RECENT BESEAECHES ON THE RADIOLAEIA. 



155 



chambers on all sides) winding spirally round an axis placed at 

 right angles to the plane of the original median partition, which 

 we have ideally enlarged. 



Another mode of conceiving the structure of this shell is to first 

 conceive a central hollow die with successive series of hollow dice 

 placed, one set outside the preceding, all round it so as to form 

 one whole. Secondly, to imagine these dice to have their surfaces 

 cut in curves so that they may be mutually adjusted, and thus 

 form a sphere of successive concentric layers of hollow dice. Then 

 to imagine each die to have a perforation made in each face (six, 

 of course, to each), the perforations of adjoining dice correspond- 

 ing in position so as to place their hollow interiors in communi- 

 cation — successive concentric layers of hollow communicating 

 chambers being thus produced. Finally, we must imagine a slight 

 twist to be given to the complex whole, so that instead of con- 

 centric layers we may have a spiral arrangement of the chambers 

 and of their perforated separating walls, which would thus have 

 come to wind round a single longitudinal axis. 



We come now to that system of confused and confounded ra- 

 diating and tangential parts which together form the spongy kind 

 of skeleton. 



The entire skeleton may consist of this kind of structure ex- 

 clusively and be discoidal, as in Spongodiscus *, or cylindrical with 

 radiating spines attached, as in Spongurus f. 



The spongy tissue may include a concentric shell like that of 

 Trematodiscus, as is the case in Spo7igocyclia or surround two 

 or three concentric shells, as in Dictyosphagma § and Spongodic- 

 tyum !| (or Bictyosoma). 



But not only concentric spheres, but also strong radii proceed- 

 ing from their exterior may coexist with a spongy mass, as in 

 Spongosplicera ^. 



Finally, a spongy mass may lie entirely outside the capsule and 

 be connected with two spheroidal intracapsular concentric shells 

 by means of radial spines, as is the case in Bhizosphcsra^^ . 



Such are the main forms presented by the skeletons of the 

 Eadiolaria which are as yet known. 



Before leaving this portion of my subject, I may be permitted 

 to remark that to my mind it seems evident that these beautiful, 



* L. c. pi. xii. figs. 14 & 15. t Z. c. pi. xxvii. fig. 1. 



X L. c. pi. xxviii. fig. 2. § Abh. k. Ak. Berlin, 1858, pi. ii. figs. 9-11. 



II ' Radiolarien,' pi. xxvi. figs. 4-6. 



^ L. c. pi. xxvi. figs. 1-3. L. c. pi. XXV. figs. 1-10. 



