MORPHOLOGY OF THE GALLIXACE.E. 219 



this stage — a very important stage, on account of the remarkable modification which 

 is seen in it of the notochord (Plate XXIII. figs. 7 and 8, n.c). The ichthyic 

 condition of the notochord, that in which the constrictions correspond with the centra 

 (figs. 1-4), has given place to an ampliihian condition, in which the principal con- 

 strictions correspond with the intercentral segmental tracts. Nor is this all, for two 

 more constrictions appear within the centrum ; thus the whole chord is moniliform, 

 and suggests the arrest of tic o out of three of the ancestral vertebra; (see Proc. Roy. Soc. 

 1888, p. 4GG). 



Now, if this fine silken thread of suggestion holds, then we may follow it, as a clue, and 

 fasten to it another, namely, that, long as is the vertebral chain of a modern bird, it has up 

 to the sacral region only one-third the number of vertebrcc in it that tbe old amphibio- 

 dipnoan ancestor possessed. If this be a true suggestion, we must give up the notion 

 (hypothesis) of the development of the bird gradually, and, as it were by chance, out of a 

 full-blown amniotic Reptile, and figure to ourselves a swarthy, worm-like creature, 

 long and unsightly, Avith no neat five-toed paws, but with simple paddles, supported from 

 within by many rays, with many joints in them; in morphological language, they pos- 

 sessed not cheiroj)terygia, but ichthyopterygia. The tendency to further subdivision 

 of the vertebral chain is not seen in the sacral and caudal regions, or only to a slight 

 extent in the fore part of the sacrum (Plate XXIII. fig. 9). These regions are still un- 

 ossified, although the hip-girdle is undergoing that change. There are twenty vertebrae in 

 front of the sacrum ; that region and the tail give us twenty-five more. The fore limbs 

 being attached very loosely to the spine, the latter is but little modified in relation to 

 these limbs ; the intervertebral passages simply being larger to allow the exit of the 

 large nerves that form the brachial plexus. But the necessity for fixity and strength 

 in the hind-quarters has operated upon the third of the four regions of this chain ; in 

 all this region and in the hinder half of the fourth, or caudal, the vertebrae lose tlieir 

 freedom of motion. The Iguanodon, to a great degree, and the Bird to a greater, has the 

 need for a long sacral series of vertebrte, which, if harmonized with the homologous 

 vertebrae of a Lizard or a Crocodile, must have secondary regional names ; but this human 

 anatomy term, useful in spite of its etymological absurdity, is quite arbitrary, and can 

 be applied to one or two, or to two dozen segments. 



Here, in the Eowl, the first has developed ribs and helps to form tlio thoi-ax ; the next 

 three, with their arrested ribs, are similar to our luml^ar vertebraj ; then come the four 

 true sacrals. Now, after two or three days' further incubation, the sacrals show their 

 greatest expansion, having to hold that swollen and ventricular part of the myelon 

 which gives ofi" the large sacral nerves : thus the axial extent of these segments is only 

 one-fourth of their bilateral growth. 



This bellying of the sacral vertebra? is not sudden ; it begins in front of the sacrals 

 proper, and is continued into the foi'c part of the uro-sacral region ; the first and second 

 vertebrie of that part are seen to have a rib-bar, right and left. The third caiidal has the 

 longest transverse processes ; in the uropygial series they are lost ; the spinal projections 

 are low, they begin again on the uro-sacral series, and they have died out on the sacrals 

 proper ; the dorso-lumbar joints of the general sacrum have spines like those of the 



