236 M. K. A. Zittel on Fossil LitMstidce. 



in very close connexion. Thus, among the five groups be- 

 longing to it we find, on the one hand, the Renierida, Suberi- 

 tida, and Potamospongida {Sporigilla) , with uniaxial spicules; 

 and on the other, the Pachytragida and Pachastrellida, with 

 tri- or quadriradiate siliceous elements. Under the Pachas- 

 trellida the family Lithistina occurs as a section of the third 

 rank. 



I confine myself to the mention of these three most recent 

 classifications of the sponges, as I have already endeavoured 

 to show how little warrant there is for the older opinion of 

 Bowerbank, Gray, and Wyville Thomson, that the Lithistid^e 

 and Hexactinellidge should be brought close together. 



In one point Oscar Schmidt, Glaus, and Carter agree : they 

 all place the Lithistidaj close to the Geodinidse and Ancori- 

 nidae (Pachytragidse, Carter). But whilst Claus assigns to 

 them the rank of a distinct order, they appear only as a family 

 of an order in Oscar Schmidt's arrangement, and by Carter 

 they are even degraded into a subsection (family) of the 

 Pachastrellidffi. 



The agreement of the Lithistidse with the above-mentioned 

 sponges consists in that the anchor-shaped surface-spicules of 

 many Lithistid genera are deceptively like certain skeleton- 

 spicules of the Ancorinidee and Geodinidas. With this, how- 

 ever, we have exhausted all that can be said in favour of the 

 affinity of these latter sponges to the Lithistidee. But if we 

 consider tliat in the Lithistidai neither typical quadriradiates 

 (as in Steletta), nor octoradiatcs, nor siliceous stellules, or 

 radiating siliceous spherules and siliceous disks are observed, 

 it is clear that even in the free siliceous structures there is a 

 considerable difference. This, however, becomes positively 

 fundamental so soon as we take into consideration the true 

 skeletal elements. No other order of sponges at present known 

 possesses similar composite and multifariously branched sili- 

 ceous corpuscles. Although a quadriradiate axial cross lies 

 at the foundation of the Tetracladina, there nevertheless 

 exists a profound difference between the quadriradiate stars of 

 the Pachytragida3, in which the individual arms are straight 

 and pointed, and the Lithistid corpuscles, which are more or 

 less ramified at the ends. Moreover, as regards the peculiar 

 union of the latter to form a generally intimately interlocking 

 tissue, and the stony constitution of the whole sponge-body, 

 we may perhaps find a certain analogy with this in the Hex- 

 actinellidaj, but certainly not in the other siliceous sponges. 

 Finally, if we take into account the complicated canal-system 

 and the external ajipearance of the Lithistida;, it is again only 

 the Hexactinellid{\2 among the siliceous sponges, and an ex- 



