An attempt has been made to correlate the roentgenographic and anatomical findings at the lumbosacral junction by careful comparison with dissected specimens. Besides the statistical data on the various anomalies which occur in this region, two findings were of interest. The first was that the degree of osteo-arthritis present in a given case is always great deal more advanced than the roentgenograms would indicate, particularly as far as the posterior elements of the vertebral segments are concerned. The second was that there are confusing findings in the roentgenograms at times which, because of the varying projection of the central roentgen-ray beam and the asymmetrical development of the articular processes, will produce an appearance suggestive of failure of fusion, but in reality the segments will be found quite intact. The variations in the anatomical structure have been briefly discussed.
The oblique, coronal (curved) plane is perhaps the most desirable for the human lumbosacral articulations, and the approach to the coronal plane should be considered the "evolutionary normal".