Recent experiments, in which barriers were inserted between anterior and posterior tissues of the chick wing bud, resulted in deletion of structures anterior to the barrier (Summerbell, 1979). From these data it was concluded that blockage of morphogen from the polarizing zone by the barrier resulted in the observed failure of specification of anterior structures. We suggest an alternative interpretation, viz. the interruption of the apical ridge by the barrier caused the deletions. This hypothesis was tested by removal of increasing lengths of ridge. This was done beginning at either the anterior or posterior junction of the wing bud with the body wall and proceeding posteriorly or anteriorly, respectively, to each half-somite level between 16/17 and 19/20. With removal of progressively greater lengths of anterior ridge, more anterior limb elements failed to develop. These data were used to construct a map of the ridge responsible for each digit. To test our hypothesis we removed posterior sections of apical ridge, as described above. Removal of posterior ridge to a level which was expected to allow outgrowth of digits anterior to the level of removal resulted in wings without digits in the majority of cases. An exception occurred when ridge posterior to the mid-19 somite level was removed. In almost half of these cases digits 2 and 3 did develop. In most cases the retention of only a half-somite piece of ridge with all other ridge removed, also resulted in deletion of all digits. Again the exception occurred when ridge posterior to somite level mid-19 and anterior to level 18/19 was removed, leaving only that ridge between somite level 18/19 and mid-19. In many of these cases digit 3 did develop. We conclude from these data that, in the wing bud, ridge anterior to the mid-19 somite level must be connected to more posterior ridge to function.
The leg ridge does not exhibit the asymmetrical, low anterior, high posterior configuration, which appears in the wing. Because the leg ridge is symmetrically high anteriorly and posteriorly, we questioned whether or not leg would also require a continuity between anterior and posterior ridge for anterior ridge to function. It did not. When posterior ridge was removed, structures developed under remaining anterior ridge and the elements which developed were complementary to those which developed after anterior ridge removal to the same somite level. Those leg elements, which failed to develop, were truncated at the appropriate proximodistal levels as indicated by the fate map we have constructed for the leg.
The data reported here do not rule out a role for the polarizing zone in specification of anterior structures. It is apparent that posterior ridge removal in the wing results in loss of structures anterior to the removal. However, this is not true for the leg.
Zygotic expression and activity of the Drosophila Toll gene, a gene required maternally for embryonic dorsal-ventral pattern formation.