368 
seventh vertebra, and HAsse (20, p. 254) confirms the statement, but in 
none of the embryos which I have examined have I found the hypo- 
chordal cartilage reaching farther forward than the ninth vertebra- 
Perhaps my specimens were not young enough to show the cartilage 
at its maximum development. The splint of bone in the hypochordal 
cartilage is, in the earlier of the two stages here figured, separated 
from the undifferentiated neural arches of the postsacral vertebre by 
the thickness of the notochord. As the notochord becomes absorbed, 
the ventral splint of bone rises to the level of the latter, which have 
by this time undergone considerable ossification; and the urostyle of 
the adult is the result of the coalescence of the two. Fig. 2 shows 
an interesting stage in which the ventral and dorsal constituents of 
the urostyle have come into contact but are still readily separable. 
The urostyle is fused up with the sacrum in Pipa’), as it also is in 
Xenopus, Breviceps (40, p. 16), Phyllomedusa (1, p. 362), 
Pelobates, and, with less constancy, in some other genera. A study 
of the embryonic spine of Pipa shows that this is not a post-embryonic 
coalescence, but is rather due to the failure of the parts to separate, 
for the cartilaginous neural arches 9 and 10 are continuous from their 
first appearance. 
Xenopus levis. 
In Xenopus the developing vertebre are much narrower in pro- 
portion to the size of the notochord than in Pipa, so that in a ventral 
view but little of the neural arch is to be seen (Fig. 3). At the stage 
here figured the notochord is beginning to undergo absorption. This 
fact, taken in conjunction with the relative size of the sacral diapo- 
physes, would indicate that the vertebral column has reached about 
the same stage of development as in the earlier of the two Pipa 
embryos figured. Owing to the great difference in the habits of the 
free-living Xenopus larva and the Pipa embryo, comfortably en- 
sconced in the integumentary pits on the back of the mother, it is 
not possible to judge of the relative ages of the two by general-ex- 
ternal features. The development of the vertebral column is probably 
less affected by external conditions than most of the organs of the 
body, and may safely be taken as a criterion of age. 
The larval vertebral column of Xenopus has, so far as I can 
gather, never been studied in detail: it has always been assumed to 
be similar to that of the young Pipa, thus presuming upon the pos- 
sible intimacy of the relationship existing between the two genera. 
1) Copr's statement (9, p. 98, and 10, p. 252) that in Pipa the 
urostyle is simple and attached to a single condyle is, if not misleading, 
liable to considerable misinterpretation. 
