PLATE ARRANGEMENT OF RHOECHINUS GRACILIS. 201 



when the sum of the characters of both genera link them closely with 

 Oligoporus and Melonites as members of the Melonitidse. The whole de- 

 velopment and structural details of the several genera bear out this con- 

 clusion and as distinctly remove them from the Archseocidaridse. In his 

 paper Dr Duncan (8) gives a list of the species to be included in Eho- 

 echimis and Palasechinus, which list has been followed for these genera in 

 the systematic classification of Paleozoic Echini (table facing paoje 242). 



STRUCTURE AND PLATE ARRANGEMENT OF RHOECHINUS GRACILIS. 



The species Rhoechinus gracilis., (M. and W.), Duncan, as described 

 by the authors (31), was from the Burlington limestone of Burlington, 

 Iowa. The type specimen consisted of portions of two interambulacral 

 and an included ambulacral area. It was incomplete dorsally and ven- 

 trally, and the figure was incorrectly oriented, as ascertained by the 

 position of the terminal pentagonal plate of column 7 and its adjacent 

 heptagon. In the Student Geological Collection of Harvard University 

 there are two specimens of Rhoechinus gracilis which are very well pre- 

 served. The specimens (catalogue number 115) are in a single small 

 slab of sandstone from the Waverly group, collected by A. E. Crandall 

 on Beaver creek, one"mile above the mouth of Leatherwood creek, Menifer 

 county, Kentucky. Though the specimens are from a formation just 

 below the Burlington group, where the original material was collected, I 

 see no evidence in the fossils for ascribing them to another species. The 

 specimen figured (plate 7, figure 36) is a sandstone cast of the exterior 

 of the test, but viewed from the interior. The original curved form of 

 the test is retained in considerable degree, so that the genital ring is in 

 the center of a quite deep depression in the rock. The lettering of areas 

 is reversed from what it would be if the specimen were viewed from the 

 outside, but this makes little confusion in studying the specimen. 



The original specimen of the species showed but 7 columns of inter- 

 ambulacral plates, but it was fragmentary, and would probably have 

 shown the addition of one more column had it been perfect at the dorsal 

 area. In the specimen here figured in all the interambulacral areas there 

 are 8 columns of plates, and one of the areas exceeds that number, attain- 

 ing 9 columns. Interambulacrum A is the most perfect one. On its 

 ventral border, as far as preserved, it has a row of 5 plates, indicating but 

 5 columns at this level. In the next row from the base a sixth column 

 is introduced by the terminal pentagon 6. This column at its origin has 

 2 columns on the left and 3 on the right, which, when reversed, as it 

 must be for comparison with the outside, is seen to be the usual position 

 for this column. In the second row above pentagon 6, column 7 is in- 

 troduced by the pentagon number 7, with an equal number of columns 



