134 INDIVIDUALITY IN ORGANISMS 



extent than in longer pieces the establishment of a new 

 gradient in the opposite direction. In pieces more than 

 eight or ten millimeters long, however, the local condi- 

 tions at the basal end usually determine the result 

 sooner or later, and the new gradient is established and a 

 hydranth develops here. 



In pieces between eight or ten and two or three milli- 

 meters in length neither hydranth nor any other out- 

 growth arises at the basal end in most cases. In these 

 shorter pieces the dominance of the apical region is 

 sufficient to inhibit the new gradient at the basal end 

 to a sufficient degree to prevent hydranth formation, 

 and the general metabolic rate in these as in most 

 other experimental pieces is not high enough for stolon- 

 formation to occur. 



In the very short pieces described in chap, iv 

 (pp. 96-99) the difference in metabolic rate between 

 the two ends of the piece dependent upon the original 

 gradient is so slight that in many cases the local condi- 

 tions at the two ends become the determining factors, 

 and hydranths begin to form simultaneously or nearly 

 so at both ends, the portion of each hydranth formed 

 depending on the length of the piece. If the original 

 gradient in the piece is sufficient to determine the more 

 rapid reaction at the apical end this becomes dominant 

 and a single, instead of a double, structure arises. 



These are the chief facts of reconstitution in Tubu- 

 laria under ordinary 'conditions and their interpretation 

 in terms of metabolic gradients and dominance. It is 

 possible, however, to obtain more positive evidence in 

 support of these interpretations by controlling and 

 altering the experimental conditions. By diluting the 



