96 BRITISH PALAOZOIC SPONGES. 
attach the organism to the sea bottom. These bundles of anchoring spicules are 
more frequently preserved than the six-rayed spicules of the body of the Sponge, 
and they have been found in the Tremadoc rocks of Wales, and the Ordovician 
of Girvan, Ayrshire; but they are more particularly abundant in the Sponge-beds 
of the Yoredale series in Yorkshire and in the decayed chert of the corresponding 
rocks in the West of Scotland and Ireland. Entire beds of rock in Yorkshire are 
filled with them. 
The genus Dictyophyton is only represented by a small form, which appears to 
be limited to Ludlow strata near Kendal, whilst it seems to be entirely absent 
from the Devonian strata in this country. In North America, on the other hand, 
and in Belgium, this genus has a great development in the Upper Devonian. 
Plectoderma is as yet only known by a fragment of the connected skeleton, it is 
limited to the Upper Silurian of the Pentland Hills. Phormosella is likewise 
restricted to a single horizon of the Silurian in Shropshire. Like many of the 
Paleozoic hexactinellids, it appears to have been a gregarious form. 
Whilst the spicular characters of many of the Paleozoic hexactinellids correspond 
closely with those of existing Sponges of this division, there are others in which 
the spicules are greatly modified, and widely diverge from the normal type. Thus 
in the Carboniferous genus Spiractinella, Hinde, the rays are ornamented with a 
spiral ridge; and, though simple six-rayed spicules occur, in the majority the rays 
divide and subdivide, so that the extreme forms are stellate. In Holasterella, 
Carter, stellate spicules are likewise present, and the larger spicules of the skeleton 
also appear to be very irregular in form. Still more abnormal are the large 
branched and spined spicules of Acanthactinella, Hinde, from the Carboniferous 
of Ayrshire. In Amphispongia again, limited to a single horizon of Upper 
Ludlow age, and to a single locality in the Pentland Hills, the spicular structure 
strikingly differs from that of any other hexactinellid genus. In the well-marked 
family of the Receptaculitidee, one ray of the normal hexactinellid spicule is modi- 
fied into a delicate plate. 
Only detached spicules of the genus Astr@ospongia have as yet been met with 
in the Silurian and Devonian strata of Shropshire and Devonshire, though in North 
America, and in Germany, entire Sponges are occasionally preserved. 
The most striking of any of the Paleozoic Sponges are the forms which I have 
described under Tholiasterella and Asteractinella, and placed in a new sub-order, 
the Heteractinellide. Their spicules are of unusually large dimensions; the 
number of rays is variable, ranging from six to thirty, and they are disposed so as 
to form either stellate or umbrella-shaped spicules. The spicules appear to have 
been partially free and partially fused together in the skeleton. These remarkable 
forms have as yet only been found in the Carboniferous Rocks of Ayrshire. 
