VARIATIONS IN THE INTERAMBULACRUM. 151 



These are the only cases observed of a reduction in the number of columns 

 passing dorsally in any Paleozoic echini.* 



Variations in the Interambulacrum. 



Having described an ideal normal type of plate arrangement of the 

 interambulacrum, it is desired to call attention to one of the most irregu- 

 lar specimens of Melonites multiporus seen. The specimen is in the col- 

 lections of the Wagner Free Institute of Philadelphia, accession number 

 3226, It is in a large slab with several other individuals. The specimen 

 (plate 3, figure 14) has eight columns in the interambulacrum. It thus 

 gives an added column of plates as compared with that shown in plate 3, 

 figure 12. Part of the plates of the area are hidden in the matrix, but 

 those visible show the essential features for a comparison. At the ventral 

 end, as far as seen, there are four columns of plates (plate 3, figure 14), 

 but a fifth column is soon introduced by the pentagon number 5. This 

 column at its point of origin has two columns on either side. The sixth 

 column is introduced by pentagon 6. It has a heptagon on the left, which 

 is a member of column 5, and has three columns on the left and two on the 

 right at its point of origin, as in plate 3, figure 12. The seventh column is 

 introduced by pentagon 7, having a heptagon on the left, which is a mem- 

 ber of column 5 instead of column 6, as in plate 3, figure 12. Column 7 

 has the usual number of columns on either side. Column 8 is peculiar 

 in that it originates with a hexagonal plate 8 instead of with a pentagon, 

 as in the cases previously mentioned and as seen in column 8 of the 

 cases tabulated on pages 165 to 170. I'his column at its point of origin 

 has four columns on the left and three on the right, the theoretically cor- 

 rect position for an even-numbered column (compare with plate 5, figure 

 20, and plate 3, figure 12). Hexagon 8 has an octagonal plate, 0, on the 

 left, which is a member of column 7. The octagonal form of this plate 

 is caused by the reentrant angle made by the hexagonal form of plate 8. 

 It is seen in the diagram (plate 3, figure 15) that a change to the pen- 

 tagonal form of plate 8, together with a slight shifting of associated plates, 

 would make both these plates of the usual form. The specimen is of in- 

 terest as showing the development of the eighth column in its correct 

 position by our law of alternation ; also as showing that the first formed 

 plate of a newly introduced column is not always a pentagonal plate ; 



* In the E. M. Museum at Princeton, New Jersey, there are three specimens (catalogue numbers 

 1464, 1466, 1466a) which show the jaws in place. These consist of 10 dental pyramids lying opposite 

 the interambulacral areas. When complete they meet orally in acute terminal angles. In two 

 specimens (1464 and 1466) part of the ambulacra! and a few interambulacral plates of the peristome 

 are preserved. These plates are more or less irregular, and the ambulacral plates have two pores. 

 The specimens were seen too late to include them in this paper further than by this note. 



XXI— Bull. Geol. Soc. Am., Vol. 7, 1895. 



