(Chapter 14) Footnote 3:

As pointed out by Saunders and O’Malley (1950, p. 62), “In the illustration of the vertebral column as a whole, its various curves have been very imperfectly reproduced due to the technique employed in mounting the preparation, which consisted of threading the vertebrae on a rigid iron bar bent by pure guesswork to fit the curvatures. This same defect is therefore apparent in the full-length plate of the skeleton.” This same error has been observed in several modern articulated skeletons in which the vertebrae were similarly mounted: the posterior lordatic curve of the lumbar region was either missing, or the lumbar region continued the kyphotic curve of the thoracic region. An additional reason for Vesalius’ simplification of vertebral curvature may be found in his belief that its structure was comparable to that of a keystone arch. See Chapter 16, note 59. In chapter 39 where he gives instructions for the mounting of a skeleton, Vesalius states his distaste for a spinal column with too much forward and backward curvature: “the greatest care should be taken with the iron rod [placed inside the vertebrae] that it not be ineptly bent backward and forward and hold the body upright in an unbecoming way.”