STRUCTURE OF THE INTERAMBULACRUM. 145 



in its peculiar position it would very easily drop out after the death of 

 the individual or in the processes of fossilization.^ To our mind the 

 angles for its reception are almost as stronoj evidence as the plate itself '■> 

 so it may be confidently stated that the interambulacrum of Melonites 

 originates with a single plate, as shown by Loven in Goniocidaris and 

 Strongylocentrotus, and in the succeeding paper in LepidecMnus and 

 Fholidocidaris. 



The specimen represented by figure 10, plate 3, evidently indicates a 

 condition in which some resorption of the ventral border of the corona 

 has taken place, and would correspond closel}'' to the same condition in 

 Strongylocentrotus (plate 3, figure 9). Reconstructing the ventral plate of 

 Melonites to the condition it probably had before resorption cut into its 

 ventral border, we have the condition shown in plate 3, figure 11, which 

 is an adaptation from the ventral area of interambulacrum A in plate 3, 

 figure 2. In this reconstruction of Melonites we have a ventral plate 1' 

 comparable to that seen in Strongylocentrotus before ventral resorption has 

 commenced (plate 3, figure 8) ; also as seen in Lepidechinus (plate 7, 

 figure 42) and Pholidocidaris (plate 9, figure 54). 



The single plate found at the ventral border of the interambulacrum 

 in echinoids, as shown by Loven in the young of modern forms and here 

 in Paleozoic forms, points directly to the conclusion that this stage in 

 growth represents an ancestral form having a single column of interam- 

 bulacral plates. The only such form known is the Lower Silurian genus 

 Bothriocldaris (see figure 4 of the succeeding paper, in the chapter " Conclu- 

 sions" ), which from this fact assumes the greatest importance. In Melo- 

 nites we show that newly added columns normally alternate to left and 

 right as introduced, even-numbered columns typically appearing on the 

 right of odd ones (plate 2, figure 2 ; plate 3, figures 12 and 14, and plate 

 4, figures 18 and 19). This being the case, it seems fair to argue that the 

 initial plate 1' oi Melonites (plate 3, figure 11) was the basal member of 

 column number 1, the left adambulacral, column 2, being introduced to 

 the right of it. This assumption w^ould carry with it the conclusion 

 that the left adambulacral column of Melonites and allies is probably the 

 genetic equivalent of the single column of Bothriocldaris (figure 4 of suc- 

 ceeding paper). 



Returning to the consideration of the interambulacrum of Melonites 

 multiporus, in the next row above the ventral plates 1 and 2 (plate 2, figure 

 2) there are three plates. The median plate, number 3, is hexagonal in 

 form, and it is the first formed plate of a new column; it makes, with 

 the addition of new plates dorsally, the first column of median hexag- 



* Similar angles in the two ventral plates are shown in Oligoporus (plate 6, figure 26), and the 

 initial plate is shown in Lepidechinus (plate 7, figure 42). 



