426 COLLECTIONS TKOSf MELANESIA. 



and the fibre appears to agree with Ehlers's account of the species, 

 but I do not find the acuate spicules mentioned by him as occurring 

 less abundantly than the acerates ; the size of the latter in the pre- 

 sent specimens is -15 to -16 by -0055 to -007 millim. (Ehlers gives 

 •17 millim. for the length) ; and I find (what Ehlers does not mention) 

 fine bihamates measuring -02 by -001 to -0016 miUim., But a more 

 extraordinary fact connected with the species is that the sponge- 

 tissue is almost entirely replaced (this seems to be the true expla- 

 nation of the facts) by a ramifying and anastomosing algal fibre, 

 •1 to "18 millim in diameter, of a semitransparent appearance and- 

 tough elastic texture in the dry state, like that of dry isinglass ; the 

 component cells are about "007 millim. in their smallest diameter. 

 It appears to be the same species as that which forms the sub- 

 stratum of the mass described byBowerbank (P. Z. S. 1876, p. 771, 

 pi. Ixxx.) as OpMitaspongia fucoides, which is nothing more than 

 a coating Suberitid Sponge running over the fibrous filaments of 

 this same alga, which Bo,werba,nk has taken, though not without 

 hesitation (see p. 772, I. c.) for the horny fibre of an Ophliti- 

 spongia, although he has identified isolated portions as alga. In 

 this case also it is not until examined with the microscope that the 

 algal nature of most of the structure is identified with certainty. 

 This form of symbiosis has been lately noticed by Prof. K. Semper 

 in ' Die natiirlichen Existenz-Bediugungen ' (' Animal Life,' Inter- 

 national Scientific Series), where <Spo«piia cartilaginea, Esper, is used 

 in illustration ; it is probably of not uncommon occurrence in the 

 Spongida. A Formosa specimen agrees closely in the characters 

 both of the sponge and alga with those from Australia. Mr. Carter 

 (Ann. &Mag. Nat. Hist. 1878, ii. p. 163) describes exactly the same 

 circumstance in an allied species from Hong Kong, and adds other 

 similar instances. Several specimens, dry and in spirit. 



Hah. Thursday Island, Prince of Wales Channel, Torres Straits, 

 7-9 fms. ; Port Molle, Queensland, coral-reef. 



Distribution. Ceylon {Esper) ; Formosa (coll. Brit. Mus."). 



This species has a similar habit to G. varius, if the form may be 

 regarded as that of the sponge and not of the alga ; but its acerate 

 spicules are not so long and scarcely half as thick as those of that 

 species. 



GELLIODES, g. n. 



Desmacidinidse of erect habit and well-defined form, fibre distinct 

 and compact ; . outer surface of sponge beset with pointed eminences. 

 Spicules smooth skeleton acerate and bihamate. 



This genus unites the habit of Echinonemata with the fibre of 

 Desmacidinidse and the spiculation of Gellius (Desmaeodes). Mr. 

 Carter (Ann. & Mag. N. H. 1882, ix. p. 288) has referred his spe- 

 cies Axos fibulata to the genus PTiorbas, Duch. and Mich., together 

 with his Amos anchorata, which can hardly be generically identical 

 with it, as its spiculation is an acerate and an anchorate, while 

 Phorbas amaranihus, the second species of the genus, has only an 



