SPONGES. 65 



characters, the " Micro trieenosa " are what MINCHIN terms " a heterogeneous collection 

 of sponges of divers affinities," and the " Oligosilicina " (" Chondrosidse," of SCHUL/,K) 

 are simple only by the reduction of the skeleton and not primitively. I therefore 

 consider that TOPSENT'S revived "Carnosa" should be again abandoned, the 

 " Microtrisenosa " placed amongst the other trisene-bearing Tetractinellida, and the 

 " Oligosilicina " placed near the Tethyidse, to which they are evidently allied by 

 their corticate character, the general arrangement of their canal-system, and the 

 astrose microscleres of Chondrilla. 



The Homosclerophora are thus left as the natural starting point of the Tetraxonid 

 series, a position already clearly indicated for them by LENDENFELD in his work on 

 the Tetractinellida of the Adriatic. TOPSENT, it is true, fully recognises the artificial 

 character of his "Carnosa" and places them after his "Tetractinellida." MINCHIN, 

 on the other hand, places them first, a diversity of opinion which clearly indicates 

 the impracticability of associating together such widely different forms as are included 

 in the group. 



Owing to their primitive character, the Homosclerophora form a very important 

 and interesting sub-order, and it is here that we must seek for the origin, not only 

 of the various types of canal-system, but also of the almost innumerable types of 

 both micro- and megascleres met with amongst the Tetraxonida. 



FAMILY: PLAKINID^E. 

 Homosclerophora in which no distinct cortex is developed. 



This appears to be the most primitive family of the Tetraxonida. The genus 

 Plakinastrella, formerly included herein, but possessing more or less well-developed 

 trirenes, finds a more natural position amongst the Pachastrellidae. 



Dercitopsis, n. gen. 



Plakinidye with calthrops, triods and smooth oxea, but no candelabra. The oxea vary 

 greatly in size, and some of the smaller ones form a special dermal layer, in 

 which they are commonly arranged at right angles to the surface. 



It is, perhaps, doubtful whether this genus ought to be separated from Plakortis, 

 but the species for which it is founded differs from the type species of Plakortis 

 (P. simplex) in three respects: (l) Calthrops are present as well as triods; 

 (2) some of the oxea are so large as to deserve the name of megascleres, although 

 they pass gradually into the smaller forms (microxea) ; (3) some of the microxea 

 form a special dermal layer, in which they are commonly arranged at right angles to 

 the surface. 



In this genus we have a refutation of SOLLAS'S statement that " megascleres are 

 absent" from the so-called " Microsclerophora " (including the PlakinidEe), and we 



