CORBITELLA AND HETEEOTELLA. 17 



Especially close seems to be the resemblance in spiculation 

 between G. pulchra and G speciosa. Under general agreement in 

 shape and structure, the two species have in common not only 

 the smooth and the spiny microxyhexactins, the graphiocome and 

 the floricome, but also similarly characterized discohexasters in 

 both the ordinary and the hexactinose form (=F. E. Schulze's 

 Discohexactin). 



So far as our knowledge goes, the following points in the 

 structure and spiculation of G. pulchra seem to be noteworthy as 

 offering probably useful data for its differential diagnosis : 



1. The body, whose length (200 mm.) docs not fall much 

 short of that of the known specimens of C. speciosa and elegans, 

 is in shape ventricose, not clavate. 



2. The upper end of the body, instead of being covered by a 

 sieve-plate, is overarched by a wreath of projecting rays belonging 

 to the principalia marginalia, This condition has been assumed 

 by F. E. Schulze ('95, p. 35) as possibly due to a partial loss of 

 a sieve-plate, such as is found in Dlctyaulus elegans F. E. Sch. 

 Now while that possibility can not be wholly set aside, it seems 

 to me equally possible that we have here to do with the same 

 perfectly natural phenomenon as the coronal wreath of Regadrella 

 komeyamai I J. ('01, p. 253). 



3. Assuming there was originally a sieve-plate, this could 

 not have passed over (to judge from Schulze's figure) so in- 

 sensibly into the lattice-work of the lateral wall as in either 

 C. speciosa or elegans. Moreover there exists a marginal ridge 

 which seems to be tolerably persistent. 



4. The parenchymal bundles, while irregularly crossing one 

 another in the greater part of the body, are arranged in regular 

 longitudinal and transverse systems near the upper end. 



