CALCAREA. 47 



l)a.sal ray, the paired rays are seen to be folded to oue side ; the fold appears in some 

 cases to occur sharply near the base of the rays, which are otherwise straight ; in other 

 cases the fold is gradual, the rays being curved for the greater part of their length. 

 The apical ray springs from the junction of the facial rays on the side to which they 

 are folded, but is immediately bent down nearly into line with the basal ray. Viewed 

 from the side the apical ray appears as a continuation of the basal ray, with a sharp 

 "set-over" at the jun(;tion! The amount of the set-over varies widely (see Sireptocomcs 

 austmlis and 3fcgapogon cruciferus). The axis of the apical ray is usually slightly 

 inclined to the axis of the basal ray, sloping towards the side to which the paired rays 

 are folded. In facial view the chiactine does not difler much from the ordinary 

 quadriradiate, since the apical rays of the latter are usually curved orally, but in side 

 view tlie ditierence is wide. 



The ordinary position of quadriradiates in calcareous sponges is tangential to the 

 gastral surface, witli the l)asal ray downwards and the apical ray projecting into the 

 gastral cavity, so that the paired rays lie circumferentially embracing the gastral 

 cavity. In Amphoriscus and Syculmis there are quadriradiates which lie tangentially 

 to the gastral surface with the apical ray pointing radially outwards, and in all the 

 Ainphoriscidae there are quadriradiates which lie tangentially to the dermal surface with 

 the apical ray pointing radially inwards. The chiactines lie in a plane perpendicular 

 to all these positions. Their basal ray is directed radially outwards, the paired rays 

 lie in the gastral cortex and the apical ray points radially inwards and projects into 

 the gastral cavity. The plane of the facial rays is therefore perpendicular to the 

 gastral and dermal surfaces. The po.sition of the paired rays is no longer fixed ; they 

 may lie circumferentially as before, but they may turn round the axes of the basal ray 

 to any extent without coming out of the gastral cortex, and in fact they are found 

 lying in all positions between circumferential and longitudinal. 



The relation between the chiactines and the ordinary quadriradiates is an 

 interesting pro])lcm. All the nine new species containing chiactines have ordinary 

 quadriradiates lying tangentially round the oscule, and also lining the whole oscular 

 collar, if any. In six * of the new species the ordinary (juadriradiates cease 

 entirely at the Ijase of the collar and are replaced l)y the chiactines which form 

 the body skeleton. At the junction between the oscular collar and the body-wall 

 there is a short space in which quadriradiate spicules are found in every intermediate 

 position Itetween tangential and centrifugal. This fact suggests most forcibly that the 

 spicules turn round. The top of the body appears to be the most recently developed, 

 and the spicules formed in the oscular collar might be supposed to be turned round 

 by the development of flagellated chambers under their basal rays. A very similar 

 tipping up of dermal triradijites, due to the growth under them of the flagellated 

 cliambers, occurs at the base of the oscular coHar in TentJireiiodes aufarcficiis. There is 



* In three of the new species the ordinary quadriradiates in the oscular collar continue throughout the whole 

 gastral cavitj- in conjunction with the chiactines in the body. 



VOL. IV. 2 



