1 Porifera. 



B. Special. 

 1. Hexactinellida. 



See also supra p 9 Marshall. 



Schulzel 1 ) notes that the Hyalonematidae must be considered to have sepa- 

 rated from the other Hexactinellida late or early according as we believe the am- 

 phidisc to be a highly modified hexaster or an independent derivative of the 

 original hexactine. Considering together outward form, root-tuft, and general 

 organisation, the following relationships are shown : 



Hyalonema 



Pheronema 

 Urhyalonematidae 



Poliopogon 



- Semperella 



A detailed scheme of the species (two new of Hyalonema} is given. Stylocalyx, 

 the subgenus of Hyalonema, is abandoned ; the presence or absence of an oscular 

 sieve-plate not being of systematic value. The sieve-plate in Hyalonematidae ap- 

 pears keineswegs mit der terminalen Siebplatte von Euplectella oder Aphrocal- 

 listes homolog, soudern nur ein mehr oder minder weit von der Gastralmembran 

 abgehobenes Balkennetz zu sein , unter welcher zwar ein Subgastralraum, aber 

 nicht die eigentliche Gastralhohle zu suchen ist. 



Schulze( 2 ) points out that in his Challenger monograph he showed the deriva- 

 tion from the hexactine for most forms of Hexactinellid spicules, even when shor 

 tening of the primary branches has produced the delusive appearance of a 

 polyaxon. There appears at first special difficulty for the discoctaster of Acan- 

 thascus and Rhabdocalyptus , which has 8 arms, each bearing a disc of secondary 

 branches. Examination now shows that each arm is trihedral at its base, and 

 alternating with the arms are 6 swellings [geometrically, the 6 swellings are on 

 the 6 surfaces of a cube, from which the 8 arms pass out at the 8 angles]. The 

 spicule is really a hexactine, each ray of which branches into 4, the 24 branches fus- 

 ing in groups of 3 to form 8 arms. This view is supported by the following facts : 



- (1) In the axis effusion of each arm is a line of irregular spaces, differing 

 completely from the normal axial canal. (2) The 6 swellings are in exceptional 

 cases prolonged into actual branched rays. (3) The terminal branches on an arm 

 are in most cases 3, 6, or 9 in number. (4) Occasionally one of the elements 

 belonging to an arm remains isolated from the other two. (5) In Trachycaulus 

 Gurlitti the author has already recorded fusion of neighbouring branches. - - In 

 Pheronema giganteum occur large knobbed siliceous spheres, 0,2 mm in diameter, 

 with a small, less refractile centrum ; similar bodies of elongated and less regular 

 form sometimes show a short straight axis-canal. S. suggests that they are si- 

 liceous pearls, the latter being formed round a broken spicule, the former on 

 some small intrusive body; this body, however, could not be demonstrated. 



2. Demospongiae. 



See also, except Dendy. all the papers under General (supra p 2 ff.) and see 

 Berghaus. 



Celesia. after a summary of previous knowledge, gives measurements of the 

 sponge, the shell, the Pagurus, and the vestibule in 34 examples of Suberites do- 

 muncula. There are no signs of corrosive action by the sponge on the shell, 

 Krukenberg's denial being justified ; the vestibule, a cavity beyond the lip of 

 the shell, had been mistaken for the mould of a dissolved shell. It is formed by 



