364 
plete (Fig. 2). The terminal cartilages of the first ribs remain small, 
but those of the second pair expand considerably in a postero-internal 
direction, so that even before the embryos leave the mother the verte- 
bral skeleton has already assumed the characteristic aspect of that 
of the fully grown animal. 
The foregoing observations raise that difficult question of the 
striet application of the terms rib and transverse process. According 
to the latest views (GApow, 15) the transverse process and rib are 
to be regarded as two parts of the same chondrifying tract, the former 
being continuous with the neural arch of the vertebra, while the latter 
is autogenous, although attached to the distal extremity of the former. 
If the tract chondrifies without any break, the term transverse process 
is applied to the whole. This may be a very convenient hard and 
fast rule for general application, but what is to be done in cases 
where ribs are distinct in the embryo and fused up in the adult, as 
happens in both Xenopus and Pipa? This is one of those interest- 
ing cases which serve to show how limited in application are our 
conceptions of morphology. 
It is a fact well known among zoologists (although, I must con- 
fess, I have never seen the statement in print), that the sacrum of 
Pipa is compound. The diapophyses that support the ilium of the 
pelvis are not merely those of the ninth vertebra, but represent the 
combined diapophyses of the ninth and tenth vertebrae. The diapo- 
physes of vertebra 9 form by far the greater part of the expanded 
plate, but those of the tenth vertebra, representing the small diapo- 
physes that stand out freely from the urostyle in the Discoglosside, 
reach fully to the lateral edge of the plate, and are connected with 
the ilia. A shallow groove is to be seen in some specimens marking 
out the limit between the two diapophyses, but the strongest evidence 
is to be found in the position of the aperture transmitting the ninth 
spinal nerve, not behind the sacral diapophysis as in most Anura, but 
below it. The part of the bony plate behind this aperture belongs 
to the tenth vertebra. 
Although this fact is so generally admitted, no one seems to 
have ventured to criticise v. JHERING’s figure (25, Fig. 2 B, p. 301) 
and description of the vertebral column of Pipa, in which he repre- 
sents the sacral diapophysis as formed entirely by the ninth vertebra 
(his vertebra G), and with the ninth nerve issuing behind it. Should, 
however, any further evidence of the compound nature of the sacrum 
be required, it is to be found in abundance in the earlier develop- 
ment of the vertebral column. In tailed embryos the sides of the 
neural arches of the vertebre 9 and 10 are connected by a continuous 
