8 Porifera. 



instance hitherto known suggesting recapitulation in a gemmule appears to be 

 the hydromedusa Epenthesis McCradyi, described by Brooks [v. Bericht f. 1888 

 Coelenterata p 5]. - - The young ovum in Tedanione has a single central nucleolus, 

 with growth there are seen two nucleoli on opposite sides of the nucleus: the 

 nucleus of the mature ovum contains either one peripheral nucleolus or none at 

 all; in the latter case a nucleolus may be often seen lying outside the nucleus. 

 There is a follicle, surrounded in young stages by closely packed mesoderm cells. 



- Segmentation seems to be normal, the free larva is uniformly ciliated; one 

 larva fixed on the parent was observed, ciliate but already containing flagellate 

 chambers. - - InHircim'a acuta the nucleolus of the ovum behaves as in Tedanione, 

 the follicle until maturity is indistinguishable from the closely packed granular 

 mesoderm cells surrounding it. 



Topsent ( 6 ) notes that HalichonJria inops (Roscoff in August) has larvae which 

 prove it an Esperelline. Renicra fulva n. , Suberites sulphur eus, and Esperclla 

 modesta reproduce by superficial buds like Tcthya. 



Topsent ( 7 ) finds Sanidastrella coronata constricting off its exhalant papillae as 

 buds. The smaller kind of gem mules in Craniella cranium are entirely aspi- 

 culous, hence to be termed >jgemmules inermes. A delicate hyaline cuticle 

 overlies a thin granular zone, the whole interior mass consisting of large cellules 

 embryonnairesa. The larger, always more numerous gemmules arrnees have 

 radial spicules, a delicate cuticle, and granular cells imbedded in a semi-cartila- 

 ginous matrix. From October to April at Banyuls , in August and September on 

 the Channel no Tetractinellids were ever found with ova or larvae. 



Weltner (*) describes a Chalina oculata full of embryos and larvae, in which the 

 maternal soft parts had nearly vanished. 



According to Lendenfeld( (i ) Sollas is quite rip-htinhis description of the embryo 

 of Oscarella as a folded blastula [see Heider] ; possibly the same is true of Corti- 

 cium. - The ova described by Kolliker for Ancorina cerebrum belong to a sym- 

 biotic worm. - - Sperm-balls in Erylus discophorus contained irregular, darkly 

 staining bodies of 10 JA )( 5 jx. The ova of Geodium cydonium sometimes con- 

 tain no nucleus. 



Garbini finds in Euspongilla lacustris embryonal spicules always spined: in 

 summer sexual elements and gemmules may be found coexisting in one sponge. 

 In Ephydatia ftmiatilis embryonal spicules are smooth. Possibly the more confined 

 distribution of this sponge is due to the amphidiscs causing its gemmule to sink 

 [cf. Bericht f. 1893 Porif. p 12 Weltner]. 



Topsent ( s ) gives a list of sponges at Le Portel, near Boulogne, found with larvae 

 in August. 



B. Special. 



1. Hexactinellida. 

 See Jjima and supra p 3 Topsent( 7 ). 



2. Demospongiae. 



See also all papers in A, and see Herdman, Lendenfeld('), Topsent ( 3 ). 



Vosmaer( 2 ) describes 11 species (4 n.) of Choristida. In Cydonium gigas, 

 as in many sponges considered to be purely radial, the central parts show 

 oxeas concentrically arranged at right angles to the radial bundles, hops macu- 

 losus n., as all Geodidae, has sterrasters in the parenchym, probably formed there 

 to be transported afterwards to the cortex. - - Canal-system - .' C. gigas. 

 Scattered groups of stomions lead each into a wide subdermal funnel, succeeded 



