124 THE VOYAGE OF H.M.S. CHALLENGER. 



are well defined and prominent bourrelets. The anal system is circular, deeply sunken at 

 the extremity of the wide anal groove. 



The only specimens of this species collected were denuded of spines. The whole test 

 is covered by a uniform granulation much as in PI. XX. fig. 17, formed of the primary 

 spines with their sunken areolas, and the intertubercular space is closely covered by 

 minute mdiaries ; the areolas become less sunken towards the actinostome, and the 

 primaries are also more distant. 



The ambulacral pores continue from the petals to the phyllodes in a single row, one 

 pore to each plate placed close to the lower suture (PI. XX. fig. 17) ; they are not 

 prominent on the abactinal surface of the test. 



Station 192. September. 26, 1874. Lat. 5° 42' S., long. 132° 25' E. ; 129 

 fathonis ; mud. 



Spatangid^. 

 Family SPATANGiDiB, Agassiz, 1836, Prod. Mon. Ead. (emend.). 



POUKTALESI^. 



Sub-family Poubtalesi^, Wy. Thomson, 1877, Voyage of the Challenger, Atlantic, vol. i. pp. 376, 396. 



From the analysis given of the characters of the genera allied to Pourtalesia proper, 

 it becomes quite evident that we have to deal with a very remarkable group of Ecliinids, 

 standing in many respects in striking contrast to other Spatangoids, while possessing 

 at the same time affinities to other families of the Petalosticlia, and thus showing a close 

 structural relationship hitherto unsuspected between apparently widely diff'erent forms. 

 For these reasons it has seemed best to separate this group of Echinids as a new family 

 from the other Spatangoids, and in this I foUow the suggestion made by Thomson that 

 these species would probably form a new family characterised by their simple ambulacral 

 system ; but as I have shown this character, while well marked in Pourtalesia, 

 Paleopneustes, Echinocrejns, Cystechinus, Aceste, and the like, becomes less apparent in 

 Genicopatagus, Homolampas and Argopatagus ; and such genera as Aceste and Aerope, 

 are evidently more closely allied to the Brissina than to the PourtalesiaB, or more 

 properly Aerope, is an embryonic Brissopsis while Aceste is an embryonic Schizaster. 



The course of the alimentary canal in the Pourtalesias seems quite variable. In 

 Echinocrepis (PI. XXXV." fig. 13), after lea\dng the actinostome the narrow oesophagus 

 opens into a broader alimentary canal, which extends vertically towards the apical system, 

 turns back upon itself over the actinostome curving to the left, then sweeps in a horse-shoe 

 round the edge of the test to the opposite side of the test where ij; forms a reverse loop 

 to the left (seen from the actinal side) side of the test, and then goes straight to the anal 

 system as a somewhat narrow intestine. The alimentary canal is supported in its place in 

 this genus by a few thin mesenteries, and the two main loops are connected by a broad 



