The developmental potentiality of embryonic endoderm nuclei in X. laevis is shown to change as the tissue to which they belong becomes differentiated. This change has been demonstrated by taking nuclei from endoderm tissue in different stages of differentiation, and transplanting them into enucleated unfertilized eggs; the development of the resulting transplant-embryos indicates the developmental capacity of their nuclei.

Since the proportion of total transplantations which become late blastulae is affected by the technique, the main conclusions have been drawn from the further development of late blastula transplant-embryos. The quality of recipient eggs, which is always variable, may also affect transplant-embryo development, and conclusions have therefore been mainly derived from selected experiments in which control donor nuclei have shown that egg quality was good. Reasons have been given for believing that non-hereditary nuclear qualities, such as their size and stage in mitosis, do not affect the conclusions drawn from transplantembryo development; their development is therefore solely dependent on the specific genetic qualities of the donor nuclei used.

No decline in the developmental potentiality of endoderm nuclei was found in blastulae and gastrulae. However, the capacity of nuclei to form normal tadpoles decreased progressively from after gastrulation until the beginning of torsion in the gut of swimming tadpoles; at this late stage there was still a small proportion of undifferentiated nuclei from which normal tadpoles have been obtained. Thus nuclear differentiation affects an increasing proportion of nuclei to an increasing extent, as the endoderm becomes differentiated. Serial transplantation has shown that the changes involved in nuclear differentiation are heritable and, at least to a large extent, irreversible.

These results are compared with those of Briggs & King who transplanted nuclei from endoderm cells of R. pipiens. In both Xenopus and Rana, nuclei are undifferentiated at the late blastula stage, but after this the rate and time of onset of nuclear differentiation is very different for the endoderm of the two species. These differences show that there is no exact correlation between nuclear differentiation and tissue differentiation. It is suggested that nuclear differentiation may be concerned in the processes which take place when individual cells become differentiated into their final functional state.