REPOET ON THE RADIOLARIA. 717 



chemical difference, an important morphological character of the skeleton also separates 

 the ACANTHARIA from all other Radiolaria : in the latter the skeleton is never centrogenous 

 or arising from the centre of the capsule ; in strict opposition to this general fact the 

 skeleton of all ACANTHARIA is centrogenous, composed of radial spines, which arise from the 

 central point of the capsule and pierce its membrane. These characteristic " radial spines 

 of acanthin," arising from the centre, are never hollow (as formerly was supposed), but 

 constantly solid. Their form is extremely variable, and most important for the distinction 

 of genera and species ; but more interesting from a general point of view is their peculiar 

 arrangement or disposition. 



The regular disposition of twenty radial spines has general value almost for all 

 ACANTHARIA, with the exception only of the small group ofActinelida. In this latter 

 group the number of radial spines is either more or less than twenty, and their disposition 

 is either quite irregular or follows a peculiar rule. The number of individuals of these 

 Actinelida, compared with that of the other ACANTHARIA, may be scarcely 1 per cent., 

 whilst the latter have more than 99 per cent. ; the number of observed species is in 

 the former about 5 per cent., in the latter about 95 per cent. Nevertheless the small 

 group of A c t i n e 1 i d a is very important, being probably the ancestral group from which 

 all other ACANTHARIA have been phylogenetically derived. These other ACANTHARIA, with 

 twenty regularly disposed radial spines, represent the two large groups ofAcanthonida 

 and Acanthophracta. For short and clear distinction of these two groups of 

 ACANTHARIA, we will call the Actinelida (with irregular number and disposition of 

 radial spines) Adelacantha, in opposition to the Icosacantha (Acanthonida and 

 Acanthophracta), which all possess twenty regularly disposed radial spines. 



Johannes Miiller, the great zoologist, to whom we are indebted for the first detection 

 and accurate knowledge of the Acanthometra, already recognised the regularity in the 

 peculiar disposition of their twenty radial spines (Abhandl. d. k. Akad. d. Wiss. Berlin, 

 1858, pp. 12, 37). In honour of my great master I have called this regular disposition 

 the " Miillerian law of spine disposition," and have given a full explanation of it in my 

 Monograph (1862, pp. 40-45, 371, 372). With regard to its general value for all 

 Icosacantha (Acanthonida and Acanthophracta), we might also call this pro- 

 morphological Miillerian law " the Icosacanthan law." 



In 1862 I had already given the following precise definition of this "Icosacanthan 

 law" (loc. cit., p. 40) :" Between two poles of a spineless axis are regularly 

 disposed five parallel zones, each with four radial spines ; the four spines of each zone 

 are equidistant one from another, and also equidistant from each pole ; and the 

 four spines of each zone are so alternating with those of each neighbouring zone, 

 that all twenty spines together lie in four meridian planes, which intersect one 

 another at an angle of 45." For the clear conception of this remarkable 

 Miillerian law, and for the complete understanding of its high value for the complicated 



