R. W. H. ROW—REPORT ON THE SPONGES : NON-CALCGAREA. 333 
The smaller anisochele are not arranged in rosettes and are much more 
typicalin shape. They measure (:026 mm. in length. 
YEN The palms of the larger end are well developed, 
especially the median one, but they are very small 
indeed at the small end. 
i) j (6) Trichodragmata. 
We 
The trichodragmata occur in enormous numbers 
throughout the sponge; they measure 0:065 mm. 
to 0°08 mm. in length. ‘They are composed of 
numerous very slender raphides. 
‘ c) Raphides. 
Ah (c) Ray 
Hxtremely slender raphides also occur in large 
Fig. 15.—Esperella dendyi sys ) 
oes a ee ~~" quantities through the sponge. They are exactly 
Anisochele, x 450. : 
like the individual raphides of the trichodragmata. 
Locality. No definite locality is given, but the bottle bears a label “ Sponge 
Itrmnaliss he 
Distribution. Red Sea. 
EXSPERELLA EUPLECTELLIOIDES, n. sp. (PI. 3%. fig. 12 ; Text-fig. 16.) 
A single specimen of this remarkable sponge occurs in the collection. As 
the name indicates, the external form of the sponge is very similar to that of 
the Hexactinellid genus Kuplectella. 
The sponge is attached to its support by a small base measuring about 
35 mm. X 22 mm. in extent, but the actual area of attachment consists of an 
irregular surface, which appears to have been situated vertically and which 
does not entirely cover the base of the sponge. From this base the sponge 
grows erect, the lower part being curved abruptly to bring the sponge into 
this position. 
The body of the sponge consists of a hollow cylindrical tube, 200 mm. in 
height and about 100 mm. in diameter at the widest part, which occurs about 
halfway up. The cylinder is widely open at the top, the aperture being 
slightly oval and 75 mm. across. 
The wall of the eylinder consists of a coarse and very irregular reticulation 
of skeletal fibre, covered with sarcode. ‘This fibre consists of a core of 
spicules coated with spongin. The diameter of the fibres varies very much, 
some of them being very delicate, while the largest fibres, which occur towards 
the base of the sponge, reach 1:0 mm. in diameter, or possibly even more. 
The fibre reticulation is very irregular, and large and small fibres are mixed 
up together in such a way that it is quite impossible to call some primary and 
others secondary fibres. In fact, there does not seem to be any distinction 
between the fibres, save that of size. In the same way the meshwork made by 
