastoderm corresponds to the cells ofthe animal pole ofthefrog's egg. The embryo arises from the posteromedian portion ofthe blastoderm, so that only the dorsal lip oftheblastopore is intra- embryonic, while the lateral and ventral lips, formed of the extra-embryonic 202 George M. Wyburn blastoderm, growandenclosetheyolk, theresultantsacbeingoutsidethebody ofthe embryo to which it is connected by a stalk. Thus at an early stage the embryo is devoid of a ventral wall. The midgut, for a varying period, com- municates with the yolk sac by the yolk stalk; this connexion is later obliter- ated. In the Gymnophiona the anterior edge and the major portion ofthe lateral edges of the blastoderm take no part in the formation of the blastopore, and -there is not only an absence ofventral wall, but the yolk remains uncovered. The embryonic shield oftheAmniota, accordingtoJenkinson, isthe equiv- alent of the blastoderm of Gymnophiona. Blastopore formation affects only the shield, there being no true ventral lip. The edge of the extra-embryonic blastoderm grows steadily round the yolk entirely independent of blastopore formation, and finally encloses it at the vegetative pole. The yolk sac is extra- embryonic. Inthe chickthe envelopingblastoderm is continuous withthe false amnion or chorion and is gradually separated from the yolk sac by the encroachment of the extra-embryonic coelom between somatopleure and splanchnopleure. In placental mammals the egg has undergone loss of yolk and cleavage is accordingly holoblastic. The blastocyst with its surrounding trophoblast and enclosed embryonic material is possibly a modification to meet nutritional and protectiverequirements. Thetrophoblast, whichfroman earlystagesurrounds the yolk sac, is analogous with,ifnotthehomologue of, thefalseamnionofthe chick. This (the false amnion) is comparable to the extra-embryonic blasto- derm of the Anamnia which encircles the yolk. The modified form of blasto- pore formation affects only theembryonic shield. The embryo is therefore without a ventral wall and the yolk sac is extra-embryonic. The completion and closure of the anterior abdominal wall occurs at a late stage of develop- ment in placental mammals. Itis the opinion ofKeith that the trophoblast in the humanblastocyst can beregarded as theprecociously developed bellywallofthe embryo. Inview of the comparativelackofknowledgeoftheearliest stages intheformation ofthe human blastocyst, which acquires in a relatively short time its own salient characteristics, and taking into account the fundamental differences in the early ontogeny of the placental mammals which represent not a simple or common evolutionaryplanbutrather agreat diversityofmethods ofattaining similar physiological requirements, it seems a hazardous experiment to affix a phylogenetic label to any portion of the human blastocyst. In the heavily yolked eggs the inert yolk substance is responsible for the partial cleavage and limitation ofgastrulation to the more active animal pole. The placental mammals show the early differentiation ofthe tissues concerned with protection and nourishment, and gastrulation is again much modified. The absence of a true ventral wall in the early stages and the extra-em- bryonic position of the yolk sac may perhaps be regarded as a modification incurred by nutritional necessity, which modification could be explained as a Infra-umbilical Portion of the Abdominal Wall 203 concomitant of incomplete and suppressed gastrulationfollowing on the more prolonged embryological phase with its demand for "yolk" or "membranes". The anterior abdominal wall in its development falls naturally into three parts-supra-umbilical, umbilical and infra-umbilical. The present portion of the work deals with the infra-umbilical abdominal wall. THE INFRA-UMBILICAL ABDOMINAL WALL This partofthe ventral wall cannot be saidto existuntil the appearance of the genital tubercle with the subsequent development of an interval between the tubercle and lower attachment of the body stalk. In the series examined the interval first becomes apparent in an embryo of 12-5 mm. which shows a well-marked genital tubercle with a narrow dense band ofmesoderm between it and the attachment ofthe stalk. Keibel & Mall give the first appearance of the tubercle as occurring in an embryo of 13 mm. This mesoderm, caudally merging into the genital tubercle and superiorly continuous with the tissue of the stalk, would therefore seem to be the basis of the future infra-umbilical portion of the belly wall. An endeavour to trace the origin ofthis mesodermal mass through a series ofyounger embryos in the Glasgow Collection has involved a consideration of theprimitive streak, ofthe cloacal membrane, ofthemesoderm ingeneral, and has led to a study ofexisting descriptions ofearly embryos in connexion with the individual points within this question. EMBRYO McINTYRE I The first ofthe series examined was embryo McIntyre I. Observations on this embryo were published by Bryce (1924), but in his paper particular in- terest was centred on the more cranial sections of the embryonic shield. The embryo is of the presomite class, having a pronounced ventral curve at the hinder end, aprimitive streak measuring 033 mm., the majorportion ofwhich occupies theventralaspectofthe caudal flexure, andatotallengthof1*4 mm., including the primitive streak. Bryce is ofthe opinion that the McIntyre embryo is younger than embryo Gle (Graf v. Spee) and embryo Vull (1Aternod), to which it bears a close re- semblanceandisprobablyastagesucceedingthatofIngallsinwhich, however, the disc has the dimensions of 2x0 75 mm. A graphic reconstruction was made of the caudal half of the embryo from sections 109-150 (Text-fig. 1A, B),-the exact position of the sections being indicated on the drawing ofthe wax models (P1. I, fig. 1). Section 146(P1I,fig. 3)isplacedattheflexureandshowstheopeningofthe archenteric canal. Sections 150-147 are through the region ofthe caudalbend, consequently there is some difficulty in interpreting the nature of the tissues anditis impossible to discriminate between ectoderm andmesoderm, butfrom the position it is probable that the section is cut through Hensen's knot. ~~~~~~~~~~~~~~~~~~~~~antois: 204 George M. Wyburn Cranial to 146 the sections show the primitive groove and primitive streak on the ventral aspect, i.e. on the caudal portion ofthe shield which has been bent round ventrally to become continuous with the body stalk. From sections 145 to 142 (PI. I, fig. 4) the primitive streak appears as a fusion between the ectoderm and mesoderm. The endoderm dorsally is in contiguity with the mesoderm, but there is no distinct fusion ofthe layers. In the sections cranial to 142 the primitive streak on the ventral aspect has the appearance ofmesoderm budding offfrom the ectoderm, more especially from the walls of the primitive groove. The endoderm forms a distinct layer con- nected to the mesoderm by a few cellular strands. This appearance persists in the next twelve sections, the only change being a gradual deepening of the yolk-sac pocket with consequent approximation of endoderm and ectoderm, and decrease of intervening mesoderm. 145 130 1918 112 .-...--.- t1 -- ~~~~~~~~130 - Ainnio-embryoniccavity 112 Text-fig. 1. EmbryoMcIntyreI. x100. The numbers refer to the sections. A. Reconstruction ofthesurfaceofthecaudalportionoftheembryonicshield,i.e.theportionbentventrallyto form the tail fold. Inthe midlineistheprimitivegroove. Areainwhichmesoderm crosses the midline is shown in black. B. Reconstruction ofa median sagittal section (seeP1. I, fig. 1). Description in text. Ectoderm-interrupted horizontal lines; endoderm-fine stippling; densemesoderm-black; primitive streak-horizontallines. Section 130 (PI. II, fig. 5). The primitive groove is still present. The ecto- derm forms a thickened cord of cells with the endoderm lying immediately above. There is no intervening mesoderm, but mesoderm is seen extending laterally on each side, separated by an interval from the ectoderm and endo- derm. The appearance is that of a cloacal membrane, and extends only over one section. Sections 129-127 (PI. II, fig. 6) again show mesoderm between the ecto- derm and the endoderm with a gradual flattening out ofthe primitive groove. The amount ofmesoderm between the two layers gradually increases. There is Infra-umbilical Portion of the Abdominal Wall 205 no sign ofbudding offofthe mesodermfromthe ectoderm, and whilethe three layers are in contiguity at the midline this may be in some measure artificial, as both ectoderm and endoderm can be traced as distinct layers while the mesoderm stretches across from side to side between them. Section 126. There is no primitive groove. The ectoderm seems to be thick- ened in the midline, forming aplate ofcells which at first glance would appear to extend upto andestablishcontactwiththeendoderm. Onfurtherexamina- tion with an apochromatic 4 mm. lens this resemblance to a cloacal plate is found tobe duetoshrinkage ofthemesoderm oneach sideleavinganapparent space between the "plate" and the mesoderm. Similar spaces can be seen in the mesoderm morelaterallyas aresult offormalinfixation. The ectodermand endoderm can again be traced as distinct layers across the middle line. Section 125 (PI. II, fig. 7). Is similar to 126, but the resemblance to a cloacal plate is even more striking. Sections 124-120 (PI. II, fig. 8). Mesoderm is again visible separating the ectoderm and endoderm. Section 119 (PI. III, fig. 9). The amniotic space between the disc and body stalk is diamond-shaped in the section. From the lower apex a distinct cord composed of a double row of cells can be made out extending down to and establishing contact with the endoderm of the allantois which at this level is oval in section and extending up the cord. Section 118 (PI. III, fig. 10). The amniotic space is still diamond-shaped. The cord ofcells noted in the previous section is not now evident, the amnion and allantois being separated by definite mesoderm. From the upper angle of the space ashortprocess extends up to butdoes notreachthe endodermofthe yolk-sac diverticulum, a single row of mesodermal cells intervening. Sections 117-109 (PI. III, fig. 11). The amniotic space gradually diminishes and disappears, the mesoderm of the disc fusing with that of the body stalk, theresultant mass beingplaced between the endoderm oftheyolksacdorsally and that ofallantois ventrally. Bryce (1924) in the description of the embryo states that the primitive streak extends up to section 117, from which point itis replaced by the cloacal plate connecting the yolk-sac diverticulum with the ectoderm andfound up to the commencement ofthe allantoic duct proper. As before mentioned, in the original description, interest wasfocusedalmostentirelyonthecranialsections and not on the region now examined. No evidence ofa cloacal plate is tobefound in section 117, orthose cranial to that; in fact in those sections where the amniotic space is gradually closing in, the ectoderm can be distinguished only with difficulty and ultimately dis- appears entirely. Definite contact between ectoderm and endoderm is established at two points, viz. sections 130 and 119. The possibility of a cloacal plate in sections 125 and 126 has been rejected because of the alternative explanation of an artefactduringfixation. Inthisembryotherearethereforetwopointsatwhich 206 George M. Wyburn anundoubted cloacalmembranecanbesaidtoexist. Betweenthesetwopoints mesoderm is continuous on the one hand with that ofthe primitive streak and on the other with the mesodermal basis of the body stalk. Regarding these findings there are two questions which occur-(A) the source ofthe mesoderm ofthebody stalk; (B) an explanation ofthe division of the cloacal membrane. SOURCE OF MESODERM OF THE BODY STALK Itisfirstlynecessarythatthereshouldbeaclearconceptionofthestructure of the primitive streak. Bryce (1924) states that all three layers are fused over the length of the streak, but while the mesoderm can be observed budding off from the ecto- derm, the basal part of which is indistinguishable from that layer, the endo- derm, except in the most caudal section where it is not present, preserves its identity and can be traced as a distinct layer across the middle line. There is certainly no interval between the mesoderm and endoderm, but in none ofthe sections of the primitive streak did there appear to be any actual budding off of mesoderm from endoderm. Hill & Florian (1931), describing the primitive streak of the "Dobbin" embryo (0.98 mm.), state that the endoderm is clearly visible on the ventral aspect of the streak. Ingalls (1918) in his embryo of 2 mm. describes a con- siderable interval between the endoderm and the primitive streak, while Heuser (1930), in one of1-6 mm., shows the streak as a fusion ofectoderm and mesoderm, the endoderm lying quite free. Debeyre (1912), on the other hand, states ofthe region ofthe primitive streak: "Tout le long de la ligne primitive ou A son niveau on remarkera l'union intime qui existe entre les feuillets." Florian (1934), from reconstructions ofembryos WO, Op. and Fetzer, is ofthe opinion that in the primitive streak primordium there is a fusion ofendoderm and mesoderm in the midline and that this persists for some time, while Rossenbeck (1923), writing on the embryo Peh-Hochstetter, considers the primitive streak as a fusion of ectoderm and mesoderm, the endoderm being separated by an interval from the actual primitive streak. This disparity in the description ofthe character ofthe streak is probably due to variations in its form according to the degree of development, not necessarily corresponding to the size of the embryonic shield; but most ob- servers seem agreed that while thereis a fusion ofall three layers in the region ofHenson's knot, theactualstreakitselfisrepresented byafusionofectoderm and mesoderm. Grosser (1927) expresses the opinion thattheprimitive groove is due to a sinking of the ectoderm on account of the heavy emigration of mesodermal cells. The blastocyst of man and higher mammals is characterized by the pre- cocious development of mesoderm, which being in the later stages an angio- Infra-umbilical Portion of the Abdominal Wall 207 blastic tissue arises in accordance with the mode of nutrition of the embryo (Bryce). Of three of the earliest ova showing the embryonic rudiment, in that of Peters (1899) the amniotic and yolk-sac vesicles are separated by mesoderm except inthemiddlelineatthecranial end; inT.B. 1 (1908)thereisaconsider- able interval betweenthe two vesicleswhich are situated in a cavity already filled withmesenchyme-theunusualdistancebetweentheamnioticcavityand the yolk sac may betosomedegreeartificial; Streeter(1927)inadescriptionof section 4 ofthe Miller ovum-the sixth section through the rudiment-states that the amniotic and yolk-sac cavities are in a common compartment surrounded by mesenchyme, the ectoderm of the amnion being in contact with the endoderm of the yolk sac with but a narrow intervening cleft. There is, however, some doubt as to what constitutes the yolk sac in the Miller ovum. Fetzer (1910), describing an embryoof0-23 mm. inlength, statesthatthere is no primitive groove, medullary fold, or neurenteric canal. Nevertheless the yolk-sac endoderm and ectoderm of the shield are separated by a layer of mesenchymal cells passing towards the cranial end and fusing in the midline with theectoderm. AccordingtoRossenbeck this isundoubtedly theprimitive streak. Grafvon Spee, discussing Peter's ovum, is ofthe opinion that the anterior and posterior poles of the amniotic and yolk-sac vesicles grow into the solid mesenchyme separating them from the chorion which mesenchyme, acting as a wedge, ultimately completely separates the ectoderm and endoderm except for an area in the midline at the caudal end-the cloacal membrane. It is now well known that the primary mesenchyme in the human ovum and that ofthe higher primates makes its appearance long before there is any indication ofa primitive streak, thereby differing fundamentally from that in the lower mammals. Streeter (1927) is of the opinion that this mesenchyme arises either from the inner cell mass during the formation of the segmentation cavity or alter- nately is split off from the trophoblast. The close relation of the mesenchyme to the trophoblast inclines him to the latter explanation. Bryce (1924) accepts the former hypothesis and regards the differentiation oftheprimitive mesenchyme as simultaneous with that ofectoderm and endo- derm from the formative cell mass. Strahl & Beneke (1910) express the opinion that the primary mesoderm is split off at the morula stage, the ectoderm and endoderm remaining widely connected in the early stages. Later this connexion is restricted to the primitive streak region. Hubrecht (1909) finds that in Tarsius the mesoderm of the connecting stalkis derived from aproliferation ofthe ectoderm at the posteriormargin of the shield which Bryce (Quain) suggests is a precocious development of the primitive streak formed later as an extension of this area. 208 George M. Wyburn In a history ofthe primates Hill (1932) shows four stages in the genesis of the body stalk: (1) Lemur, e.g. Loris, where the mesoderm arises relatively late (blasto- cyst3mm.)andbothextra- andintra-embryonic mesoderm are derived from a primitive streak. The mesenchyme of the allantoic outgrowth which is the homologue ofthe connecting stalkis formed from the primitive streak, arising from its hind end and extending round the cloacal membrane. (2) In Tarsius the connecting stalk mesenchyme is in the first instance a proliferation of the posterior margin of the shield ectoderm, later being rein- forced by a backward extension from the primitive streak. (3) In "Hapale" embryo there is an area of proliferating mesenchyme at the caudal margin of the shield which Hill suggests corresponds to the pre- cociously developed hinder end oftheprimitive streak in Tarsiusand whichin his opinion forms the whole of the primary mesoderm, including that of the stalk. The extra-embryonic mesenchyme is differentiated some considerable time before the appearance of the primitive streak. (4) The fourth stage is the human embryo. Florian (1930) has demonstrated the existence ofa proliferating areain the posteromedial marginoftheshield ectodermintheembryoFetzerandbelieves this tobethe source ofthestalkmesenchyme and themajorportion oftherest of the primary mesenchyme in the human. In a recent publication (1933) he averstohavesatisfiedhimselfastotheexistenceofsucharegionintheembryos Peters, Werner, andvonMollendorff's Wo andOp. This accords withtheviews ofRossenbeck that inthe human the primitive streak forms relatively little of the mesoderm. It is to be noted thatin the embryos ofLoris, Tarsius and Selenka's Kleim S., the ectoderm and endoderm of the rudiment are in contact in the early stages; the duration of contact diminishes as the scale is ascended. InanswertoquestionA(p.206)itcanthereforebe statedthatinthehuman the mesenchyme of the,stalk is in all probability derived from a proliferation of the caudal margin of the shield ectoderm, and one can postulate a stage, perhaps transient, when the ectoderm and endoderm are connected in the midline over a considerable extent of the embryonic shield. As seen in the sections of McIntyre I and also demonstrated by Florian (1934) in embryo Beneke (size 0 75 mm.) and by Hill and Florian (1931) in the Dobbin embryo (0-98 mm.), the mesoderm of the primitive streak is continuous round the cloacal membrane with that of the stalk. As there is no histological difference between the primary as compared with the secondary mesoderm it is impossible from a study of the sections to saywhere theoneendsandthe otherbegins, butitisreasonabletoassumethat the mass ofmesoderm beyond the cloacal membrane and on the caudal aspect of the stalk is compounded ofprimary and secondary mesoderm. Infra-umbilical Portion of the Abdominal Wall 209 DIVISION OF THE CLOACAL MEMBRANE Until the recent work of Florian, scant attention was given by embryolo- gists to the caudal region of the human embryo. Any semblance of union between the ectoderm and endoderm in the expected region by direct appo- sition or indefinite cords of cells was designated "cloacal membrane". The approximation of endoderm and ectoderm with no intervening mesoderm in section 130 ofMcIntyre I gives the appearance ofan undoubted cloacal mem- brane. Similarlytheconnectingcordofcellsfromtheamnioticectodermtothe endoderm ofthe allantois is clearly demonstrable in section 119. Onthe other hand, a cursory examination ofsections 125 and 126 might lead to their inter- pretation as cloacal plates, whereas a more careful study shows the impossi- bility of excluding mesoderm intervention. Section 118 with the cord ofcells from the amnion extending almost to the endoderm suggests that at this point former fusion between the two layers is being broken down by a pressing in of mesoderm from each side. Most observers are agreed as to the difficulty in defining the limits of the cloacal membrane. Hill & Florian (1931), describing the Dobbin embryo, comment on this, more especially the differentiation between the membrane andtheprimitive streak. OfthisregionDebeyre(1912) says: "Asonextremite posterieure commedansl'embryon de Keibel-Frassi (1907) se trouve l'ebauche d'une membrane cloacale tries probablement au moins (bien que nous ne puissons le verifier directement sur les coupes)." Much of this difficulty, how- ever, arose from the failure to recognize the incompleteness offusion in many parts ofthe "membrane". Sternberg (1927), in an embryo of4 somites, described afusionofectoderm and hindgut endoderm, and also separate connexions between amniotic ecto- derm and the endoderm of the allantois. Since then similar conditions have been noted in other specimens-Florian (1930) in embryo Bi II (also of 4 somites); Hill and Florian (1931) in the Dobbin embryo; and West (1930), describing an embryo of 8 somites, states that the cloacal membrane is very indefinite, actual contact being established in only two sections. In the Ingalls (1918) embryo there exists what the author designates as an amnio-allantoic duct characterized by a pocketing of the amniotic cavity towards the allantois with contiguity ofectoderm and endoderm. Rossenbeck (1923), after noting the cloacal membrane as a cord ofcells from the ectoderm to the endoderm in the Peh-Hochstetter embryo, in a subsequent paragraph describes an apposition of the dorsal wall of the allantoic passage with the floor of the amniotic cavity. FromastudyofMcIntyre I, alongwiththefactsavailablefromrecentwork on this subject, more especially the researches ofFlorian, I am ofthe opinion that the primordium ofthe cloacal membrane is in the early stages relatively extensive, the caudal fusion of endoderm and ectoderm being between the amnion and the future allantois, at this stage not yet separated off from the AnatomyLXXI 14 210 George M. Wyburn Table I. Table showing extent ofcloacal membrane relative to the size ofthe embryo Lengthof embryonicdisc Embryo Presomite Cloacalmembrane W.O. 0-25mm. What von Mollendorff describes as the cloacal (V. Mollendorff, 1925) membraneis, accordingtoFlorian,theprimitive streak, and von Mollendorff's allantois is the cloacalmembrane ofFlorian "Fetzer" 026mm. Thecloacalmembraneisseparatedbyaninterval (Fetzer& Florian, 1930) fromtheprimitive streak T. F. 0-4mm. Thecloacalmembraneisa cordofcellswhichex- (Florian (1928)) tendsontoandconnectstheallantoicendoderm withtheamniotic ectoderm BiI ? Primitive streakreachesthe cloacalmembrane Bi24 0-625mm. Length0-06mm. "Hugo" Stieve (1926) 0-635mm. Length0035mm. "Beneke" 0-75mm. Cloacal membrane separated by aninterval from (Strahl& Beneke, 1910) theprimitivestreak W.A. 17 0-85mm. Length0-13mm. andextends ontotheallantois (Grosser, 1930-1) "Dobbin" 0-98mm. Length0-08mm. (Hill& Florian, 1931) Peh-Hochstetter 1-4mm. Cloacalmembraneextendswellontoallantois Rossenbeck (1923) Heuser (1930) 1-6mm. Cloacalmembraneextendswellontoallantois Ingalls (1918) 2-0mm. Length 0-12mm., andthereis an additional con- nexion between the allantoic endoderm andam- nioticectoderm Somites F. von Orts Llorca 2 Cloacal membrane in two parts-one ofwhich is (1934) allantoic H3 (Wilson) 2-3 Cloacalmembraneatthecommencementofallan- tois BiII 4 Cloacal membrane broken up-one portion in- volvingallantois Sternberg (1927) 4 Cloacalmembraneis infourportions Pfannenstiel-Kromer 5-6 Very short cloacal membrane extending some distanceinto the allantois Dandy (1910) 7 Notmentioned Payne (1925) 7 Notmentioned West (1930) 8 Length009mm. Eternod (1899) 8 Length0-05mm. Corner (1929) 10 Cloacalmembraneterminates beforereachingthe allantois Embryo B1. XI 10 Cloacal membrane terminates beforereachingthe allantois PfannenstielIII 13-14 Has trebled its length compared to Pfannenstiel- Kromer Heuser (1930) 14 Length0-18mm. Atwell(1930) 17 Length0-18mm. Extendstotheallantois Davis (1924) 20 Length0-16mm.