373 
concave below, but this curvature goes a very little way towards les- 
sening the great difference that exists between the mode of develop- 
ment of the first and second vertebre. Short, stunted diapophyses 
appear on the neural arches of vertebre 5—9, and soon assume the 
form of hemispherical tubercles (d Fig. 4). The second, third and 
fourth vertebrae have no diapophyses, but bear movable ribs, which 
later become confluent with the neural arch and form the rigid so- 
called transverse processes of the adult. ‘The circular area on the side 
of the neural arch upon which the rib articulates is shown in Fig. 4 at d’. 
The ribs ossify early, but both proximal and distal extremities 
remain for a long time cartilaginous. The distal cartilage on the third 
rib is backwardly directed, thus foreshadowing the large terminal plate 
of cartilage which occupies the same position in the adult. A similar 
pair of cartilages occur on the second pair of ribs, but these become 
reduced later, and are insignificant in the adult. The second and third 
ribs are parallel to one another; they are long and slender and slope 
slightly backwards. The first rib, in relation with the second vertebra, 
is not more than half the length of the second, and has its sharply 
pointed distal end buried in a tract of embryonic connective tissue 
which continues outwards as far as the lateral extremity of the second 
and third ribs. The rib appears to have no distal cartilage equivalent 
to that on the ribs behind; indeed, since its course is continued out- 
wards in unchondrified embryonic tissue, it is probable that only the 
proximal part of the rib ever develops. 
In Xenopus the urostyle is confluent with the sacrum, and the 
sacral diapophyses are greatly expanded. The sacrum, however, is 
simple, the lateral plates consisting of the diapophyses of the ninth 
vertebra only. The sacral diapophyses are quite insignificant in the 
earlier stages of development and do not differ materially from those 
of the vertebre 5—8 in size or shape. They enlarge very rapidly, 
however, and although their outline as seen from the ventral aspect 
is for some time semicircular (Fig. 3), they expand peripherally and 
acquire their characteristic triangular shape long before they come into 
actual contact with the ilia of the pelvis. 
The neural arches behind and including the ninth are continuous 
from the very first, so that, although by a loose mode of speech we 
may say that the sacral vertebra of Xenopus is fused with the 
urostyle, it would be more exact to say that the sacral and post-sacral 
vertebree never become separate, that the sacral vertebra does not 
differentiate from the urostyle, or that the sacral and post-sacral ver- 
tebree do not have an independent existence. (See GEGENBAUR 17, 
p. 34, in respect of the same in embryos of Pipa.) In spite of this 
