The spines have bony structures as well as a spinal canal and spinal cord. The lumber spines are most commonly affected by congenital or acquired diseases because of their widest range of motion and heaviest weight bearing. There are various kinds of radiographic examinations to diagnose the disease of the lumber spine : simple spine radiography (A-P, lateral and both oblique views), tomography, myelography, spinal venography and discography, etc. The lesions affecting the bony struct ures are bestseen in the simple radiogram and tomogram. The lesion, however, developed within the spinal canal or in the cord are mainly demonstrated in the myelogram. So, without any evidence of bone changes in a case of spinal cord tumor, it is difficult to diagnose only by simple radiogram. In 1934, Elsberg and Dyke published normal values for interpediculate distances as determined on simple radiograms of the adult human spine, and they insisted that the measurement of the interpediculate dis ances was helpful in the diagnosis of the spinal cord tumor. Thereafter, Hinck and Schwarz published normal values for interpediculate distances f normal child and adult. The sagittal diameters of the cervical and lumbar spines were also measured by Hinck and Burrows. But, there was no previous report about the measurement for the diameter of subarachnoid space using normal myelogram. The author measured the tranaverse diameters of the dye column in the normal lumbar myelograms of 70 male and 30 female adults. The magnification ratio and individual differences were studied by the measurements of the transverse diameters of the 3rd lumbar spine in simple A-P radiogram and fluoroscopic radiogram transverse diameters of the 3rd lumbar spine in simple A-P radiogram and fluoroscopic radiogram of 10 males and 5 females. The magnification ratio of the fluoroscopic shadow to the simple of 10 males and 5 females. The magnification ratio of the fluoroscopic shadow to the simple radiogram was 1 2 times, and the individual differences were around 2 per cent of their values. The measurements of the dye column were carried out in 11 different levels. The widest dye column at the level of L-1 was determined to be the first measurement level, and the narrowest dye column just below the first level, to be second measurement level. The odd numbered levels were wider than the eve numbered levels because the nerve roots were at the odd numbered levels. The 12th measurement was taken between horizontal line draw through the bilateral neural sleeves of the 5th lumbar spine and the end of the cul-de-sac. All the values of the male were slightly bigger than those of female, but the differences were less than 1mm. The widest transverse diameter of the dye column was observed in the 7th measurement level where the 4th lumbar neural sleeves were located, and the values were 23.1$pm$3.36mm. inmale and 22.5$pm$3.30mm. in female. The narrowest transverse diameter was observed in the second and th rd lumbar interspace, and the values were 16.8$pm$2.97mm. in male and 16.9$pm$2.50mm in female. The transverse diameters of the 3rd lumber spines were measured in the same fluoroscopic films, and the means were 59.8$pm$4.05mm. in male and 51.9$pm$2.74mm. in female. But the values of each level were almost same even with the differences of the transversediameters of the 3rd lumbar spines. The individual differences of the intedpediculate distances of L-3 were very small, and there was no significant correlation between each value for dye column and each interpediculate distance of L-3. The diameter of the dye column in the L-3 was 71 per cent of interpediculate distance at the same level.