FIGURE 8.

Schema representing the prosomeric model as actualized in Puelles and Rubenstein (2003). Note that the secondary prosencephalon (SP), comprising hypothalamus and dorsal telencephalon, was left devoid of any prosomeric subdivisions. We knew that our former three units were incorrect, but had not yet found the solution for the neuromeric connection between the hypothalamus and the telencephalon. Note that at this stage we still believed in the ascription of the SP to a prechordal region of the brain. This was part of the problem and was resolved subsequently jointly with the hypothalamus issue (see Figure 9). We also used in this schema the concepts ‘rostral’ and ‘caudal’ diencephalon, which were later suppressed to avoid confusion with columnar notions (it is preferred to wholly separate the hypothalamus concept from the diencephalon proper). The three alar diencephalic domains show tentative schematic subdivision patterns that were largely corroborated subsequently, particularly the rostrocaudally tripartite pretectum (PT; Ferran et al., 2007), and the dorsoventrally tiered thalamus structure (d, i, v; Th; Redies et al., 2000; Puelles and Martinez, 2013). The prethalamus (PTh) was recently found to be rostrocaudally tripartite (Puelles et al., 2020). The inclusion of the prethalamic eminence in the prethalamus (Em), just rostral to the thalamic habenular area (Hb), was a change introduced in this schema. The hypothalamo-amygdalar bipartite spike was then based on Fan et al. (1996) and has been confirmed recently as the ‘hypothalamo-amygdalar corridor’ in Garcia-Calero et al. (2021).