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



In my Monogra^sli (1862, p. 412) I founded a separate subfamily, Dorataspida, for 

 the " Acaiithometra' cutaphractce" which I considered as the first subfamily of the 

 " Ommatida." That sujjfamily contained at that time only two genera, Dorataspis 

 (with seven species) and Haliommatidium (with five species). A thii-d genus, 

 Aspidomina (with two species), was united by me with the Haliommatida (because of 

 its double shell). For a fourth genus (Diploconus) with a single species I founded 

 the peculiar family of Diploeonida. Therefore the whole number of A c a n t h o- 

 p h r a c t a descriljed in my Monograph amounted only to four genera and fifteen 

 species. Now the rich collections of the Challenger have added such a great number 

 of new forms, that we may distinguish here thirty-eight genera and two hundred and 

 twelve species. 



Richard Hertwig in his excellent work (Der Organismus der Eadiolarien, 1879, p. 25) 

 separated his " Acanthophractida " perfectly from the "Ommatida" (or the siliceous 

 S p h 86 r i d e a), and united them with the " Acanthometrida " in his order 

 " Acanthometrea." But he separated them also from the nearly allied Diplo- 

 eonida, following my former arrangement. He distinctly noted that the skeleton 

 in all these Acanthophractida (as well as in the Acanthometrida) consists not of silex 

 but of the organic substance " acanthin." 



The astonishing number of new and interesting forms of Acanthophracta 

 which I have found in the rich collection of the Challeno;er enables me to distinguish 

 now in this suborder six difi"erent families, two of which are perfectly new (the 

 Sphserocapsida and the Hexalaspida). But the four other families also are so much 

 enlarged that their interesting morphology appears in quite a new and clear light. Far 

 the largest and most important of these six families is that of the true Doratasjiida, 

 which embraces seventeen genera and one hundred and eight species (more than the 

 other five families together). From this largest and oldest ancestral family four other 

 families have afterwards arisen, whilst a single family, the Sphterocapsida, seems to 

 possess no direct phylogenetic connection with the five other families. 



The peculiar and quite new family of Sphserocapsida (PI. 133, figs. 7—11 ; 

 PI. 135, figs. 6—10) di3"ers from all other Acanthophracta in the singular 

 structure of the spherical acanthinic shell, composed of innumerable small plates or 

 aglets, each of which is pierced by a very small porule. This peculiar pavemented shell 

 (enclosing the central capsule and separated from it by the jelly-like calymma) seems 

 to be produced on the surface of the spherical calymma, immediately by secretion of 

 the pseudopodia, and independently from the twenty radial spines, united in the centre 

 of the sphere. On the twenty points, where the spines perforate the shell, there are 

 originally eighty larger pores (four around each piercing spine) ; but there is no certain 

 indication that the shell is produced by the meeting apophyses of the twenty spines, as 

 is the case in the five other families of Acanthophracta. Therefore perhaps it is 



