REPOET ON THE TETEACTINELLIDA. 
cxiii 
not to be distinguished from a species of Stelletta. But there is another family, also 
possessing sterrasters, the Placospongidae, which differs widely from the Geodiidae, partly 
in the character of the external openings of the canal-system, and still more in the total 
absence of triaene or tetraxon spicules ; the sterraster, however, is such a highly peculiar 
form, known in no other group outside the Geodiidse, that its presence in this case may 
be taken as allying the Placospongidse with the Geodiidse ; for the two families thus 
united we find the common name Sterrastrosa. 
The Sterrastrosa, Euastrosa, and Streptastrosa are all united together by the common 
possession of some form of aster, and, as a rule, they possess as well a second distinctive 
aster, they may, therefore, be united together in a single order, the Astrophora, as 
opposed to the Sigmatophora. 
The remaining families are the Placinidse, Corticidse, and Thrombidse ; in all three 
there is an entire absence of megascleres, and in the first two the microscleres are 
variously modified tetractinose asters, in the last the microsclere resembles a minute 
trisene, and in one of its species is accompanied by a still smaller microsclere, which 
bears much the same relation to the larger form that the microscleres in the Astrophora 
and Sigmatophora do to the megascleres ; this second form of microsclere, which is one 
of the minutest of all known spicules, is an amphiaster, the straight, slender axis being 
terminated by four minute recurved spines at each end ; its form is quite peculiar and 
its origin obscure, so that it throws no light on the relationship of the family; as micro- 
scleres of only one order are present in the other two families, they will not help us 
much in an investigation into the relative constancy of characters, but it may be 
pointed out that, while the Corticidse differ from the Placinidse widely so far as the 
character of the mesoderm and type of chamber-system are concerned, they are 
evidently closely united together by the characters of the spicules, which in this case as 
in others are found to be-more persistent than those of the soft parts. The character by 
which all three families are united into a single group is the presence in all of small 
spicules, which are all derivable from a tetractinose aster; thus in the Placinidse the aster 
may assume the character of a candelabrum or of a trisene, the trisene in one species, 
Placina trilopha, resembling that of the Thrombidse, except that it is not spinose ; in 
the Corticidse the candelabrum has been preserved as the characteristic form, in the 
Thrombidse the trisene. In none of these Sponges do the spicules occur aggregated into 
fibres, and while the Placinidse are evidently a primitive group from which the other two 
have been derived, these last agree together in possessing a diplodal chamber-system and 
a collenchymatous mesoderm ; the three families may, therefore, be classed together as a 
single group, forming a third suborder, the Microsclerophora. 
In searching for relatively constant characters we have been able to unfold at the 
same time the nature of our classification, and we have arrived at the result that such a 
phenomenon as an absolutely constant character does not exist ; Sponges like other 
(zooL. CHALL. EXP. — PART Lxiii. — 1888 .) Rrrp 
