152 



PROF. ST. GEORGE MIVART ON 



sively investing the innermost but extracapsular shell from 

 which such radii start. This condition is seen in AracJmosphcera 

 myrlacantlia, and has been already referred to as occupying cir- 

 cumferential structures depending upon radial parts. 



The I'adial parts hitherto referred to are all extracapsular ; but 

 radii may proceed from an intracapsular spheroidal shell, as in 

 Heliodiscus'^ and Actinomma, and also in jRhaphidococcios'f . 



They may so radiate and at the same time assume the most 

 beautiful arborescent structure, which may be solid, as in Olado- 

 coccusX, or hollow, as in Gcelodendrum. 



Fig. 8. 



Xiphacantha obtained during the ' Challenger ' Expedition §. 

 (After Wyv. Thomson.) 



The truly radiating structure ^«r excellence is found, however, 

 in AcantJiometra and the allied genera, where we often find radii 

 only, with no vestige of a circumferential skeleton. The radii 



* L. c. pl.xvii. figs. 5-7. t L. c. pi. xiii. fig. 5. | L. c. pi. xiv. fig. 6. 

 § ' Vovagp of Challenger,' p. 2^r>, fig. ftS. 



