SPONGES 



49 



eurypylous Rhagon type, but the aphodal is not unknown. . The 

 Ceratosa contain all sponges with a homy skeleton, except those 

 in which the horny fibres are cored or spined with silicious spicules 

 secreted by the sponge ("proper" spicules) ; these are arbitrarily 

 assigned to the Monaxona. There is convenience in this proceed- 

 ing, for horny matter is widely disseminated throughout the Demo- 

 spongigi, occurring even in the Lithistida, and it frequently serves 

 to cement the oxeate spicules of the Monaoama into a fibre, without 

 at the same time forming a preponderant part of the skeleton. It 

 would be wellnigh impossible to say where the line should be drawn 

 between a fibre composed of spicules cemented by spongin and one 

 consisting of spongin with embedded spicules, while there is com- 

 paratively no difficulty in distinguishing between fibres containing 

 spicules and fibres devoid of them. That the distinction, however, 

 is entirely artificial is shown by the fact that, after spicules have 

 disappeared from the homy fibre, they may still persist in the 

 mesoderm ; thus Von Lendenfeld announces the discovery of micro- 

 scleres (cymba) in an Aplysillid sponge and of strongyles in a 

 Cacospongia, both horny sponges. (A form intermediate between 

 this Aplysillid and the Desmaeidonidss would appear to be Toxo- 

 chalina, Ridley.) The Ceratosa frequently enclose sand, Fora- 

 minifera, deciduous spicules of other sponges and of compound 

 Ascidians, and other foreign bodies within the horny fibres of their 

 skeleton ; they also sometimes attach this material, probably by a 

 secretion of spongin, to their outer surface, and thus invest them- 

 selves in a thick protective crust. In some Ceratosa no other 

 skeleton than that provided by foreign enclosures is present. The 

 canal- system is syconate or eurypylous in the simpler forms and 

 diplodal in the higher. The Monaxonida make their earliest ap- 

 pearance in the Silurian rocks (Climacospongia, Hinde), and are 

 now found in all seas at all depths. The only sponges inhabiting 

 freshwater belong to this group. The Tetractinellida adhere to 

 the Monaxonida at more than one point, and one of these groups 

 has probably been a fruitful parent to the other, but which is 

 offspring and which parent is still a subject for discussion. The 

 Choristida in its simplest forms presents a eurypylous Ehagon 

 system, in the higher an aphodal system. It is in this group that 

 the most highly complex cortex is met with ; in the Geodinidse, 

 for instance, it consists usually of at least five distinct layers. 

 Thus, proceeding outwards, next to the choanosome is a layer of 

 thickly felted desmachyme, passing into collenchyme on its inner 

 face ; then follows a thick stratum of sterrasters united together 

 by desmacytes ; this is succeeded by a layer of cystenchyme or 

 other tissue of variable thickness ; extemal to this is a single layer 

 of small granular cells and associated dermal asters ; and finally, 

 the surface is invested by a layer of pavement epithelium. The 

 Lithistida, like the Ceratosa, are possibly of polyphylitic origin ; 

 in one group {Tetracladina) the articulated scleres are evidently 

 modified calthrops spicules (see fig. 14 e), and associated with them 

 are free trisenes, which support the dermis and resemble precisely 

 the trisenes of the Choristida. In another group {JRhabdocrepida) 

 the scleres are moulded on a Monaxonid base (see fig. 13 q-s) ; but, 

 associated with them, trisenes sometimes occur similar to those of 

 the Tetracladina. Both these groups are in all probability derived 

 from the Choristida, and a distinct passage can be traced from the 

 Tetraoladose to the Khabdocrepid group. In the Ehabdocrepida 

 we find forms without trisenes ; these may possibly be degenerate 

 forms. The third group of Lithistids is derived from the Ehabdo- 

 crepida, the Anomocladine desma being derivable from the Khabdo- 

 crepid by a shortening of the main axis into a centrum. The 

 thick centrum, from which the arms, variable in number, ori- 

 ginate, is hollowed out by a cavity, which appears during life to 

 have been occupied by a large nucleus, like that of a scleroblast, 

 and it is quite conceivable th'at the scleroblast, which in the 

 Tetracladine Lithistids lies in an angle between the arms, may 

 have become enclosed in an overgrowth of silica, from which addi- 

 tional arms were produced. ^ The constancy with which spicules 

 in other sponges maintain their independence is very striking. 

 When once a persistent character like this is disturbed, excessive 

 variability may be predicted, as in the Anomocladine scleres. 

 Classifi- The classification of the sponges into families is shown in the 

 cation in following scheme. 

 *»™'lies. Class CALCAREA. 



Order 1. HOMOCCELA, Pol. 

 Family 1. AscONiDiE, Hk. — Homoccela which are simple or com- 

 posite, but never develop radial tubes. Examples : Ascetta, Hk. 

 (fig. 1) ; Leucosolenia, Bwk. 



Family 2. HoMODEEMiDiE, Lfd. — Eomoccela with radial tubes. 

 Example : Homoderma, Lfd. (figs. 3, 4). 



Order 2. HBTEBOOCELA, Pol. 

 Tribe a. tSYCONARiA.^ 

 The flagellated chambers are either radial tubes or cylindrical 



Family 1. SYCONiDai. — The radial tubes open directly into the 

 paragastrio cavity. 



Sub-family a. Syconina. — The radial tubes are free for their whole 

 length, or at least distally. Examples : Sycetta, Hk. ; Sycon, O.S. 



Sub-family i. Uteina, Lfd. ^- The radial tubes are simple and 

 entirely united. The ectosome is differentiated from the choanosome 

 and sometimes develops into a cortex. Examples : Grantissa, Lfd. ; 

 Ute, O.S. (fig. 5); Syeortiisa, Hk.; Amphoriscus, Pol. 



Sub-family c. Grautina, Lfd. — The radial tubes are branched. 

 The incurrent canal system is consequently complicated. An ecto- 

 some is present. Examples : Grantia, Fl. ; Heteropegma, Pol. (fig. 

 4) ; Anamaxilla, Pol. 



Family 2. Sylleibid^;, Lfd. — The choanosome is folded. The 

 flagellated chambers (which are partly rhagose in Vosmaeria) 

 communicate with the paragasti'ic cavity by excurrent canals. 

 Examples : Polejna, Lfd. (fig. 6) ; Vosmaeria, Lfd. 



Family 3. TEiCHONBLLlDiE, Carter. — Composite Sylleibidx with 

 the oscules and pores occurring on different parts of the surface. 

 Example : Teichonella, Crtr. 



Tribe b. tLEUcoNARiA. 



The canal system belongs to the eurypylous Ehagon type. 



Family LXEUOONiDiB, Hk. — The outer surface is not differentiated 

 into osculiferous and poriferous areas. Examples : Leucetia, Hk. ; 

 Leucaltis, Hk. ; Zeucortis, Hk. 



Family 2. Eilhakdid^, Pol. — Composite Leuconaria, with the 

 outer surface differentiated into special osculiferous and poriferous 

 areas. Example : Eilhardia, Pol. 



The arrangement adopted above is founded on Von Lendenfeld's 

 revision (//) of the classification propounded by Polejaeff (j6), who 

 in a masterly survey has thrown an unexpected light on the struc- 

 ture and inter-relationships of a group which Haeckel has rendered 

 famous. It should not be overlooked that Vosmaer {31) had pre- 

 viously explained the structure of the Leucones. However errone- 

 ous in detail, Haeckel's views are confirmed in their broad outlines, 

 and it was with true insight that he pronounced the Calcarea to 

 offer one of the most luminous expositions of the evolutional theory. 

 In this single group the development in general of the canal system 

 of the sponges is revealed from its starting-point in the simple 

 Ascon to its almost completed stage in the Leucon, with a complete- 

 ness that leaves little further to be hoped for, unless it be the re- 

 quisite physiological explanation. 



Class MYXOSPONGIu^. 

 Order 1. HALISAECDfA. 

 Family 1. HALisAECiDiE, Lfd. — The flagellated chambers are 

 syconate. Examples: Ealisarca, Duj. (with branched chambers) ; 

 Bajalns, Lfd. (with simple chambers). 



Family 2. Oscaeehid^, Lfd. — The flagellated chambers are 

 eurypylous and rhagose. Example : OscareUa, Vosm. 



Order 2. CHONDKOSINA. 

 Family 1. CHONDKOSliDiE. — With the characters of the order. 

 Example : Chondrosia, 0. S. 



Class SILICISPONGIM. 



Sub-class I. HEXACTINELLIDA. 



Order 1. tLYSSACINA. 



Family 1. EtrPLEOTELLlD.a). — The spicules of the dermal mem- 

 brane are "daggers" (fig. 15 a). Examples : Eupkctella, Owen 

 Solascus, E. Sch. ; Sabrodictyum, W.T. 



Family 2. AscONEMATiDiE. — The dermal spicules are " pinnuli ' 

 (fig. 15 b, c). Examples: Asconema, S. Kent; Sympagella, O.S. 

 CaulophsBTis, Schulze. 



Family 3. HYALONEMATira;. — The dermal spicules are pinnuli 

 and amphidisks (fig. 15 d). Example : Hyalonema, Gray. 



Family 4. fEossELiD^. — The dermal spicules are gomphi, stauri 

 (fig. 15/), and oxeas. Examples: Mossella, Crtr.; Craleromorpha, 

 Gray ; Aulochona, E. Sch. 



Family 5. *RE0EPTAOirLiDa;, Hinde. — The distal ray of the 

 dermal spicules is expanded horizontally into a polygonal plate. 

 Example : *Beceptacv,lUes, Defr. 



Order 2. tDIOTYONINA. 

 Sub-order 1. UNCINITARIA. 

 Uncinate spicules are present. 



Tribe as. Clavularia. 

 Clavulse (fig. 16 c) are present. 



Family 1. FAKEEiDiB. — Characters those of the tribe. Example : 

 Farrea, Bwk. 



Tribe b. Scopulaeia. 

 The dermal spicules are scopularise (fig. 16 b). 

 Family 1. tEuREXiDiE. — Branched anastomosing tubes, or goblet- 

 shaped, with lateral outlets. Examples : Eurete, Marshall ; Peri- 

 phragella, Marshall ; Lefroyella, Schulze. 



Family 2. tMELLiTTONlD^. — Tubular or goblet-shaped, with 

 honeycomb-like walls. Example : Aphrocallistes, Gray. 



1 An • indicates that the group is only known in the fossil state, a t that it 

 is both recent and fossil. 



