EEPORT ON THE GREEN TURTLE. 13 



clioudrocrauium, from having a flat-bottomed Batracliian form, grows up into a high wall 

 between the eyes, as well as developing a partition in the nasal region. 



Other Batrachian characters show themselves in later stages in the Turtle's skull, 

 but this is the stage at which it approximates nearest to that of the Tadpole ; this is 

 especially seen in the flat, out- turned ends (cornua) of the paired trabeculse (figs. 5-7). 



This stage is especially valuable as helping, in comparison with what is to iDe 

 seen in Tadpoles and larval Urodeles, to a clear conception of the true nature of the 

 prochordal part of the trabeculte. 



I apply the term hasineural to the paired elements of the skull-base and skull-walls, 

 and their homology with the series of paired cartillages of the spine (neural arches) is 

 clearly to be seen in the stage before us. 



The axial part, the notochord, with its thin mesoblastic sheath, stops behind the 

 oral involution ; but the three mesoblastic tracts are carried on to the frontal wall 

 of the embryo, the trabeculse continuously, and the intertrabecula, with only a short 

 tract suppressed. This, however, becomes all filled up afterwards, or long before the 

 time of hatching. 



Now, however, and during the next stage, until the middle of the incubating period, 

 the prochordal part of the trabeculse is segmented off from the parachordal part.^ 



The tissue is continuous, but the cartilage divides and forms a temporary joint, 

 inherited, I have no doubt, from some old tjrpe to whom such a joint was useful. 



The prochordal tracts (seen from above in PL II. figs. 5 and 7 ; from below, in fig. 

 6; and in section in figs. 9 and 10, tr.,i.tr.) occupy half the lower region of the skull, 

 which, however, is shorter by two-fifths than it would have been if the mid-brain had 

 stretched out in a straight line instead of folding itself into a sudden loop. 



But the chondrocranium only shows a tendency, in the curved form of the noto- 

 chord, to loop itself ; it really begins again at a new point, and the prochordal part 

 of the trabeculse is set on at the base of the ascending part. 



In the next stage we shall see a more perfect looping, even of the basal part of the 

 skull ; now, we see the floor merely breaking out again under the fore-brain. The 

 prenasal end of the intertrabecula bends, already, somewhat downwards in the front of 

 the head ; behind, it runs under the fore half of the pituitary space ; at that part 

 the three rods are all rounded (fig. 10, tr.,i.tr.). 



Further forwards, between the eyes (fig. 9), the intertrabecula is on a still lower 

 plane than in the pituitary space ; this is an important fact to be noticed, as the cari- 

 nation of the orbital septum in this type is clue to this low position of the middle bar. 

 There also the three bars are only separated by less solid tissue ; further forwards 

 (PI. IV. fig. 1, tr.,i.t)'.) they are more distinct again, and there the trabeculse give out 



' For some years I supposed this clistinetion of parts to be primary ; I am now satisfied that it is secondary (see 

 Proc. Roy. See, Feb. 13, 1879, p. .3.39). 



