EEPORT ON THE RADIOLAETA. 1729 



bivalved mantle. The central capsule is so enclosed between the two inner valves, 

 that its three openings lie in the open frontal fissure between them. 



The family C ce 1 o d e n d r i d a differs from the preceding Concharida (its probable 

 ancestral group) in the development of a conical galea or pyramidal cupola on the 

 apical pole of each valve, and of three or more hollow radial tubes arising from each 

 galea. They do not possess, however, the peculiar sagittal nasal tube or rhinocanna, 

 which is constantly developed from the base of each cupola (and connected with its 

 apex by a frenulum) in the following family, the Ccelographida. These latter differ 

 also from the former in the constant possession of prominent verticillate styles. 



The family Coelodendrida was founded in 1862 in my Monograph (p. 360) and 

 represented hitherto only by two species of the genus Ccelodendrum, there described 

 (p. 361, Taf. xiii. figs. 1-3, and Taf. xxxii. fig. 1). This first description, however, 

 contained some errors, which were afterwards (in 1879) corrected by Richard Hertwig ; 

 this author also gave the first accurate description of the central capsule and its three 

 openings. In the rich collection of the Challenger, the Coelodendrida are represented 

 by four genera, but only seventeen species, some of which, however, are cosmopolitan 

 and very common, particularly Coelodendrum. 



The two valves of the lattice-shell, dorsal and ventral, are either hemispherical, or 

 somewhat more flatly vaulted or cap-shaped. They are never connected in the equatorial 

 zone of the body, as I supposed in my Monograph (1862, loc. cit.) ; but they are separated 

 by the girdle-fissure, a free circular equatorial interval, in which lie the three openings 

 of the enclosed central capsule. Though the two valves, therefore, have no direct 

 connection, they are, however, always opposed so accurately, that their equal free 

 circular edges correspond exactly one to the other, so that the apex of each valve lies in 

 one pole of the sagittal axis. From this apex there arises on each valve an irregular 

 conical or three-sided pyramidal cupola, the galea (PL 121, figs. 3, 4, 8). The 

 Coelodendrida differ in the possession of this galea from the Concharida, and agree with 

 the Ccelographida ; but they never exhibit the peculiar rhinocanna or nasal tube, which 

 arises from each galea in the latter family. 



The siliceous lattice-plate of the two valves, and of the galea arising from them, is 

 very thin and fragile, and its irregular roundish pores are extremely variable in size, 

 number, and disposition. Sometimes the pores are so small and so scarce, that the plate 

 appears nearly solid. At other times the siliceous plate seems to be really solid, and 

 covered by a network of thin crests, the small dimjDles between which give to it the appear- 

 ance of being fenestrated. Often the pores or the dimj)les are wanting in the central part 

 of each valve, while they are very numerous and dense in the peripheral part. The same 

 may be said of the lattice-plate of the galea, which is sometimes nearly solid, at other 

 times richly fenestrated. The Ccelodendi'ida agree in this structure with the following 



(ZOOL. CHALL. EXP. — PAET XL. — 1886.) Er 217 



