16 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



The correlated changes in coccyx and sacrum are due to 

 atrophy of the caudal vertebrae, along with fixation of the 

 pelvis, the erect attitude, &c. Correlated variations of 

 the thoracic and lumbar vertebrae are dependent upon 

 elongation or diminution of the length of the trunk, 

 possibly associated with differences in the size and rate 

 of growth of viscera, and upon the separation or approxi- 

 mation of the region of the attachment of the limbs. 



Lumbo-sacral variations are associated with the 

 mode of attachment of the hip-bone to the spine, and 

 present an excellent example of individual variation. 

 Rosenberg has formulated the theory of a phvlogenetic 

 shortening of the vertebral column by a telescoping of 

 the spine on the hip-bones, and bases his view in part upon 

 the fact that among the variations in this region there are 

 examples of four instead of five lumbar vertebrae, the 

 fifth being absorbed into the sacrum, and providing an 

 attachment of the ilium. 



When however we examine a large series of human 

 spines and sacra, we find that an excessive number of pre- 

 sacral vertebrae (25) is as common as a diminished 

 number (23). If numbers are of any account — and in a 

 democratic age every vote has the same weight — these 

 variations indicate as great a tendency to elongation as 

 to shortening in the length of the spine. Rosenberg 

 looked upon the form of spine with an excessive number 

 of praesacral vertebrae as " atavistic,'' that with a 

 diminished number as a " future " form. The plain fact 

 is that we have here a mere individual variation, indi- 

 cating an oscillation of the hip-bone round its normal 

 attachment. It may in such cases (which occur equally 

 in both directions, — 5'3 per cent.) catch on the 24th (last 

 lumbar) vertebra, or it may miss its connection with the 

 25th (first sacral) vertebra, 



