KEPOET ON THE TETEACTINELLIDA. 
213 
present (PI. XXVIII. figs. 23, 24). They measure from about O'OllS to 0'0434 mm. 
in diameter, and lie in cavities a little larger than themselves, 0'0197 and 0'0474 mm. 
in diameter. A concentric or spiral arrangement of interrupted lines is to be observed 
in them ; this in some parts of the section is exchanged for a regularly dotted appear- 
ance. In very thin sections, especially in such as happen to be accidentally torn, these 
appearances are found to be produced by minute bacillar bodies, of which the whole struc- 
ture is built up. It would appear very doubtful whether these structures are part of the 
sponge ; they might with much more probability be regarded as of a Bacterial nature. 
The characters of the canal system have been partly described incidentally ; the 
flagellated chambers are small, about 0'0237 mm. in breadth by 0'0197 mm. in length, 
the apopyle and prosopyle vary greatly in diameter, from about O'OOG to O’OIG mm., the 
prosodus is always short, often absent, and the aphodus is usually also short compared 
to the length it attains in many other Geodine sponges. 
The Skeleton . — The oxeate spicules are scattered singly and apparently at random 
through the choanosome, except near the cortex, towards which they are directed at right 
angles, seldom entering it. 
Orthotrisenes are absent, except quite close to the cortex (PI. XXVIII. figs 17, 18) 
and in the neighbourhood of the cloacal wall, which appears to be a modified cortex. 
The cladomes of these trisenes lie in the inner fibrous layer of the cortex, frequently 
centrally situated between three or four surrounding chones, so that the cladi extend 
between adjacent pairs of these latter, often lying tangential to their walls. 
The sterrasters are scattered in various stages of development in the choanosome, 
and when adult are arranged in the cortex in the manner already described. They may 
be regarded as modifications of the usual more or less spherical sterraster, resulting from 
the abbreviation of all the actines lying in an equatorial zone including the hilum, and the 
extension of those lying along and adjacent to the zone-axis. This would be the natural 
result of tension acting in the direction of the zone-axis. 
The youngest sterraster met with resembles that of Geodia in presenting a very small 
centrum and long trichite-like actines ; the difference in the length of the actines had 
already, however, in the instance observed been well established (PI. XXVIII. fig. 30). 
This spicule was seen still embedded in its scleroblast, of which, however, the nucleus was 
not visible. A nucleus, however, was observed of the usual scleroblastic type in connec- 
tion with a more advanced example, it was situated over that part of the sterraster which 
remains as the non-granulated area of the completed form (PI. XXVIII. fig. 31), thus 
proving the homology of that area with the hilum of the more common form of 
sterraster. In the stage immediately preceding that of completion the sterraster is 
smooth, and frequently subdivided at the ends (PL XXVIII. fig. 12); the small spines 
which granulate the surface commence at the extremities and subsequently extend 
towards the centre. 
