PALMECHINUS GIGAS AND RHOECHINUS ELEGAJSS. 205 



tained the maximum number of interambulacral columns, and during 

 later growth only i\qw rows were added to the columns already existent. 

 A similar condition of early attaining the full number of columns exists in 

 Oligoporus nobilis, as figured by Meek and Worthen (31), and an approach 

 to this condition is seen in Oligoporus coreyi (plate 6, figure 28). Most 

 Palseechini, on the contrary, are less accelerated in their development 

 iind do not attain the full number of columns until much later in life, 

 as shown in this paper in Melonites, Oligoporus danse, Pholidocidarus, Lepi- 

 dechinus and other genera. 



The ambulacral area of Palseechinus gigas (plate 7, figure 39) is inter- 

 esting for comparison with other species of its genus and also with the 

 genus Oligoporus. M'Coy's (28) figures of the ambulacrum of this species 

 show 4 pores in each plate, which is evidently an error. In each half 

 ambulacrum the plates are drawn out so that none of the plates cross the 

 area; they must therefore be considered as 2 columns of plates b and b\ 

 each plate having 2 ambulacral pores. 



In Rhoechinus elegans, M'Coy we find a different type of ambulacrum 

 from Palaeechinus, and this type may be taken as representative of the 

 genus. In a specimen of Rhoechinus elegans from the Carboniferous of 

 Hook Head, Ireland, in the Museum of Comparative Zoology (catalogue 

 number 3002) an ambulacral area is clearly shown (plate 7, figure 40). 

 In this it is seen that while alternate ambulacral plates overlap one an- 

 other slightly or occasionally entirely on the outer border, yet each plate, 

 or practically each plate, may be said to extend quite across its own half 

 ambulacral area. It may be considered as representing the first step in 

 the change from a form of plate which extends entirely across its half 

 area unmodified, as in Bothriocidaris (figure 4, page 234), Archseocidaris 

 and Cidaris (plate 8, figures 43, 48), to that condition where each plate 

 extends only part way across its area, as in Falseechinus gigas (plate 7, 

 figure 39). • This is an important consideration in view of the changes I 

 have traced in the development of Oligoporus (plate 6, figure 25), where 

 ventrally the plates extend quite across the area, and succeeding plates 

 as introduced gradually fall short of this relative length in an increasing 

 degree until the length of each plate is only half the total width of the 

 half ambulacrum in which it is situated (plate 6, figure 30). The ana- 

 tomical and morphological relations of the ambulacral plates in these 

 several genera is diagrammatically shown in figure 1, page 191, to which 

 attention is especially called. In this diagram and the accompanying 

 plates we see that columns ab of Rhoechinus elegans are the equivalent of 

 columns a-\-a' and 6 + 6' of Polaeechinus gigas^ and consequently columns 

 a b of Rhoechinus are the morphological equivalents of columns a b at the 

 base and a-\-a' and b-\-b' at the upper portion of Oligoporus coreyi (plate 



