MATERIALS FOR A MONOGRAPH OF THE ASCONS. 3840 
leaving interspaces to which an equiangular triradiate, not a 
cruciform spicule, would be the natural adaptation. Hence 
I consider that the constant rectangular junctions of the rays 
of Hexactinellid sponges cannot be regarded as an adapta- 
tion, but as an inherent peculiarity of the spicule itself, 
probably determined by the physical, if not the crystalline, 
properties of the material. 
Strongly contrasting with the Hexactinellid spicules are 
those in the third sponge-stem, the heterogeneous assem- 
blage united under the comprehensive name Demospongix 
by Sollas. Here the primitive spicule is apparently a 
tetraxon, or possibly an aster, of which the tetraxon is but 
one modification.! From this arises every possible type of 
form. Dendy has recently * constructed a phylogeny of the 
spicules of Demospongiz. The only point which it is neces- 
sary for me to dwell upon here is, that in the Demospongiz 
the angles at which the spicule-rays meet are infinitely 
variable, and show not.the shghtest tendency to be constant 
in any way. It is sufficient to mention the modifications of 
the trizne, distinguished by a formidable nomenclature 
(protriznes, orthotriznes, anatriznes, etc.). It is evident 
that whatever may be the case in Calcarea and Hexacti- 
nellida, the physical nature of the spicule material in Demo- 
spongiee offers no obstacle to the indefinite variability of the 
spicules. I see nothing against accepting Schulze’s theory 
(1887, p. 503) that the primitive tetraxon of Demospongiz 
arose as an adaptation to the form of the interspaces between 
numerous closely-packed spherical chambers. Comparison 
of existing forms, such as Plakina and Oscarella, make 
it highly probable that in this stem forms with chambers 
preceded forms with spicules in evolution. 
It may be a stumbling-block to many that in Calcarea and 
Hexactinellida physical conditions should be regarded as a 
1 In this connection the statement of Maas (1900 [8]), that the asters of 
Tethya arise by fusion of separate tetraxons, must not be overlooked; but 
compare Woodland (1908). 
2 «British Association Reports,’ York, 1906. 
