CORBITELLA AND HETEROTELLA. 27 



(PL, figs. 20-23), which constitutes the most characteristic kind 

 of spicules in this genus and species. In my opinion it is directly 

 derived and not far removed from the spiny microxyhexactin that 

 we have seen in Corbitella speciosa, G. elegans, etc., and at the 

 same time it may represent to a degree a stage in the transforma- 

 tion of a spiny oxyhexactin into an oxyhexaster. The spicule is 

 the one that has been figured by Bowerbank ('58, PI. xxv, 

 fig. 38 ; '64, PL vnr, fig. 188) under the name of " bifurcated 

 rectangulated hexradiate spicule." Apparently the identical spicule 

 has also been figured by W. Thomson (I. c, fig. 2 a) and alleged 

 by him to belong to Corbitella speciosa. 



The spicule in question occurs in the parenchyma of both 

 Specimens A and B, not very abundantly but at intervals in 

 rather scattered distribution. Its size is somewhat variable : axial 

 length 64-100 !>■ in Specimen A ; 70-136 />- in Specimen B. In 

 the center is a small spherical node, whence arise the fine, 

 gradually tapering rays, not more than 2 /* thick at base. Some 

 of the six rays may be perfectly simple, running out to a fine 

 point, as in a smooth microxyhexactin (see PL, figs. 22 and 23). 

 But more generally the rays are armed with tolerably long, 

 slender and branch-like spines, which are however never numerous. 

 The usual number is 1-3, at most 4, to a ray. The spines arise 

 at quite indefinite points in the space from about the middle to 

 the outer end of the ray, and are generally directed obliquely 

 outwards, though cases of a retroverted or of a vertically out- 

 standing spine were sometimes met with. After taking origin 

 they are either nearly straight or curved one way or the other, 

 always running out to an exceedingly fine point. 



When a spine or spines spring out laterally from a ray that 

 keeps up a tolerably straight course to the end, as is occasionally 



