2G4 THE VOYAGE OF H.M.S. CHALLENGER. 



13. Abactinal area with flat .stellate plates. Covered with a uniform granulose 



membrane Goniodiscin^. 



a. Abactinal area with polygonal or substellate plates, overlaid witli a 



granulose membrane which invests the whole test, 

 (i. Actinal area naked. Margin of test thick. 



a. Infero-marginal plates with one or more lateral spines . . Stellaster. 



/?. Infero-marginal plates devoid of lateral spines .... Ogmaster. 

 b. Actinal area covered with a membranous skin. Margin angu- 



lated, and with a marginal fringe of small conical thornlets . . Leptogonaster. 



b. Abactinal area with stellate plates, leaving wide interspaces for the 



passage of papulae .......... GonioJiscus. 



C. Abactinal area with small stellate plates bearing true paxillae. Actinal inter- 

 mediate areas with imbricating plates in transverse series, bearing 

 paxilliform groups of spines. ....... Mimasterix.k. 



a. A single genus ............ Mimaster. 



Subfamily Pentagonasterin^e, Sladen, 1888 (non sensu Viguier, 1878). 

 Genus Pentagonaster, Linck. 



Pentagonaster, Linck, De Stellis marinis, 1733, p. 20. 



Goniaster (pars), L. Agassiz, Prod. Mon. Eadiaires, Mem. Soc. Sci. Nat. Neuchatel, 1835, t. i. p. 191. 



Astrogonium (pars), Miiller and Troschel, System der Asteriden, 1842, p. 52. 



Goniodiscus (pars), Miiller and Troschel, System der Asteriden, 1842, p. 57. 



Hosia (pars), Gray, Ann. and Mag. Nat. Hist., 1840, vol. vi. p. 279. 



Tosia, Gray, Ann. and Mag. Nat. Hist., 1840, vol. vi. p. 281. 



Two species only were referred by Linck to his genus Pentagonaster. One of them, 

 Pentagonaster regularis, has not since been recognised, and, as the type is lost and its 

 identification now could be nothing- more than a guess, the name should be discarded 

 altogether. The second form, Pentagonaster semilunatus, is a perfectly well recognised 

 species, about which there is no doubt, and in my opinion it should certainly be held as 

 the type of Linck's genus Pentagonaster. On these grounds I cannot agree with Perrier ' 

 in removing Pentagonaster semilunatus to the genus Astrogonium, as defined by him, for 

 so long as the genus Pentagonaster is preserved I consider that Pentagonaster semilunatus 

 must unquestionably be referred to it. The term Astrogonium, as used by me, is 

 restricted to limits different from those recently assigned to it by Perrier, 2 when restoring 

 it to generic rank. 



Cliorology of the Genus Pentagonaster. 



u. Geographical distribution : — 



Atlantic : Twenty species between the parallels of 75° N. and 55° S. 



In the northern area, '• Pentagonaster granulans, occurs both 

 on the eastern and the western sides, being found off' the coasts of 

 1 .4 nn. Sci. Nat. (Zool), 1S85, t. six. p. 37. 2 Loc. cit, p. 30. 



