II. EXPERIMENTS ON COAGULATED ORGANISERS IN THE CHICK. The experiments to be described in this section were begun in October, 1931, but no successful experiments were performed until the end of 1932. A short note on the results obtained was published in Nature (Waddington, 1933 b). In the meantime, Bautzmann, Holtfreter, Spemann and Mangold (1932) had shown that in the Am-phibia coagulated organisers are still able to exert the inductive capacities which they possess in the living state. This work followed on the earlier experiments of Marx (1931) who obtained inductions with narcotised organisers, and Spemann (1931 a) who found that the inductive capacity is retained when the organiser tissue is sub-jected to crushing. The same type of investigation into the nature of the inducing agent in the amphibian organiser has now been carried still further by Spemann, Fischer & Wehmeier(i933), Holtfreter (19330 and b) and Waddington, Needham & Needham (1933 a and b), all of whom have prepared cell-free organising extracts. The last-named authors have been able to show that the active substance is ether-soluble, but Fischer & Wehmeier (1933) claim that the substance is glycogen.
The investigation of the nature of the chick organiser has not proceeded so rapidly. This is partly due to a definite technical difficulty which does not arise in amphibian material. When a fragment of dead material is grafted in the usual manner between the ectoderm and endoderm of the chick blastoderm, it usually becomes clothed round with mesenchyme cells of the host embryo, and may in this way be insulated from the overlying ectoderm so that it cannot manifest any in-ductive capacity which it may possess. The only grafts which have been successful in performing inductions have been made into the head region of the host, where there is comparatively little mesenchyme. But here there is the danger of the in-duced neural plate, if any, being incorporated by the host neural plate and thus becoming indistinguishable. In fact, in the only two specimens in which induced neural plates are present, these induced plates are united laterally with the host plate, but in these cases the incorporation has not been perfect enough to obscure the real nature of the induced plates.
METHODS.
Experiments on Embryonic Induction 219
Holtfreter (193 3 a) has shown that in the Amphibia non-inducing tissue becomes capable of performing an induction after it is killed, but no experiments have been made in the chick with coagulated non-inducing tissue.
The grafts were placed between the epiblast and endoderm of chick blastoderms in the primitive streak stage in the usual way. Specimens were fixed in Bouin's fluid, sectioned at 15/A and stained in haemalum.
32-45. Host. 15^ hours, SM pr. s.
Donor. 24 hours, v. early head process.
Graft. h. pr. and Hensen's node, coagulated by wet immersion in
boiling water for 10 min.
Cultivated. 26 hours.
This specimen is cited as an example of the envelopment of a coagulated graft by mesenchyme. In the figure (PI. II, fig. 8) it can be seen how the graft is enclosed within a sac, most of which is mesenchymatous, although part of the floor is derived from the endoderm.
32-331. Host. 18 hours, L pr. s.
Donor. 18 hours, L pr. s.
Graft. pr. s. no end. coagulated by being immersed as dry as
possible in boiling water, at left anterior.
Cultivated. 24 hours.
The coagulated graft lies to the left of the host embryo, and stretches from the heart region to the most anterior part of the head. Its posterior part is clothed round with mesenchymatous cells, or with yolk-sac endoderm, but in the anterior part this covering is not complete, and the upper surface of the graft is not separated from the host ectoderm by a cellular layer, but only by a space.
In the most anterior part of the head, the right side of the neural plate is folded into a tube and the left side is flat. The right and left sides are probably again more or less equal in size and there may be no induced plate present in this region; but it is difficult to be certain of the middle line between the right and left sides.
33-17. Host. 20 hours, L pr. s.
Donor. 20 hours, LM pr. s.
Graft. pr. s. coagulated by dry immersion in boiling water. (a) anterior third at left anterior.
(b) posterior two-thirds at right. Cultivated. 23 \ hours.
The graft (a) lies at the anterior edge of the area vasculosa and is completely enveloped by mesenchyme and yolk-sac endoderm. It has fallen to pieces to some extent, and no induction has occurred.
Graft (b) can be seen as a large flat plate lying under the right side of the host neural tube in the region just anterior and posterior to the heart. In the posterior part of this region, an induced neural tube is present, accompanied by a notochord. This tube is quite separate from the host neural tube, although nearly adjoining it, for ten sections (PI. II, fig. 10), but further anteriorly, the lumina of the two tubes become continuous with one another, and the two tubes unite to form one single tube. The right side of the composite tube is probably larger than the left side for a further six or seven sections, after which they seem to become equal, although it is difficult to be certain on this point, as the two notochords disappear and there is no way of determining the mid-line. From the point of junction of the two tubes to the tip of the head is twenty-two sections.
As in 32-331 the head-fold is deeper and stretches further posteriorly on the side above the graft, but this fact is probably not significant. The anterior end of the notochord appears under the induced tube at nearly the same level along the longi-tudinal axis as it does under the host tube, and there is little doubt that the regional character of the induced tube has been determined by the nearby parts of the host, and not by the graft.
DISCUSSION.
The two successful experiments 32 331 and 33-17 described above are sufficient to show that in the chick the capacity to perform an induction can be present in dead material. Owing to technical difficulties, particularly to the envelopment of the graft by mesenchyme, the number of successful experiments is very small, but the fact of induction by dead organisers can be taken as proved.
The nature of the inducing factors.
Experiments on Embryonic Induction 221
Such a substance can probably be assumed to be present in the form of a quanti-tative gradient, being most concentrated in the centre of the organising region and becoming less concentrated in the more peripheral regions. This is suggested by the fact that in the chick the inductive capacity of the primitive streak is highest in the region near Hensen's node and becomes smaller posteriorly (Waddington, 1932). Similarly in the Amphibia, Bautzmann (1926) has shown that the inductive capacity is greater near the dorsal lip of the blastopore and falls off away from that region. Holtfreter (1933 a) has obtained inductions from dead endoderm material from Triton embryos, which suggests that the organising substance is present even in this part of the egg, though probably in small concentration.
An inducing stuff, present in the form of a gradient, might be active in either of two ways. Firstly, it is possible that its action essentially depends upon its being distributed in the form of a gradient. Such a hypothesis is implied by the Axial Gradient theory, where it is apparently the gradient as such which is conceived to be the causally effective agent. Secondly, it is possible that it is the stuff as such which is active, the fact that the stuff happens to be present in the form of a gradient being secondary. These two hypotheses are not necessarily mutually exclusive.
The first of these two hypotheses, which one may call the gradient hypothesis, as opposed to the second or stuff hypothesis, can be further developed in various ways. Perhaps the most obvious and convincing development is to suggest that the gradient obtains its effect by an inherent tendency for isolated parts of it to complete themselves. This is the hypothesis usually made in connection with the field theory of regeneration, for example. But it can be shown that this theory is quite untenable as an explanation of embryonic induction. The argument is as follows. If induction is an expression of a tendency for an isolated part of a gradient system or field to complete itself, then clearly a complete gradient system or field should have no in-ductive capacity. But it can be shown that in the chick such a complete field is still able to perform an induction. If two epiblasts are taken and the endoderm removed, they can be placed together with their mesoderm faces in contact and cultivated in vitro in that position. The primitive streak fields are then quite normal and complete, but even so it is found that the primitive streak of one epiblast can induce the forma-tion of a neural plate from the ectoderm of the second epiblast (Waddington, 1932; see also discussion, Waddington & Schmidt, 1933).
It is therefore clear that induction cannot be essentially dependent on a tendency of an isolated part of a gradient system to complete itself. In fact, it seems more natural to suggest that it is the stuff as such, and not a gradient of the stuff, which performs the induction1.
The individuation field of a dead organiser.
In the process of induction, at least two aspects must be distinguished. Firstly, there is the calling forth of part of a new embryonic axis in a part of the embryo which should normally not form such a structure; and secondly there is the
deter-1
mination of the regional character of this induced axis, that is to say, the determina-tion of which segment of the embryo, usually which segment of the anterior-posterior axis, it represents. Whereas, as we have seen, the induction of a new embryonic axis is not under the control of a field system, the determination of the regional character of such axes does seem to be performed by agencies of this nature. These fields have been named "individuation fields" (Waddington & Schmidt, 1933), and one of their main characteristics is that an isolated part of such a field tends to complete itself.
It is difficult to imagine how such individuation fields could be carried other than by a definite spatial structure in the inducing tissue. It might therefore be expected that an individuation field could not be carried by a dead organiser, in which one would imagine that the spatial structures would be very considerably altered if not altogether destroyed. The two inductions described in this paper are in agreement with this expectation; the numbers are however too small for great importance to be attached to this. But we do find that in both cases the regional structure of the induced axis has been determined by the host and not by the dead graft. Thus in 32-331 the induced neural plate is united with the host plate and probably shares its regional character. The same thing is true of the induced plate in 33-17 and here there is the additional evidence of the induced notochord, which stretches just as far, but no farther towards the anterior under the induced plate as the normal noto-chord does under the host plate.
In the chick it has not yet been possible to obtain inductions by dead material acting on tissue from which all possible influence of the host individuation field has been excluded. In the Amphibia, Holtfreter (1933 a) finds that under such conditions only regionally undetermined neural plates are induced. This again suggests that the individuation fields of the organisers have been destroyed by the boiling or other treatment to which they were subjected.
What is determined by the inducing agents and what by the individuation fields?
Experiments on Embryonic Induction 223
of transverse section characteristic of some position along the anterior-posterior axis of an embryo, and that the individuation fields then determine merely which position this shall be. One might almost say that in the first alternative the individuation fields determine two dimensions, in the second only one.
It would be possible to decide between these two alternatives by means of in-ductions by organisers whose individuation fields had been destroyed acting on tissue which was entirely isolated from the host individuation field. Such inductions, under rather abnormal mechanical conditions, have been obtained in Triton by Holt-freter. He found that no mesodermal structures were induced, which is evidence for the first alternative as against the second. But he also found that the induced neural plates assumed a tubular form and the inductions obtained by the use of the special technique of Auflagerung had a fairly definite and characteristic, though abnormal, pattern of transverse section. These two latter points tend to show that the second alternative is at least to some extent true, and that an organiser induces something which has at least a tendency towards a characteristic pattern of transverse section. If this conclusion turns out to be correct, the question then arises as to whether the characteristic pattern of transverse section in the induced structures is impressed by the inducing agent upon originally indifferent induced material, or whether the pattern is not from the beginning inherent in the induced material. The a priori probabilities would seem to favour the second alternative.
SUMMARY.
1. Two cases of induction by coagulated organisers in the chick are described. The implants consisted of pieces of chick primitive streak, and previous to implanta-tion they were killed and coagulated by immersion in boiling water. After this treatment they still retained the inducing capacity which they have been previously shown to possess in the live state.
2. Grafts of dead material into the chick blastoderm usually become enveloped in mesenchyme and thus isolated from the host ectoderm.
3. It is argued that, although there may in the normal egg be a gradient of in-ducing capacity, the inin-ducing factor itself cannot be a gradient as such: and refer-ence is made to the most recent work which shows that the factor is actually a chemical substance.
4. It is pointed out that there is as yet no evidence that dead organisers can determine the regional character of the embryonic axes which they induce, as live organisers can.
