374 HAWAIIAN AND OTHER PACIFIC ECHINI. 



reasons for not accepting the new genus proposed by Doderlein, of which oblonga 

 is the type are given above (p. 367). The series of specimens at hand is most 

 perplexing, so great is the individual diversity shown. On the one hand, there 

 are from Clarion Island, black specimens with high, compressed test, big abacti- 

 nal tubercles, and small abactinal system with all the oculars completely ex- 

 cluded from the periproct, but with spines tapering to a point and not nearly as 

 thick as they should be in oblonga. And on the other hand there are specimens 

 from Makemo in the Paumotus, which have the characteristic spines of oblonga 

 but show a remarkable diversity in color; some are deep purple, some grayish 

 lavender, and one is dull purplish red. In the presence of these puzzling 

 specimens, much sympathy is felt with de Meijere's remark, apropos of 

 oblonga, "Dass letztere eine gute Art 1st, glaube ich kaum." 



Mortensen first called attention to the curious fact that the globiferous 

 pedicellaria? of oblonga have a jointed stalk. Examination of a considerable 

 amount of material shows that this peculiarity exists in less than half the globif- 

 erous pedicellariae; at least, more than half of those examined did not show it. 

 Such a joint in the stalk of the pedicellarige is not known in any other echinoid, 

 although it is easy to see how it might be of real use. 



The "Albatross" obtained specimens of oblonga from the following stations: 



Laysan, Hawaiian Islands. Hilo, Hawaii, H. I. 



Necker Island, H. I. Puako Bay, Hawaii, H. I. 



Lanai, H. I. Rangiroa, Paumotu Islands. 



Kamalina Bay, Niihau, H. I. Makemo, Paumotus. 



Hanalei, Kauai, H. I. Fakarava, Paumotus. 



Waianae, Oahu, H. I. Clarion Island, Mexico. 



Honolulu Reef, Oahu, H. I. 

 Sixty-seven specimens. 



Echinometra insularis, 1 sp. nov. 



Plates 95, figs. 13-15; 114, figs. 3, 4. 



On first examination the Echinometras from Easter Island showed considerable 

 resemblance to Van Brunti, but further study soon proved that they are more 

 nearly related to Mathcei. The test is rather wide and flat, the width being 

 .S5-.95 of the length and the height about half the length or a trifle less. The 



1 Insularis = pertaining to an island. 



