At the end of the nineteenth century, controversy arose as to precisely when the first glial cells originate during development of the central nervous system, and to date, the issue has not been satisfactorily resolved. His (1889, 1890) noted that, even in the earliest developmental stages of the germinallayer, there appeared to be two distinct cell types. The cells which he called Spongioblasten were thought to be glial precursors from which all mature glial cells derive; Keimzellen, in contrast, were regarded as forming 1 neurons. His was working on the assumption that the very first preneurons migrate into a preexisting framework of glial eelIs. In contrast to this view, Schaper (1897) regarded both Keimzellen and Spongioblasten as belonging to a common population of proliferating and pluripotent stem cells which begin differentiation into glial and neuronal cells at late developmental stages. It is this latter view which is the basis of the most recent studies on the subject (e. g. , Caley and Maxwell1968a, 1968b; DeVitry et al. 1980).The concept of one common stem cell seemed to be supported both by experiments using 3H-thymidine autoradiography (Fujita 1963, 1965b, 1966; Sauer and Walker 1959; Sidman et al. 1959) and by ultrastructural studies (Fu- jita 1966; Hinds and Ruffet 1971; Wechseler and Meller 1967) indicating that structural differences, which His presumably used to define his two cell types, could be related to different stages of the mitotic cycle.

The basic thesis for this study was that the telencephalon is needed to make decisions in new situations. Subsidary hypotheses were that the telencephalon consists of: (a) a sensorimotor system which generates motor activity from sensory input and (b) a selection system which makes choices from possible motor programs. It was postulated that the selection system should fulfil the following requirements: be accessible for past and present events, have the capacity to process this information in a nondetermined way with a possibility for ordering, and have access to motor-affecting systems (the sensorimotor system). The ability of the selection system to correlate information in a nonpredetermined way was considered most important. In short: The selection system should be able to associate any information in any combination, and have the capability for internal control of neuronal activity and external selection of motor programs (see Fig. IA. ) Xenopus laevis was chosen as a subject, since it has a relatively simple tel­ encephalon, with characteristics that it shares with "primitive" species of different vertebrate classes, and because it is easy to maintain as a laboratory animal. The main method used was the determination of connections with HRP. The pallium was in the focus of attention, since it was considered to be the core of the selection system. Immunohistochemistry was used as an additional parameter to compare Xenopus laevis forebrain with those of other vertebrates.

At the end of the nineteenth century, controversy arose as to precisely when the first glial cells originate during development of the central nervous system, and to date, the issue has not been satisfactorily resolved. His (1889, 1890) noted that, even in the earliest developmental stages of the germinallayer, there appeared to be two distinct cell types. The cells which he called Spongioblasten were thought to be glial precursors from which all mature glial cells derive; Keimzellen, in contrast, were regarded as forming 1 neurons. His was working on the assumption that the very first preneurons migrate into a preexisting framework of glial eelIs. In contrast to this view, Schaper (1897) regarded both Keimzellen and Spongioblasten as belonging to a common population of proliferating and pluripotent stem cells which begin differentiation into glial and neuronal cells at late developmental stages. It is this latter view which is the basis of the most recent studies on the subject (e. g. , Caley and Maxwell1968a, 1968b; DeVitry et al. 1980).The concept of one common stem cell seemed to be supported both by experiments using 3H-thymidine autoradiography (Fujita 1963, 1965b, 1966; Sauer and Walker 1959; Sidman et al. 1959) and by ultrastructural studies (Fu- jita 1966; Hinds and Ruffet 1971; Wechseler and Meller 1967) indicating that structural differences, which His presumably used to define his two cell types, could be related to different stages of the mitotic cycle.

The basic thesis for this study was that the telencephalon is needed to make decisions in new situations. Subsidary hypotheses were that the telencephalon consists of: (a) a sensorimotor system which generates motor activity from sensory input and (b) a selection system which makes choices from possible motor programs. It was postulated that the selection system should fulfil the following requirements: be accessible for past and present events, have the capacity to process this information in a nondetermined way with a possibility for ordering, and have access to motor-affecting systems (the sensorimotor system). The ability of the selection system to correlate information in a nonpredetermined way was considered most important. In short: The selection system should be able to associate any information in any combination, and have the capability for internal control of neuronal activity and external selection of motor programs (see Fig. IA. ) Xenopus laevis was chosen as a subject, since it has a relatively simple tel­ encephalon, with characteristics that it shares with "primitive" species of different vertebrate classes, and because it is easy to maintain as a laboratory animal. The main method used was the determination of connections with HRP. The pallium was in the focus of attention, since it was considered to be the core of the selection system. Immunohistochemistry was used as an additional parameter to compare Xenopus laevis forebrain with those of other vertebrates.