358 Mr. H. G. Seeley on a Theory 
be maintained that the inferior arch of a cervical vertebra of a bird 
differs less from the inferior arch of a dorsal vertebra than does 
the ordinary upper arch of a vertebra from the upper arch of a 
segment of the skull. In the thoracic region the growth and 
development of viscera is chiefly in depth, as is the weight of the 
lungs; and in Amphioxus lanceolatus the notochord extends 
anterior to the neural cord, whereas in mammals, even in a very 
early embryonic state, the neural rudiment which becomes the 
brain is prolonged far in front of the notochord; and thus it is 
seen that with its development in height the brain undergoes a 
development in length, which the thorax did not. And nothing 
can be more evident than that, restrained by the structures in 
front and by the vertebra behind, the growth in length must 
exercise a pressure and tension in that direction exactly corre- 
sponding to the forces which gave rise to the epiphysial bones 
which roof in the brain as it developes in height. And there- 
fore, since by the influence of such enormous and equable 
pressure and tension epiphyses are developed in height, exactly 
the same forces exerted in length cannot but have produced 
epiphyses at each end; and so, remembering how, up to a 
certain point, the plan of the brain and the spinal cord must 
have been the same, it is curious to observe that while the basi- 
sphenoid developes the basioccipital and presphenoid for its 
epiphyses much after the plan of an ordinary centrum, the bones 
of the neural arch also develope epiphyses in length just as they 
do in height, as we saw was the case with some fishes—the en- 
tire occipital segment answering to the posterior epiphysis, and 
the entire frontal segment beg the anterior epiphysis of the 
parietal segment of the skull. And accordingly it is found that 
the elementary bones of these epiphyses converge and close in 
the brain at both ends, thus demonstrating that they owe their 
growth to its growth, and extend no further than they are forced 
by its pressure; and therefore, though the skull will obviously 
develope quite regardless of the degree of growth in the several 
parts of the brain, by the simple law of inheritance, yet in many 
cases the relative size of several bones will be found to vary with 
the size of the division of the brain which is underneath them. 
Thus Mr. Robertson remarks that fishes may be divided into a 
sluggish group, typified by Lophius, in which the cerebellum is 
small, and an active group, in which the cerebellum is large, 
typified by the Tunny; and finds that in skulls of equal length, 
the occipital segment of-the skull measures 42 inches long in the 
Tunny, while in Lophius it only measures 2 inches: and, ascend- 
ing in organization, it is seen that as the brain rapidly expands, 
bones which before, in the lower forms, were quite exterior to 
the skull become gradually introduced to form part. of the 
cranial walls. 
