PLATE ARRANGEMENT IN OLIGOPORUS DAN^. 195 



tinuous from their point of origin in the initial plate to the dorsal area, 

 where the plates are normally separated dorso-ventrally by the string- 

 ing-out arrangement characteristic of newly formed plates, as shown 

 in Melonites (plate 3, figure 13). This want of continuity in successive 

 plates, as stated under Melonites giganteus, page 177, is to be explained 

 by supposing that after the initial plate 9 was formed, no more plates 

 were built in this column for a period in which two rows were added to 

 the other columns of the area ; then, in the third row, additions to column 

 9 were begun afresh. It is to be observed that this break of continuity 

 in the cases described usually occurred at the point of introduction of a 

 new column. No break at any other portion of a column has been ob- 

 served in any Paleozoic Echini excepting Rhoechinus gracilis, as de- 

 :gcribed on page 203. 



When new columns are introduced, the initial plate is pentagonal; 

 ■therefore after such a break of continuity as just described in Oligojjorus 

 danas (plate 6, figure 31), the first plate, 9', built after the break occurred 

 should theoretically be pentagonal, as the mechanical requirements of the 

 €ase are precisely the same as if it were a new column. It is seen, how- 

 ever, that plate 9' is hexagonal ; but an adjoining plate, i\^, of column 7 is 

 pentagonal and bears a heptagonal plate, H, which is a member of column 

 6, on its right ventral border. This shows clearly that the pentagonal 

 and adjacent heptagonal form may in the mechanical adjustment of grow- 

 ing parts be forced on to some other plates of the same row rather than be 

 taken on by the actually new column introduced. If this very rare sepa- 

 ration of the plates of the column had not taken place and initial penta- 

 gon 9 had not been formed, I should have described this case by saying 

 that column 9 originated in plate iY, with a heptagonal plate, H, on its 

 right ventral border, the usual position ; but the position of the column 

 would be anomalous in having 5 columns on its left and 3 on its right. 

 Just such a supposed condition may be the explanation of the position 

 of the niath column described in Melonites multiporus (catalogue numbers 

 3016 and 3023, tables, pages 165 and 168). This case of OUgoporus danx is 

 very instructive and suggestive, for it seems as if it might explain the 

 apparently anomalous position of newly introduced columns in any of 

 the cases of Melonites which are figured and tabulated, as in plate 9, figure 

 49. While with the addition of a new column the middle plate of a row 

 commonly takes on a pentagonal form, which we therefore call the initial 

 plate of the new column, this form may be shoved on some other plate 

 of the same row, while the new column, which is probably still central, 

 becomes continuous with the series of plates of a preceding column. To 

 explain this idea in the case of OUgoporus danse (plate 6, figure 31), column 

 9 as expressed by its upper limits, from plate 9' upward, would be a direct 



