1. An attempt was made to increase the effective ‘dose’ of inducing action by transplanting pieces of presumptive neural plate from late gastrula stages into the presumptive neural plate of younger hosts, so that they were exposed to the action of the archenteron roof a second time. No difference was seen in their mode of development, either when they were left in the second embryo or if they were explanted from it (together with the underlying mesoderm) into salt solution.

2. If a young gastrula ectoderm is submitted to the action of the archenteron roof for the period from the mid- to late-gastrula stages, and then transplanted to a new young gastrula, it develops morphologically in conformity with its host, but there is some slight indication that the histology of the tissue is characteristic of a rather more posterior level of the neural system than that in which it lies.

3. No method of increasing the dose of inductive action succeeded in producing the slightest tendency for tissue lying in the presumptive neural plate of the late gastrula to develop into mesoderm.

4. If the whole presumptive neural plate is removed from a late gastrula, and ectoderm from a young gastrula substituted for it, normal regional differentiation may occur, in spite of the fact that all parts of the grafted ectoderm will have been underlain by inducing material for the same length of time.

5. If the presumptive neural plate of a late gastrula is removed and then replaced with its anterior-posterior axis reversed, most of the embryos have a reduced head or none at all, and a tail may form over the anterior mesoderm. Thus, although it can be recognized that the main polarity of the embryo has been determined by the mesoderm, there is considerable evidence that regional determination has progressed some way during the period of invagination. The results summarized in the previous paragraph show, however, that the effects of this early period of labile regional determination are not indispensable.

6. If gastrula ectoderm is grafted directly into the archenteron roof, it nearly always develops into neural tissue, whereas if it is placed between the archenteron roof and the overlying neural plate, it often differentiates into mesoderm. This is held to demonstrate that there is no condition of the interior of the gastrula which directly converts ectoderm into mesoderm; that the induction of mesoderm requires a direct contact of the whole surface of the ectoderm with inducing mesoderm; and that, provided a certain minimum intensity of inducing action has been achieved, the decision whether the ectoderm will become mesoderm or neural plate depends on whether it lies on the surface of the embryo or beneath it.

This work received financial support from the Agricultural Research Council, for which I wish to express my gratitude.
 Agricultural Research Council Scientific Staff.