212 Mr. R. Kirkpatrick on Hexactinellid 



of centrifugal and centripetal pressure. No transitional forms 

 are known. There is no fixed intermediate stage in an 

 umbrella when it is held up in a high wind, and disaster (or 

 the pointing of ribs in an opposite direction) is imminent. 

 We get amphidisks in Monaxonellid sponges, especially in 

 Spongillidse, where these spicules may be definitely orientated 

 in the gemmule coats, or scattered, and the spines of the disks 

 may be at right angles to the shaft or bent down. Here the 

 history is entirely different, and the resemblance with 

 Hexactinellid amphidisks furnishes an instance of convergence. 

 The effects of pressure on Hexactinellid spicules are abun- 

 dantly in evidence. The subfamily Acanthascinse is appa- 

 rently an example of the effects of centripetal pressure leading 

 to the formation of discoctasters, and suppressing more or less 

 completely the actines of all the hexasters. In Anoxy calyx 

 ijimai, K., one sees many hexasters with a bayonet kink at 

 the base of the end-spines. 



(6) Discoctasters are, as Schulze has shown, simply hex- 

 asters with the actines suppressed and the end-spines fused 

 with neighbouring end-spines to form a new combination of 

 eight groups of end-spines. The six axial canals of the 

 atrophied actines are still visible in the centre of the spicule, 

 as Ijima has pointed out. There is no difficulty about the 

 position of these spicules in the morphological scheme sketched 

 out above. The central node with its cross of six axial canals 

 represents the six actines, and the eight secondary so-called 

 rays which are without axial canals are the fused end-spines. 



Turning aside from the morphological classification, it 

 is very interesting to note how the morphological groups of 

 spicules are distributed among Hexactinellida. Megascleres 

 (A), i. e. macrohexactins and, I may now say, their deriva- 

 tives, are found in both suborders ; microhexactins (B) only 

 in Amphidiscophora; centripetal hexasters and their deriva- 

 tives (C) only in Amphidiscophora; and centrifugal hex- 

 asters (D) only in Hexasterophora. Amphidiscophora has 

 groups A, B, G ; Hexasterophora has groups A, D. 



The association of B with C and the non-association of B 

 with D are deeply significant facts, and are probably due to a 

 common cause. I have suggested (Ann. & Mag. Nat. Hist. 

 (8) iv. 1909, p. 481) that the existence of the amphidisks 

 (diasters or reduced hexasters with centripetal end-spines) is 

 due to the gathering of the trabecular tissue into strong- 

 concentric laminse at the dermal, gastral, and canalar surfaces, 

 and that the effect of the pressure on the end-spines was to 

 give them a shaft ward bend. Possibly the pressure of con- 

 tractile surface lamina? may have prevented the development 



