VII. LIMB MORPHOGENESIS
311
It may be argued that digit formation occurred solely as a result of
'influences' from the proximal tissues and that the precise position of
the mesoderm (regardless of contact with ridge) was the important
determining factor. This is dubious. In the same paper Saunders et al.
reported that removal of the distal two-thirds of a limb bud invariably
resulted in complete absence of distal structures. There was no 'regulation' of distal structures (Amprino and Camosso, 1955d, reported similar
results with a roughly equivalent operation, cf their Fig. le). When,
however, Saunders removed the distal two-thirds of the bud but now
retained the ectodermal ridge and attached it to the cut surface of the
stump (i.e. the remaining one-third of the limb bud), distal structures
did form. (See also results of Hampé above).
The most convincing evidence, which leaves no doubt that the
ectodermal ridge has some sort of inductive influence on the limb
mesoderm, comes from experiments in which one limb-bud mesoblast is
provided with more than one ectodermal ridge. When all of the ectoderm,
including the original ridge, is removed from a limb bud and a ridge
(from other limb buds) is placed on each lateral limb surface (Zwilling,
1956a) outgrowth from the original distal end stops. Instead there are
two new outgrowths and two sets of distal structures form, one from
each lateral surface and in direct association with the two grafted ridges.
When part of a ridge is used instead of a whole one, only partial outgrowth occurs and only one or two digits form. In fact every portion of
the mesoblast surface (including the base which is in contact with the
somatopleure and which ordinarily forms girdle) may form distal limb
structures provided that it is in contact with an ectodermal ridge.
Tschumi (1957) has reported a few cases of double outgrowths from
Xenopus limb buds on which the ectoderm had been rotated 90° from
the original position. A new ectodermal ridge formed (presumably
induced by the mesoderm) over the site of the original, while the old
ridge persisted. An outgrowth which developed digits formed in association with each ridge.
In the majority of Tschumi's 90° rotation experiments, a new ridge
did not appear. All of these developed axis relations (i.e. anteroposterior and dorso-ventral features) which were in strict accord with
the ectoderm. They were all turned 90° to the original orientation of the
mesoderm. This was a satisfying confirmation of some of my own
experiments (Zwilling, 1956b) in which the ectoderm of chick limb buds
was re-oriented so that it was 90° to its original relation with the mesoderm. Again, in these cases, all of the outgrowth was at right angles to
the original long axis of the mesoblast and the ectoderm was the determining component of the axis relations. It is significant that only those
portions of the original distal surface of the mesoblast which were
311
It may be argued that digit formation occurred solely as a result of
'influences' from the proximal tissues and that the precise position of
the mesoderm (regardless of contact with ridge) was the important
determining factor. This is dubious. In the same paper Saunders et al.
reported that removal of the distal two-thirds of a limb bud invariably
resulted in complete absence of distal structures. There was no 'regulation' of distal structures (Amprino and Camosso, 1955d, reported similar
results with a roughly equivalent operation, cf their Fig. le). When,
however, Saunders removed the distal two-thirds of the bud but now
retained the ectodermal ridge and attached it to the cut surface of the
stump (i.e. the remaining one-third of the limb bud), distal structures
did form. (See also results of Hampé above).
The most convincing evidence, which leaves no doubt that the
ectodermal ridge has some sort of inductive influence on the limb
mesoderm, comes from experiments in which one limb-bud mesoblast is
provided with more than one ectodermal ridge. When all of the ectoderm,
including the original ridge, is removed from a limb bud and a ridge
(from other limb buds) is placed on each lateral limb surface (Zwilling,
1956a) outgrowth from the original distal end stops. Instead there are
two new outgrowths and two sets of distal structures form, one from
each lateral surface and in direct association with the two grafted ridges.
When part of a ridge is used instead of a whole one, only partial outgrowth occurs and only one or two digits form. In fact every portion of
the mesoblast surface (including the base which is in contact with the
somatopleure and which ordinarily forms girdle) may form distal limb
structures provided that it is in contact with an ectodermal ridge.
Tschumi (1957) has reported a few cases of double outgrowths from
Xenopus limb buds on which the ectoderm had been rotated 90° from
the original position. A new ectodermal ridge formed (presumably
induced by the mesoderm) over the site of the original, while the old
ridge persisted. An outgrowth which developed digits formed in association with each ridge.
In the majority of Tschumi's 90° rotation experiments, a new ridge
did not appear. All of these developed axis relations (i.e. anteroposterior and dorso-ventral features) which were in strict accord with
the ectoderm. They were all turned 90° to the original orientation of the
mesoderm. This was a satisfying confirmation of some of my own
experiments (Zwilling, 1956b) in which the ectoderm of chick limb buds
was re-oriented so that it was 90° to its original relation with the mesoderm. Again, in these cases, all of the outgrowth was at right angles to
the original long axis of the mesoblast and the ectoderm was the determining component of the axis relations. It is significant that only those
portions of the original distal surface of the mesoblast which were