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



J. Miiller as Haliomma hystrix. A second species of Aspidomma, the Acanthometra 

 mucronata of J. Miiller, was probably an Astrolonche. At that time I placed 

 Aspidomma among the Haliommatida, led by the erroneous opinion that it might 

 represent a transition -form between Dorataspis and Haliomma. But I afterwards 

 gave up this view, as I was convinced that there is no true phylogenetic connec- 

 tion between the acanthinic Dorataspida (Actipylea) and the siliceous Haliommatida 

 {Peripylea). Thei-efore in my Prodromus (1881, p. 468) I placed Aspidomma among 

 the Dorataspida and changed its name to Phractopelta, to avoid further confusion 

 with the um-elated Ommatida (S p h se r o i d e a). It formed there, with three nearly 

 related genera, the " subfamily Phractopeltida," which we now advance to the higher 

 rank of a separate family. (By a typographical mistake the words are printed in the 

 ^voAxovcms, Phractopelma and Phraetopelmida, &c., instead of PA /-acto^^eZto and Phracto- 

 peltida, &c.). The detection of other new species appertaining to this family, and a 

 closer anatomical investigation of them, has now led to the distinction of five different 

 genera, characterised by other differences than were employed in 1881 in the provisional 

 system of the " Prodromus. " *' 



The two concentric spherical lattice-shells of the Phractopeltida, connected by radial 

 beams, correspond perfectly to those of the double-shelled Dyosphaerida [Haliomma, 

 Diplosphwra, &c.), and in both cases we may call the smaller inner the " medullary 

 shell," and the larger outer the " cortical shell." There is no doubt that the double- 

 shelled Phractopeltida must be derived phylogenetically from the simple-shelled Doratas- 

 pida (just as we derive the double Dyosphserida from the simple Monosphserida). But 

 it is not yet possible to decide positively which of the two shells is the first formed. 

 Probably the small inner or medullary shell of the Phractopeltida is the first formed, 

 and corresponds to the simple spherical lattice-shell of the Dorataspida ; and the larger 

 outer or cortical shell of the former is a later new formation, absent in the latter 

 family. This opinion seems to be confirmed by the genus Orophaspis, the only form 

 among the Dorataspida, in which the radial spines outside the shell bear free latticed 

 apophyses. If these twenty apophyses grow further and meet one another, the second 

 or outer shell of Phractopelta may be formed. But some objections may be raised 

 to this opinion from the peci;liar structure and the very small size of the inner 

 shell ; and there is some possibility that this latter is a secondary later product inside 

 of the primary cortical shell. The probable phylogenetic series which reveals the origin 

 of the Phractopeltida is the following : — Acanthometron, Zygacantha, Lithophyllium, 

 Phractacantha, Doracantha, Dorataspis, Orophaspis, Phractopelta. 



The twenty radial spines exhibit in all Phractopeltida the same characteristic 

 position and relation as in all other Icosacantha, and are constantly arranged according 

 to the Miillerian law in four meridian planes, their distal ends falling into five 

 parallel zones. Their distinction in the majority of the Phractopeltida is not difficult, 



