Planarians have a long history of attracting many biologists for their amazing regenerative ability, which absolutely depends on a unique population of somatic pluripotent stem cells called neoblasts that are distributed throughout the body. After amputation, neoblasts proliferate to form a head blastema at an anterior-facing wound and a tail blastema at a posterior-facing wound, and finally regenerate a whole-body anterior-posterior (A-P) pattern, even from tiny tissue fragments. More than a century ago, Thomas Hunt Morgan, one of the great early investigators of planarian regeneration, proposed that axial patterning is determined by two opposing morphogenetic gradients of formative substances, head stuff and tail stuff, along the A-P axis. However, to date few attempts have been made to assess Morgan's hypothesis. Recent molecular studies using RNA interference (RNAi) have revisited planarian regeneration and revealed key signaling pathways involved in the regulation of neoblast dynamics, including Wnt/β-catenin signaling acting as a posterior tissue determinant (which accordingly fits with the instructive role of the tail stuff), during planarian regeneration. One of our recent great advances was identification of extracellular signal-regulated kinase (ERK) signaling that acts as a cell differentiator in the planarian Dugesia japonica. Furthermore, we found that interplay between anterior ERK signaling and posterior β-catenin signaling can account for the reconstruction of a complete head-to-tail axis, in which the absence of β-catenin signaling allows neoblasts to achieve the ERK activation necessary for head regeneration. These findings suggest that ERK signaling plays a crucial role in the circuitry of the head stuff in Morgan's hypothesis.