386 ROBERT TRACY JACKSON ON ECHINI. 



is smaller, but is more complete and instructive (Plate 58, fig. 3; Plate 59, fig. 5). Besides 

 these there are several fragmentary specimens known. 



The test is large and was probably spheroidal. Melon-like ribs are entirely wanting, and 

 in this character it is like M. keepingi and equally difi'ers from American species. The ambul- 

 acra form a gently rounded curve, doubtless corresponding to the curvature of the whole test. 

 Keeping thought that the ambulacra formed ten broad elevated ribs; but I see no support 

 for this view. The width of a half-ambulacrum (Plate 58, fig. 3; Plate 59, fig. 5) is about 24 

 mm., which would give about 48 mm. as the width of the whole ambulacrum at or near the 

 mid-zone. In the large slab which was figured by Keeping, the ambulacra are somewhat nar- 

 rower, a half-ambulacrum in the widest part seen measuring 16 mm. in width. 



In the interambulacra as shown in one specimen there are at least seven columns of plates 

 in an area. There are six columns in place and a left adradial column is restored (Plate 58, 

 fig. 3; Plate 59, fig. 5). On this basis the width of the area would be about 48 mm., the same 

 width as the ambulacrum in the same individual. Keeping assumed that there were eight or 

 nine columns of interambulacral plates. This may be correct, but there is no proof of it. The 

 larger interambulacral plates measure from 8 to 8.2 mm. in width by about 6 mm. in height 

 and 4 mm. in thickness. On the basis of the measurements from the more nearly complete 

 specimen (Plate 58, fig. 3), with ambulacra about 48 mm. in width and interambulacra of the 

 same width, the circumference would be about 480 mm. and the diameter about 153 mm., a 

 very large sea-urchin, exceeding any species known in the genus. 



The ambulacra are very wide as shown by the above measurements, and in one specimen 

 (Plate 58, fig. 5) a half-area is quite complete at a zone probably near the mid-zone. This 

 has six columns in the left half-area, which shows that there were twelve columns of plates in 

 a complete ai-ea. They are composed of relatively narrow occluded plates, narrow demi-, and 

 four columns of isolated plates in each half- area. The occluded plates (as in M. keepingi) are 

 relatively narrower than in American species, and are not built up into an elevated melon-like 

 ridge as in American species. The two occluded columns are clearly distinguished because 

 the pore-pairs are set on opposite sides of the plates (p. 27), which definitely locates the center 

 of the area. Demi-plates bevel over the adambulacrals on the adradial sutures. 



In most of the specimens the interambulacral plates are dissociated, so that the number of 

 columns is quite doubtful, but in one (Plate 58, fig. 3; Plate 59, fig. 5) there are one pentagonal 

 adambulacral and five hexagonal median columns in place. To complete the area, at least one 

 left adambulacral column must be added, as indicated by dotted lines. There may have been 

 more than seven columns. With seven columns the interambulacrum has about the same width 

 as the ambulacrum, but, if there were more than seven, then the interambulacrum would be 

 so much the wider. Ambulacra! and interambulacral plates bear numerous secondary tubercles 

 and spines. The spines are tapering, the most nearly complete being about 2.5 mm. in length. 



