In this fore-half of the dorsal shield a median furrow quickly makes its appearance (Figure 1.123 rf). This is the broader dorsal furrow or medullary groove, the first beginning of the central nervous system. The two parallel dorsal or medullary swellings that enclose it grow together over it afterwards, and form the medullary tube. As is seen in transverse sections, it is formed only of the outer germinal layer (Figures 1.95 and 1.136). The lips of the primitive mouth, however, lie, as we know, at the important point where the outer layer bends over the inner, and from which the two coelom pouches grow between the primary germinal layers. (FIGURE 1.121. Oval embryonic shield of the rabbit (A of six days eighteen hours, B of eight days). (From Kolliker.) ps primitive streak, pr primitive groove, arg area germinalis, sw sickle-shaped germinal growth. FIGURE 1.122. Dorsal shield (ag) and germinative area of a rabbit-embryo of eight days. (From Kolliker.) pr primitive groove, rf dorsal furrow. FIGURE 1.123. Embryonic shield of a rabbit of eight days. (From Van Beneden.) pr primitive groove, cn canalis neurentericus, nk nodus neurentericus (or "Hensen's ganglion"), kf head-process (chorda). FIGURE 1.124. Longitudinal section of the coelomula of amphioxus (from the left). i entoderm, d primitive gut, cn medullary duct, n nerve tube, m mesoderm, s first primitive segment, c coelom-pouches. (From Hatschek.)) Thus the median primitive furrow (pr) in the hind-half and the median medullary furrow (rf) in the fore-half of the oval shield are totally different structures, although the latter seems to a superficial observer to be merely the forward continuation of the former. Hence they were formerly always confused. This error was the more pardonable as immediately afterwards the two grooves do actually pass into each other in a very remarkable way. The point of transition is the remarkable neurenteric canal (Figure 1.124 cn). But the direct connection which is thus established does not last long; the two are soon definitely separated by a partition. The enigmatic neurenteric canal is a very old embryonic organ, and of great phylogenetic interest, because it arises in the same way in all the chordonia (both tunicates and vertebrates). In every case it touches or embraces like an arch the posterior end of the chorda, which has been developed here in front out of the middle line of the primitive gut (between the two coelom-folds of the sickle groove) ("head-process," Figure 1.123 kf). These very ancient and strictly hereditary structures, which have no physiological significance to-day, deserve (as "rudimentary organs") our closest attention. The tenacity with which the useless neurenteric canal has been transmitted down to man through the whole series of vertebrates is of equal interest for the theory of descent in general, and the phylogeny of the chordonia in particular.