STRUCTURE OF THE INTERAMBULACRUM. 149 



the preceding column to that one to which the terminal pentagon belongs 

 (see plate 3, figure 12; plate 4, figures 18 and 19; plate 5, figure 20). 

 This position of heptagonal plates, however, is a character which is sub- 

 ject to somewhat frequent exceptions, as shown in the tables on pages 

 165 to 170. The above may be accepted as the rule in the method of 

 introduction of new columns in Melonites, Oligoporus,Sind other Paleozoic 

 echini. 



As new columns were added as the animal progressively grew dorsally,* 

 and as new columns were introduced by a pentagonal plate, an apex of 

 which pointed ventrally or toward the oral area (plate 3, figure 12), there 

 is in this feature an important aid in diagnosing the relative position 

 of the axes in even fragmentary specimens. It has been ascertained 

 that this conclusion is correct by finding the termination of columns as 

 stated in 14 specimens in which the axes were positively known from the 

 presence of genital and ocular plates (plate 3, figure 13). This feature is 

 one that has been overlooked by previous writers, for in almost all pub- 

 lished figures of Melonitidse, and other Paleozoic echinoids as well, in 

 which the termination of columns is shown, the apex of the terminal 

 pentagon points toward the upper pole of the figure, which should be the 

 anal pole. This demonstrates that such figures were not correctly 

 oriented, the oral and anal ends being transposed. This remark applies 

 to Meek and Worthen's (30, 31) .figures of Rhoechinus gracilis, Oligoporus 

 danse and Lepidocidaris squamosus, Hambach's (18) figure of Melonites 

 crassus, and M'Coy's (28) figure of Palseechiniis sphsericus, as well as some 

 others. 



In passing toward the anal area where the younger plates are situated, 

 we find that there is a gradual passage from the hexagonal form char- 

 acteristic of the older plates of median columns. This change of form is 

 shown well in a choice specimen of Melonites multiporus which Professor 

 Wm. B. Clark, of Johns Hopkins University, kindly loaned for study 

 (plate 3, figure 13). It is also shown in plate 5, figure 20. In the pro- 

 gressively younger plates, the form is gradually modified by the progres- 

 sively shorter lengths of the upper and lower sides of the hexagon until 

 the plates have a more or less rhombic form. At this part of the test it 

 would be more convenient to study the plates in a descending order, for 

 the growth is properly from the early rhombic form to the mature hex- 

 agonal form. In passing from the rhombic form to the hexagonal it is 

 occasionally seen that while the upper end of the rhomb retains its 

 character the lower portion is truncated by a horizontal edge, thus form- 

 ing a pentagonal plate with its apex directed dorsally, as in plate P, 

 of figure 13, plate 3. There is, however, no need of confusing this type 



* See further discussion under " Conclusions " in the succeeding paper. 



