380 C. M. Child 
in large measure. As each region grows it bulges out more and 
more from the level of the body-wall, forming a small, usually 
rounded elevation on which the tentacles later appear. It seems 
probable from my point of view that with the formation of this 
slight elevation the new polarity is established both as regards 
direction and position of the two poles. This establishment of 
the new polarity does not consist in any special “formative” 
process, but is merely the necessary consequence of the localiza- 
tion of the region of growth and of the growth occurring there. 
In fact it seems to me inconceivable that some sort of polarity 
should not arise in every region of growth which is definitely 
directed either by internal or external conditions. Even if the 
constitution of all parts involved is originally the same, the term- 
inal region for example must become different from a middle region 
because the correlations cannot possibly be identical in these two 
regions and because they are differently related to the external 
world. In short, any organic system in which either the correla- 
tions between parts or the relations between the system and the 
external world differ characteristically in different regions possesses 
a polar specification in some degree. 
In Harenactis as in most elongated forms the oral region is in 
general the initiating or dominant region in reaction, i. e., it reacts 
more rapidly than other regions in the presence of many, probably 
of most stimuli. So far as this is the case this region must play 
an important role in determining the development of other cor- 
related parts, since the fact that it reacts to a given stimulus before 
they do usually or always changes the character of the reaction in 
them, according to their correlation with the dominant region. 
In the regions of localized growth on the rings the central part, 
or after elevations have been formed, the terminal portions of these 
are undoubtedly the physiologically dominant parts, and as such 
develop oral structures. These regions of growth are then briefly 
more or less completely radially symmetrical regions in which 
polarity arises in consequence of their constitution and their cor- 
relations with adjoining parts. Such a system composed of Hare- 
nactis protoplasm being given, it seems probable that structural 
localization and differentiation must occur in greater or less degree, 
