REPOET ON THE RADIOLARIA. 403 



" Calodictya and Lithocydidina" of Ehrenberg and a great part of his " Haliommatina." 

 As three different subfamilies of that family I separated the Coccodiscida (with 

 five genera), the Trematodiscida (with seven genera), and the Discospirida (with two 

 genera; Monogr. d. Radiol., p. 485). A fourth group of Discoidea was constituted 

 by the Spongodiscida (with eight genera, including the Spongocydida) , which at that 

 time I united with the Spongurida, because of their spongy structure (loc. cit., p. 452). 



As the number of fossil Discoidea found in the Tertiary rocks of Barbados and of 

 the Mediterranean shores (Sicily and Greece) is comparatively very large, we find even 

 in the first system of Polycystina of Ehrenberg (1847), not less than twelve genera 

 distinguished, viz., six Calodictya, two Haliommatina, and four Lithocydidina (Monatsber. 

 d. k. preuss. Akad. d. Wiss. Berlin, 1847, p. 53). The whole number of Radiolarian 

 genera distinguished in that first system was forty-four. The diagnoses of them given 

 by Ehrenberg were as usual very insufficient. The characters of the three families 

 given by him were the following: Calodictya "Testarum intus spongiosarum et nucleo 

 destitutarum orbes ; Haliommatina Testae subglobosse nucleus radiatus; Lithocydidina 

 Testarum disci in media parte nucleati margine cellulose." In the latest work of 

 Ehrenberg (1875, p. 157) the same system was repeated, but some new genera added; 

 and thirty-eight different species, appertaining to the Discoidea, were figured in the 

 same work (Abhandl. d. k. Akad. d. Wiss. Berlin, 1875, Tafs. xx.-xxx.). 



Richard Hertwig, 1879, in his excellent work, Der Organismus der Radiolarien 

 (pp. 57-68), gave a detailed description of the skeleton of some Discoidea, and 

 arrived at the conclusion that this whole family had a spirally constructed skeleton, 

 and should therefore be derived from the Lithelida. But this conclusion is certainly 

 erroneous, and in my opinion the whole explanation of that spiral structure, and of its 

 signification in the development of the Discoidea, is the weakest part of that other- 

 wise very important work. 



In my Prodromus (1881, p. 456) I gave a provisional system of the Discida 

 or Discoidea from the immense quantity of new material collected by the Challenger, 

 and could distinguish not less than eighty-four genera. This number is from sub- 

 sequent research only augmented by seven, so that in the following pages ninety-one 

 genera with five hundred and one species are described. In the Prodromus I had 

 disposed them in four different families, which number is now increased to six. These 

 six families can be again disposed in two main groups or sections, the Phacodiscaria 

 and the Cyclodiscaria, each section with three families. 



The Phacodiscaria are characterised by the possession of a typical "phacoid 

 shell," and contain the three families Cenodiscida, Phacodiscida, and Coccodiscida. On the 

 other hand, the Cyclodiscaria are distinguished by the absence of such a " phacoid 

 shell," and contain the three families Porodiscida, Pylodiscida, and Spongodiscida. Both 

 sections exhibit an analogous development. 



