go C. M. Child 
level, 1. e., composed of the same cells, shows very clearly that the 
factor involved is correlative, not constitutional. 
Finally we have to consider the quantitative regional differences 
which appear along the axis in restitution. Harenactis affords 
a fairly good illustration of such differences. Here the rapidity 
of tentacle development and to some extent also the size of the 
tentacles decrease as the distance between the level from which 
they arise and the original oral end increases: in the extreme prox- 
imal region tentacles do not appear at all. Similarly, though less 
clearly, the ability to form a foot decreases toward the oral end 
until in the cesophageal region it does not occur at all. Similar 
differences appear in all animals in which the phenomena of resti- 
tution are not confined to appendages and peripheral parts, except 
where they are complicated and modified by various other 
factors, e. g., the intensity of the correlative stimulus, unsymmetri- 
cal organization along the axis, cephalization of the nervous system, 
etc. Discussion of these factors is not necessary here since they 
do not alter the fundamental fact that quantitative regional dif- 
ferences in restitution do occur at both poles of the axis, wnich is 
all that concerns us at present. These diffferences are very evi- 
dently expressions of the constitution of different regions. In 
Harenactis, for example, from a level just proximal to the cesopha- 
gus tentacles develop rapidly and attain a large size when this 
level forms the oral end of a piece, but a foot is rarely if ever formed 
when the same level lies at the aboral end of a piece. At a level 
near the proximal end of the body tentacles appear slowly and 
remain small, but a foot forms readily from the same level. Evi- 
dently the new correlations established after isolation of the pieces 
do not completely obliterate the previously existing differences of 
constitution. If these were the only phenomena of polarity Mor- 
gan’s hypothesis of a graded constitution would account for the 
facts. 
To sum up: The qualitative regional differences as exhibited 
in primary heteromorphoses and the quantitative regional differ- 
ences which appear in the structures of either pole at different 
levels are determined by the constitution of the material which 
existed before isolation of the pieces: the qualitative axial or polar 
