Differentiation in culture of pieces of the early chick blastoderm. I. The definitive primitive streak and head-process stagesкод для вставкиСкачать
DIFFERENTIATION I N CULTURE O F PIECES O F THE EARLY CHICK BLASTODERM I. THE DEFINITIVE PRIMITIVE STREAK A N D HEAD-PROCESS STAGES DOROTI-IEA RUDNICX Department of Zoulogy, Uniaersity of Rochester ELEVEN FIGURES INTBODUCTION In the course of exploration of various isolation techniques, several fairly extensive series of short-time cultures of pieces of the chick blastoderm have been prepared and studied. The present communication deals with the definitive primitive streak and head-process stages. There have been many recent experimental investigations of the organization of these stages in which isolation methods have been used. Wetxel ('24, '36) and Hoadley ( '26) have used an in situ sectioning technique. Waddington and his collaborators ( '32, etc.) have combined explantation methods with transplantation of embryonic parts. Waddington ( '32, '35) has also at various times investigated the differentiation of large pieces of the blastoderm on plasma clots, as well as the morphogenesis of the whole blastoderm under these conditions ; Waterman ( '36) has briefly reported defect expcriments using this method. Hoadlev ( '26 b), Willier and Rawles ('31), Hunt ('31, '32)' Dalton ( ' 3 5 ) ' Rawles ('36), t o mention oiily a few contributions, have tested various parts of blastoderms in the axis-forming stage by means of the chorioallantoic graft. The preliminary question raised by the various results obtained in such isolations seems to me to be the following: In what degree do the structures differentiating from pieces 351 THB ANATOMICAL XICORD, VOJA. 70, NO. 3 352 DOROTHEA RUDNICK correspond t o the normal axial layout of prospective areas as determined by vital staining, and how far are they to be ascribed to secondary reactions that may vaguely be designated as reconstitutions or inductions? Only when this question is answered is it possible to ask further ones concerning the independence or dependence of various organ fields, levels of the axis, germ layers, o r other units, and their developmental role ; or to evaluate the various experimental environments used. The present experimental problem was to test in vitro areas of the blastoderm comparable to those whose behavior on the cliorio-allantois is already known. Since, with the technique used, very small pieces of the blastoderm do not differentiate in vitro, it was decided to use only transverse strips of the pellucid area, the blastoderm being so divided into approximate fourths or fifths. In two of the series reported, the entoderm was left intact; in the other two, the upper layer (mesectoderm) was explanted separately. Pieces of this size are large enough to undergo morphogenetic changes when placed on plasma clots. S s will be seen, axial structures arising under these conditions originate from their areas of prospective significance. TECHNIQUE White Leghorn and Rhode Island Red eggs were used; 16 to 22 hours' incubation gave the range of stages used. Eggs were opened in warm sterile Ringer, the blastoderm removed from the yolk and dissected in Pannett-Compton fluid kept at 38.5", under a binocular microscope. Measurements of the length of streak, length of head process if present, o r the distance from the pit to the anterior edge of the pellucid area, were made with an ocular micrometer. Transverse marks were made at the desired levels with a glass needle ; the distance of these from the pit was then measured. If the entoderm were t o be left intact, thesc marked out pieces were simply separated one from another, the opaque area trimmed off, and the piece transferred in a pipette to the surface of a cover slip clot made of two drops of plasma and two drops CULTURES O F C H I C K RLASTODERRL I 353 of dilute embryonic extract. I n the series where it was decided to remoire the entoderm, the marked blastoderm was turned over in the dish. Almost invariably the light pressure of needle necessary to make marks on the ectodermal side had been sufficient t o sever the delicate splanchnopleure completely. The mesentodermal strips could then be removed one by one, usually quite freely except at the streak or pit, where dissection would be necessary. These strips, incidentally, were also explanted ; since their differentiation capacity wa,s very slight and was limited almost entirely to mesodermal components also found in mesectodermal cultures, they will not be discussed here. The remaining heavy mesectodermal strips were trimmed and transferred to clots. The Naximow double cover slip method was used. Cultures mere run for 2 to 6 days; if f o r more than 3 days, they were washed in Pannett-Compton fluid and fed new plasma, or transferred to entirely new clots every other day. OBSERVATIONS 1. G a p a c i t i ~ ~of s various levels. Figures 1 t o 3 show the way the blastoderm was divided in the four experimental series. In the first series, diagrammed in figure 1,lthe mescctoderm was explanted separately. The anterior cut was always made directly through the pit. The second cut was made one-fourth of the way back on the streak; this distance varied from 0.27 to 0.59 mm. posterior to the first, as indicated. The third cut was made halfway back on the streak, 0.59 to 1.00 mm. posterior to the pit. Figure 2 shows a variation used in the second and third series. I n the second, the anterior cut was made 0.14 t o 0.28 mm. anterior t o the pit, the second cut 0.14 to 0.28 mm. posterior to the pit. The third was made 0.51 to 0.94 mm. behind the pit-again halfway back on the streak. The posteriormost piece was not explanted, being entirely like piece 4 in figure 1. The entoderm was removed in this series. The third series was cut as in figure 2, the first cut 0.17 to 0.34 mm. anterior to the pit, the second 0.14 to 0.41 mm. * This series was the subject of a preliminary report ( '37). 354 DOROTHEA RUDNICK posterior to the pit, the third 0.51 to 0.97 mm. back. The entoderm was included in this series, and all four levelsA, B, C, D-explanted. It will be noticed that in the foregoing two series the node was left intact, whereas in the first it was bisected transversely. Figure 3 shows the cuts made on a small series of headprocess blastoderms. The first cut was made approximately through the anterior tip of the head process. The second cut was through the pit; in a few, it was varied to a position Figs. 1and 2 Diagrams of cuts made on definitive primitive streak blastoderms; series 1 to 3; see text. Figures t o the right represent distances from the primitive pit in millimeters. Fig. 3 Diagram of cuts made 0x1 head-process blastoderms (series 4). as far as 0.34 mii. posterior to the pit. The third was about one-fourth of the way back on the streak, the fourth halfway back. Thc entoderm was left intact in these cases, its removal being impractical. Table 1presents a summary of the structures found in the various types of piece explanted. It will be noted that the piece is designated as in figures 1 to 3, the series to which it belongs (vide supra) being indicated in the second column. Thc measured limits of the pieces can be seen by reference to the figures. CELTURES O F C H I C K BLASTODEKM. 355 I To take the levels in ordcr, in the dcfinitive primitive streak and very early head-process stage : anterior material, cut at least as f a r as 0.34mm. anterior to the pit, may form medullary tube o r plate, though not invariably. If the cut be made through the pit, the proportion of medullary differenti at'ion TABLE 1 Showing distribution of dzffereictiated structures in definztive pmmzttve streak and head process semes dangramrned in figures 1 t o 3 and desertbed in the text. Figwres under pach heading show number of cases ~n which structure wus identtfied; questtonable cases are also indicated. 'Mesenchyme' zs recwrded only when formed i n d ' e a differentiated culture-not when migrated as fibroblasts. ' Epithelzum' refers t o oolwmnar and aubogal types; not to thin membranes. .Endothelium i s recorded only when clear-cut blood tessels are found c 4 s q cl z? i) Fz E B 1 __ 9 15 A 8 1 ___ B 10 B 3 8 1 2 14 1; 15;3' 9; l? 10 1 6 ; 14 15 4 5 E z - 1 1 29 ___ 1 5 2 ; 19 1 2 3 11 6 7 7 28 5 3 5; 27 3 2 ; 1;" 14 11 1 15 1 c 3 9 2 2 3; 15 8 I ~ ~ C I " 13 ~~ 2 ~3 I 1 3 I 1 4 ~ RP C 3 26 2 4 10 21 ~ ~ 4 D 4 1 1B 2 3 4 4 4 4 ~ 7 2 ~ 14 16 7 13 6 8 4 ~~ ~ 7 8 6 9 2 4 ~ 2 ~ 1 3 ~ increases, and a few dubious cases of heart muscle occur. In all this group, the presence of entoderm appears to have no effect on the differentiation of the upper layers. The node level (B), if thc node be left intact, produces medullary material nearly always. Functional heart muscle appears here rather strongly. Chorda is very frcqucnt. An 356 DOROTHEA RUDNICH effect of the eiitoderni appears at this level: it will be noticed in the second B group (series 3 ) that chorda is recorded in practically all cases: 95 t o loo%, instead of 75 to 90% as in the mesectodermal gronp. Also, body mesenchyme occurs in 100% of the former group, and is better developed in almost every individual case. The removal of the entoderm in these pieces undoubtedly has a mechanically disruptive effect: it will be recalled that dissection was almost always necessary. The post-nodal, anterior streak level (C) shows strikingly diminishing capacity for formation of medullary plate. Only the more inclusive C groups contain any of this tissue at all. These are all of inferior grade, some dediffcrcntiating, some very dubious. All are in pieces cut 0.21 mm. or less behind the pit. The 3 group, where the anterior cut mas 0.27 mm. or more posterior to the pit, contains none at all. Chorda, as might be expected, is lacking in this group, as is sensory epithelium. The series containing cntoderm ( 3 ) shows strikingly better heart differentiation; this of course is t o bc expected in view of the fact that the lower mesodermal layer usually adheres to the entoderm; the only surprising thing is that the difference did not appear previously, at the node level. The posterior (D, 4) piece produces erythroblasts uniformly; nothing else besides a little mesenchyme and epithelium. These pieces frequently do not differentiate at all, as will be seen by comparing the number of cases in the 4 group with other series 1 groups. All these levels were explanted in equal numbers; the number of cases reported indicates the proportion of successful differentiating cultures obtained. The head-process cultures show the same picture, in slightly expanded terms. The piece anterior to the process can form medullary tube, as do process and node levels (1B). These latter have strongly developed heart masses, frequently double. Chorda, body wall epithelium, and mesenchyme are regularly formed. The posterior half of the node (2) has rather poor capacity f o r forming medullary tissue and chorda; heart is more frequent. The posterior levels (3, 4) form mesenchyme, some epithelia and erythroblasts. CULTUBES O F CHICK BLASTODERM. I 357 Thus a consistent picture is provided by the distribution of differentiated structures in these series. Medullary tube or plate is found regularly in node level strips, and for some distance anteriorly. This corresponds with the prospective significance of the anterior material ; and the relations of the structures as they are observed differentiating coiifirm the interpretation that these medullary tubes actually correspond to the prospective brain region. Posterior to the node, however, the capacity for medullary plate formation falls off rapidly; most of the prospective medullary field at streak levels is incapable even of an attempt at forming medullary plate under these experimental conditions. The material classified as ‘sensory epithelium’ will be discussed subsequently; here it may be noted that this epithelium is formed with considerable frequency in the node and anterior levels in streak stages ; relatively rarely in corresponding regions of the head-process blastoderm. The notochord arises only from the immediate vicinity of the primitive pit ; heart from lateral material at and somewhat posterior to the node level. Body wall ectoderm is found in all but posterior pieces ; mesenchyme and indeterminate epithelia all through the levels tested ; blood-forming tissue may arise almost anywhere, although most strikingly in the posterior streak region. 2. Diferentiation. of the explaiated strips. The form assumed by the explanted pieces, always established in its essentials after 24 hours in vitro, was absolutely characteristic for each level, although, of course, variations on the groundpattern mere abundant. Pieces anterior to the pit or process tip, whether of the 1, A, o r 1A type, formed very large thinwalled vesicles, simple or complex. Usually such cultures contained, in addition, small dense vesicles found subsequently to be composed of mednllary tube, i.e., forebrain (fig. 9). Strips containing all o r part of the node o r head process behaved quite differently on the clot. The median region would thicken considerably, remaining opaque, very rarely forming a vesicular structure. In this mass, medullary tissue and notochord could sometimes be distinguished in the living 358 DOROTHEA RUDNICK culture. The lateral ends of the strip would swell, becoming thin-walled vesicles which often contained pulsating masses of heart tissue. This sort of culture is illustrated in figure 4, where one lateral vesicle and a median one are shown; the other lateral end had failed t o vesiculate. Fignre 6 illustrates a section of such a culture. Pieces of the 3 or C type underwent a rather striking morphogenesis, of the sort shown in figure 5. The median portion, instead of forming medullary plate, would become a dense cord. This, on later days, might become extremely long, thin and contorted. The lateral regions formed large thin-walled vesicles, sometimes with a little heart tissue inside, but usually quite empty except for a few delicate mesenchyme strands, cell nests, tubules o r endothelial vessels (figs. 7, 8). Posterior pieces (4,D) did not form vesicles in more than half the cases. These pieces remain homogeneous ; the streak disappears; usually the explant flattens out as a dense membrane, with slight peripheral migration. When vesiculation does occur, the whole explant takes the form of a blister rather than of a complete spherical structure. The interior remains full of cells, which after a few days are seen to be erythrocytes. Posterior pieces, whether vesicles do or do not fopm, yield dense masses of erythroblasts, which acquire haemoglobin usually on the third or fourth day. The changes in form described above are well marked by the end of the first day of cultivation. Variations in form usually may be traced to the failure of one lateral portion to vesiculate : in that case, migration sets in, and the region in question becomes a membrane closely applied to the surface of the clot (this is occurring in the culture shown in fig. 4). The presence of the entoderm appears t o make no difference. The fate of these structures is somewhat diverse. Compact vesicles and masses of medullary tissue may persist through several washings or transfers, and histological differentiation take place, provided cell migration does not occur. Thin-walled vesicles usnally collapse after a day of two, forming very thin membranes on the clot. Rarely, they may shrink, become CIJLTUBES O F C I I I C K BLASTOUERM. 1 359 dense, arid persist ; differentiation of tissues has not been observd in these cases. Posterior pieces nsnally persist as dense membraiies on tlie clot, with a minimum of migration. As between strips from definitive primitive streak aiid hcadprocess stages, the differerices a r e about what rnight he ex- 4 5 Fig. 4 P h o t o g r a p h of a k i n g culture of a mesectoderma! strip o€ level 2 (tip. 1) a f t e r 24 hours explantalioii. Relow i s one lateral iTesicle; above t h e niedian region, i ~ l i i c his iiow heyinning migration; tlie other lateral portion is cut off at the top; this was not vrsieiilsterl. X 40. Fig. 5 Photograph of living ciiltiire of a mesectoderimal Rtrip of lexel 3 (fig. 1) aftcr 34 houru. The m d a n axial region has formed t h e dark pro.jection directcrl toward the lower right hand corner; attached a r e seen the l a r g e thin walled l a t e r a l vesicles, contniiimg dark mesenchyme aiid bluod cores, besides a delicate membranous complex. X 40. 1)ected. The medidlary tnbes dereloping from head-process explants are much more extensive than the others aiid are more apt to be of recognizablc form, coiitaiiiing rclgions tori-esponding to optic vesicles, hrain segments, etc. T n goiicral, only tlici 360 DOXOTHE.4 RT-DNTCK axis at and anterior to the node level iindergoes a normal morphogenesis into axial structures under the coiiditions irnposed; in head-process stages this region is of course of greater extent than in previous stages. The features of primitive morphogenesis that remain unaltered by transverse section of the blastoderm may be summarized as follows : llediaii axial regions conclense in vitro as they do in normal embryogenesis. Anterior and lateral regions (i.e., ventral and extra-embrvonic) show marlied vesiciilation ; thus the coelomic areas already have certain innate capacities. Node and pre-nodal regions contain the liead aiid heart already localizccl. As for the primitive streak, thc following points seem significant: The second quarter of the streak shows its destined capacity f o r regression by its marked anteroposterior sliortening in vitro. This level when left attached to the anterior p a r t of the hlastoderm, mill show posterior growth of the streak as a little ‘tail’ (Waddington, ’32; Wetzel, ’36) j when this region is left attached to the posterior part of the blastoFigures 6 to 11 are photomicrographs of Pections through cultures. Allen B-15; staiu, he id en ha in'^ iron hemntosylin. Fixation, Fig. 6 Rectioii through’ median and oiie lateral rrrass, series 1 culture, level 2, mesectoderm, gromi 2 days. Right, central mass with thin riicdullary plate, mesmchyme, aiid Eonie cords. Left, corlomicavrsicle coiitaiihg mass of iiicoiiipletely differentiated henrt niusele and soine erpthroblasts. X 75. Fig. 7 Section through one vesicle, I)PS niesectoderui, level 3, series 1, cultured 2 days. Note douhle structure, thick mesenrhynial wall. Insidu, epithelial tubule, spaces and blood vessel. This vesicle coiitaiiis the iiiaximuiii structure yroduccd by level 3. X 75. Fig. 8 Section through vesicle of %day culture, I)YR mesectodenn, level 3, wries 1. This i? the usual picture: double vesicle of meinbranous rpitheliuiu; a little ineaeiichpme and blood inside. x 75. Fig. 9 JIPS, series 1 , meseetodenn, level 1 ; cultured 3 days. Vesicle of hody wall rrtoderin, containing lobed medullary tnbe. S o t e variations i n thiekucss of mrrlullary wall, and ‘sensory epithelial’ character of upper lobe; also straiid of ~ ~ 1 eoniiecting 1s medullary tube with outer layer. x 75. Fig. 1 0 DPP, mesectoderm of level 4, series 1, cultured 2 tla3s. Blister of thin epithelium, eoiitaiiiiiig fihiohlasts an11 erythroblasts. X 75. Fig. 11 DPS, series 1 , inesectoderm o f level 2, eultnrrd 2 days. S o t e corving medullar!: tube, partly underlaid 1)) chorcla (right) ; somite inasses a t lower left of ehorda. Good drvrloprnent of meseiiehuiiic aiid entlothelial vessels. X 75. CULTUIES O F C H I C K RLASTODERM. I 361 derm, complete regression of the streak takes place (Waddiiigton, ’36). When the level is isolated, however, a sort of 90* deflection of growth takes place in the streak region. The condensed streak material becomes a cord connecting lateral vesicles; t o do this it stretches tremendously in a lateral direction. This is accomplished, i t seems from sections, much more by shape changes thaii by proliferation; it is not a direct resnlt of tension in the cultures, since the cord may tx-ist and ereiz double on itself 011 the second or third day of explantation. The posterior half of the streak, evidently, possesses no such morphogenetic urge ; it soon loses its identity in culture. 3. Microscopic difereuttintiow. Figures 9 and 11 show the sort of medullary material developing in the cultures during the first 2 days of explantation. They are ixsnally closed-ofi tubes, irregular in shape, with out-pocketings that may or may not be of morphological significance. In one case, such an evaginatioii was found to be in coiitact with the outer ectoderm, and the contact was marked by a definite spherical thickening. This is undoubtedly an optic vesicle with lens ; other similar formations are less sure of interpretation. The medullary tubes or plates may vary in thickness, as comparison of the illustrations shows. There is a well-marked ependymal layer ; on subsequent clays of culture, ganglionic and fibrous zones may differentiate, although not in any regular pattern. This differentiation takes place only if the cnlture is transferred regularly, and at the same tinie prevented from migrating. Usually, a disorganization process sets in after 2 to 6 days; the medullary plates are broken up apparently by two processes: migration of cells from the plate, and over-proliferation within the plate itself-sometimes clediffercntiatiiig masses are found black with mitoses. I n nnhealthy cultures, o r where the tissue mass is excessively bulky, necrosis may of course set i n ; but in the period under discussion the migration and proliferation factors are the major disrupting ones. Frequently continuous with typical medullary plates (compare fig. 9) is a high pseudostratified epithelium very like that CULTURES O F CHICK BLASTODERM. I 363 normally forming the primordia of the sense organs-nose o r ear. Some of this may actually be of such an extra-medultary origin; however, the relations of most of this epithelium, and the fact that it predominates in primitive streak stages and is relatively rare in the head-process series, makes one incline to interpret it as low-grade medullary plate in most cases. hIedu3lary and sensory epithelia, especially where tubes have formed, grow within vesicles of epithelium that resembles body wall ectoderm: low cuboidal type, with vacuolated cells. This differs from the thin membranes surrounding non-axial vesicles (compare figs. 9, 8) : in section these latter are extremely thin, with almost no cytoplasm. More (fig. 11) o r less (fig. 9) mesodermal structure may be interposed between the medullary tubes and the body mall. Figure 11 shows a rather extensive development of mesenchyme, notochord, somite masses, and blood spaces. The notochord at first develops as a solid cord, which may or may not follow the course of the medullary tube, but which is almost always contorted. I n some older cultures-after 3 days-the notochord may take on its characteristic vacuolated appearance; in others it never does this, but remains in its primitive condition. Heart develops as little masses of muscle. In its earliest form, this tissue cannot be distinguished from mesenchyme ; such masses may pulsate in vitro. Usually, a typical though minute mass of recognizable primitive myocardium is present. Interlacing myoblasts distinguish these masses ; striations may develop after 3 days’ culture. Figure 6 shows a typical general picture of a culture containing heart. To the left is the original median region of the strip, containing f r a p e n t s of medullary plate. To the right is one lateral vesicle, containing the heart mass. Another heart vesicle, not shown in this section, is present on the opposite side. Some sporadic formation of isolated erythrocytes seems to occur in certain anterior pieces ; there are cases containing only a few cells of this type, usually in endothelial channels THE ANATOYICAL BJWORD, VOL. 70, NO. 3 364 DOROTHEA GUDNICK or sacs. Posterior cultures tend to give a typical blood-island picture : dense masses of erythroblasts in endothelial vessels, although the latter may be absent. These masses are not always surrounded by an epithelial membrane, as in figure 10 ; but the position of the erythropoeitic masses suggests that their origin is mesodermal. The yellow color of haemoglobin is usually detected in the living explant during the third day of culture-a little later than its normal first appearance, but still while the erythroblasts are in their primitive form. Subsequently they may differentiate completely, and become dispersed in any space available. The entodermal layer, in the series where it was retained, has shown no true differentiation. Occasionally a thin epithelial layer, or even a closed vesicle, is found, which has most probably been derived from the entoderm; hut no structure beyond this. DISCXJS8ION The organization o€ the definitive primitive streak blastoderm takes on different aspects as different methods of analysis are applied. If the blastoderm be left on the yolk, and one or more cuts be made in it (Wetzel, '36), the isolated parts will, in general, develop according to their original intention ; the movements of material in the streak region, and some other features of morphogenesis, are suppressed ; but the main axial primordia develop in the position they held at the time the operation was performed. The same result is obtained if the whole blastoderm be explanted to a plasma clot (Waddington, '32) or if large pieces, as half or a third, be isolated on a clot (ibid., '35; Waterman, '36) : an embryo, or parts, arises from the material which would normally have formed these structures, even if the material is prevented from executing its usual movements. The posterior, or streak region of the axis is at a definite disadvantage when isolated from the node j nevertheless it may form primitive trunk parts independently. When the intact blastoderm is explanted, trunk and tail parts form very Ii-ell. The movements of in- CULTURES O F C H I C K BLASTODERX.1. I 365 vagination and regression of the streak, while not absolutely essential f o r the formation of the posterolateral medullary plate and somites, must account for the precision with which these events occur in the intact blastoderm. With the formation of the head process, there is no essential change in the behavior of the blastoderm when transected. Node and process levels form axis freely; streak levels do so with some difficulty when isolated, although streak movements are more uniform (Rndnick, '38). When the blastoderm is tested by cutting it into pieces and grafting these to the chorio-allantois, a different picture is obtained. I n the definitive streak stage, the head regionnode level and anterior-consists of overlapping fields, of definite histogenetic potency, but more extensive and less definite in form than the prospectire areas corresponding to the organs in question. The immediate vicinity of the node can form almost all axial tissues unaided. Behind the node, the prospective medullary fieId is weak in histogenetic potency (Hunt, '31, '32; Stein, '33; Dalton, ' 3 5 ; Rndnick, '38). I n head-process stages, the fields previously so concentrated in the node vicinity become spread out to a position more nearly corresponding to their final site : lateral and posterior organs, in particular, become localized in this and later stages (Rudnick, '32, ' 3 5 ; Clarke, '36; Rawles, '36). The post-nodal prospective medullary area, however, does not acquire any further independence (Hunt, '31 ; Rawles, '36). Thus this analysis gives a picture of the blastoderm as a system localized only in the most general way in the node field and anteriorly, becoming progressively more specific during the head-process and subsequent stages (Willier and Bawles, ' 3 5 ) . The present experiments, like the other culture experiments and the in situ sections, show in the definitive primitive streak stage a localized axis in the anterior portion of the pellucid area, with the heart material, at least, already bilaterally situated. As in the chorio-allantoic analysis, capacity for medullary or nerve differentiation falls off very sharply behind the node. A transverse strip of the blastoderm at the level 366 DOROTHEA RUDWICK of the second quarter of the streak, f o r example, will never form medullary plate in vitro, whereas the same material left attached t o the whole posterior portion will form medullary plate in culture (Waddington, '35) and may even, infrequently (Dalton, '35 ; Rudnick, '38), form nerve tissue when grafted. Dalton finds that the presence of the entoderm is necessary for such differentiation; the present results suggest that the entirety of the posterior piece is also of importance, at least for the formation of medullary plate. The present cultures also suffer restrictions in morphogenesis of mesodermal and entodermal derivatives ; indeed, the dissection of the blastoderm appears to entail all the developmental limitations of all other methods. This is due apparently to mechanical block of movements in and of the streak. In spite of a strong tendency to migrate and disorganize, the primitive structures formed in culture-medullary plate, heart, notochord, erythroblasts-are capable of completing their differentiation in vitro under suitable conditions. Burrows ('11) and Olivo ('27) have described the migration and differentiation of neurones from young medullary plate explants ; Grigorieff ( '31) describes the differentiation of nerve elements remaining in the original explant. The present eultures, if transferred, form fibrous areas distinct from ganglionic ones. No migration of nenroblasts has been observed: the cultures carried on to differentiation were selected because the outer layer remained intact. Olivo ('28) has likewise observed, among other things, the differentiation of pulsating areas, and finally of striated heart muscle, from the prospective heart areas of the stages used here. Murray ('32) has described the differentiation of erythrocytes from all parts of the blastoderm except the anterior quarter of the streak. The histological differentiation of notochord in cultured total embryos has of course been described by Waddington ('32). Thus potencies for differentiation of cell types and small tissue units appear quite complete even in the early stages studied here. CULTURES OF CHICK BLASTODERM. I 367 SUMMARY 1. Short-time cultures of transverse fourths or fifths of definitive primitive streak and head-process blastoderms, either with or without the mesentodermal layer, mere studied with reference to morphogenesis of early embryonic structures. 2. The node level and anterior region, in the stages studied, form medullary plate, body wall, notochord, and some mesoderm from their axial (median) portion, which undergoes relatively typical morphogenesis. Anterior and lateral nonaxial regions form coelomic vesicles. Heart is localized laterally. Removal of the mesentoderm reduces the incidence of chorda and heart. 3. The second quarter of the streak forms no axial structures in vitro, although the streak tissue behaves in a striking and unique manner. Lateral regions at this level form vesicles. Presence of the mesentoderm makes no difference in the behavior of this region. 4. The level including the posterior half of the streak, whether entoderm is present or no, forms large masses of erythroblasts and a few non-specific cell types. 5 . The structures formed may be histologically similar to those of a corresponding normal stage, or they may present deficiencies. The best cases, at least, are capable of complete histogenesis in vitro, although the tendency to dediffercntiation is strong. 6. Like the results obtained by the in situ sectioning method at the same stage, the present ones show a localized axis at the node and anterior levels; unlike the former, but like the results from chorio-allantoic grafts, they show complete lack of axial organization posterior to the node level. LLTERATURE CITED BURROWS, M. T. 1911 The growth of tissues of the chicken embryo outsine of the animal body with special reference t o the nervous system. J. EX^. Zool., v01. 10, pp. 63-53. CLARKE,L. F. 1936 Regional differences in eye-forming capacity of the early chick blastoderm a s studied in chorio-allantoic grafts. Physiol. Zool., VOI. 9, pp. 102-128. DALTON,A. J. 1935 The poteiicies of portions of young chick blastoderms as tested in chorio-allantoie grafts. J. Exp. Zool., vol. 71, pp. 17-52. 368 DOROTHEA RUDNICK GRIGORIBFF, L. M. 1931 Differenzierung des Nervengewebes ausserhalb des Organismus. 1 Mitt. Arch. Exp. Zcllf., Bd. 11, S. 483-519. HOADLEP, L. 1926a The in situ development of sectioned chick blastoderms. Arch, d e Biol., vol. 36, pp. 225-309. 1926 b Developmental potencies of parts of the early blastoderm of the chick. 1-111. J. Exp. Zool., vol. 43, pp. 151-224. HUNT,T. E. 1931 An experimental study of the indepenclent differentiation of the isolated Henseu’s node and its relation to the formation of axial a n d non-axial parts in the chick embryo. J . Exp. Zoijl., vol. 59, pp. 395-427. _____ 1932 Potencies of transverse levels of the chick blastoderm in the definitive-streak stage. Anat. Rec., vol. 55, pp. 41-70. MURRAY,P. D. F. 1932 The development in vitro of the blood of the early chick embryo. Proc. Roy. Soe. Lond., B, vol. I l l , pp. 4 9 7 5 2 1 . OLIVO, 0. M. 1927 Rfigrazione di elementi nervosi coltivati in vitro. Arch. Exp. Zcllf., Bd. 4, S. 43-63. 1928 ~ o b e r die friihzeitige Deterniinierung der Herzanlage beim Huhner-embryo und dercn histologische und physiologische Differenzieruiig in vitro. Anat. Aiiz., Bd. 66, Erg. Heft, s. 108-118. RAWLES,M. E. 1936 A study i n the localization of organ-forming areas in the chick blastoderm of the head-process stage. J. Exp. Zoiil., vol. 72, pp. 271-315. R L ~ N I C KD., 1932 Thyroid-forming potencies of the early chick blastoderm. J. EX^. ZOO]., v01. 62, pp. 287-313. 1935 Regiond restriction of potencies in the chick during embryogenesis. J. Exp. Zoo]., vol. 71, pp. 83-99. -___1937 Differentiatioil in vitro of transverse strips of the mesectoderm of the chick blastoderm. Anat. Rec., vol. 67, sup. no. 3, p. 43. 1938 Contribution to the problem of neurogenic potency in postnodal isolates from chick blastoderms. J. E X ~ZOO~. . ( I n press.) STEIN,K. F. 1933 The location and differentiation of the presumptive ectoderm of the forebrain and hypophysis as shown by chorio-allantoic grafts. Physiol. Zool., vol. 6, pp. 205-235. WADDINGTON, C. H. 1932 Expcrimentv on the development of chick and duck embryos cultivated in vitro. Phil. Tram. Roy. SOC.B, vol. 221, pp. 179-2 30. 1935 The development of isolated parts of the chick blastoderm. J. EXP.ZoOl., V O ~ .71, pp. 273-288. WATERMAN, A. J. 1936 Experiments on young chick embryos culturd in vitro. Proc. Nat. Acad. Sci., vol. 22, pp. 1-3. WETZEZ,R. 1924 Uber dem Primitivknoten des Huhnchens. Verh. phys. med. Ges. Wurzburg, Bd. 49, 8. 227-233. _ _ _ 1936 . ~ Primitivstreifen und Urkorper naeh Storungsrersuchen a m 1-2 Tage bebriiteten Huhnehen. Arch. Entwmech., Bd. 134, S. 357-465. WILLIER,B. H., AND M. E. RAWLES 1931 The relation of Hensen’s node to the differentiation capacity of whole chick blastoderms a s studied in chorio-allantoic grafts. J. Exp. ZoOI., vol. 59, pp. 429-465. 1935 Organ-forming a r e w in the early chick blastoderm. Proc. Soe. Exp. Biol. and Med., vol. 32, pp. 1293-1296.