24 K. KiRKFAi RKK : A New Dictyonine Sponge. [Voi,. II, 1908.] 



The dictyonal network does not fine off at the growing edges 

 into a single layer as commonly happens in Farrea^ and rarely in 

 Eurete. 



Spicules. — Scopulse of two kinds. i. With four or five 

 knobbed prongs (fig. 10), 436 /. in total length, the prongs being 

 62 jx long ; the shaft, which is slightlj^ swollen at the point of 

 origin of the pnjngs, is roughened at the upper and lower end ; 

 the lower end is blunt-pointed ; the prongs are coarsely granular ; 

 and the pyriform knobs provided with retrocedent spines; the shaft, 

 which is swollen at the cladal origin, is roughened at the upper 

 and lower ends ; the lower end rather blunt-pointed usually ; the 

 prongs, 62 /x long, are coarsely granular, and the p^^riform knobs 

 ]:)rovided with retrocedent spines. 2. vScopulae with lanceolate 

 prongs (fig. 11), total length 694 /- ; the whole surface smooth ; 

 shaft 6" 5 /' thick at centre, swollen at cladal end to 11 fx. Prongs, 

 usually five in number, smooth, lanceolate, 82 /<. in length. 



I found both kinds of scopulse on the dermal side, and was 

 unable to make out any distinction between dermal and gastral 

 scopulse. 



vSmall hexactins, \-arying a good deal in size, but with rays on 

 an average 400 to 450 /-, in length, with finely spined surface. 



Discohexasters 44*5 /, in total diameter, with smooth primary 

 rays 6*25 /<, long, each primary with four roughened curved, second- 

 ary rays 16 /• long, each ending in minute disk with finely denticu- 

 late edge. 



Affinities.— Y:\\^ new species in its general form comes nearer to 

 Eurete ere ctum, It E. Schulze [I.e., supra, p. 106), and its varieties 

 (Wilson, /.c, supra , p. 62, et seq.), than to other species of the 

 genus, but at the same time there are great differences. 



In E. ere ctum there is a tendency to form a long axial growth 

 rather than a clump of anastomosing tubes. In E. erectum var. 

 tubuliferum, Wilson, there is only a single axis with lateral branches, 

 as in the new species ; in this variety the lateral branches are at 

 first cup-like, with flaring edges, which latter in more developed 

 branches curve over and meet in such a way as to leave an orifice 

 at each end of the line of junction. 



In E. annandalei a great degree of speciaUsation has arisen , 

 the simple tube or cup expands into a hollow lamella in which the 

 coalescent edges form the margin of a leaf-like lamina. 



The upper and lower laminar orifices represent the openings 

 left by the partial fusion of the edges of the simple tubular branch. 



In spite of the high degree of specialisation attained in this 

 species by the branches and their openings, it did not seem neces- 

 sary to place the form under a new genus. In other Euretid 

 sponges we find great divergence from the anastomosing tubular 

 growth even within the limits of the same species, as in Farrca 

 occa (Bowerbank) var. laminaris, Topsent [Result. Camp. Set. 

 Monaco, fasc. xxv, p. 44, pi. vi, figs, i, 2), and F. occa var. 

 foliascens, Topsent {Bull. Mas. Oceanograph. Monaco, November 

 1906, p. I). 



