508 Development of Mouth and Gills in Bdellostoma 
is larger and more complex, and the section shows, fortunately, a portion 
of the tube leading to the eetoderm. The ectoderm itself now forms a 
slight pocket and is reduced in thickness at this point. Around the ex- 
ternal gill tube the cartilage arch, ba, is forming and it is seen to be 
entirely an extra-branchial arch in all cases except the cesophago- 
cutaneous duct. And here the cartilage extends entirely to the gut, 
oftentimes spreading out slightly along its walls, as Ayers and Jackson, 
oo, have also pointed out. As seen from Figs. 30 and 31 the gills 
in the mid region at this stage are becoming better developed than the 
anterior ones. Indeed in the hatched embryo the extreme anterior and 
posterior pouches are rarely so well formed as those nearer the middle 
of the line. 
Finally, Fig. 32 shows a section through a gill of a lately hatched 
embryo. The gills now, of course, open to the exterior and are perfectly 
pouched. In the figure e¢ is a cross section through the external gill 
tube which now leads from the pouch to the skin in an almost posterior 
direction. Thus the characteristic marsipobranch is entirely derived from 
the endodermal portion of the gill and reaches its adult complexity by a 
spreading out and subsequent folding of the walls of the originally sim- 
ple gill tube. The ectodermal portion of the final gill structure is insig- 
nificant, being nothing more than a short rim about the external gill 
opening. 
The Cartilage Arches,in their embryonic condition are,as far as I could 
judge from a comparison with Ayers and Jackson’s description, very 
similar to these structures in the adult. At some stages they appear 
to be slightly more extensive. These differences, however, are insufficient 
to base conclusions upon since all parts of these animals are subject to 
such wide variations. The branchial skeletons of the Marsipobranchii have 
long been contrasted with the typical piscian condition as extra-branchial 
instead of intra-branchial. This distinction in Bdellostoma at any rate, 
as Ayers and Jackson maintain, is not conclusive for the reason that the 
cartilage of the cesophago-cutaneous duct, which is a gill tube, extends 
entirely into the pharyngeal wall. I think that a very plausible explana- 
tion may be offered for the condition in the other cartilage arches which 
will show them to be secondarily extra-branchial. It will be recalled from 
the preceding description that all of the gills of Bdellostoma are at one 
time in their development simple tubular structures just as the cesophago- 
cutaneous duct always is. Now from analogy I assume that if these gills 
should remain tubular until the stage in development when the. cartilage 
forms they also would have sheaths of cartilage supporting them from 
their outer ends along their entire length to the pharyngeal wall just 
