64 



p()rifb;ra. I. 



those o-iven by Ridley and Dendy for the specimen of Bowerbank; Topsent also states that the 

 spicules project through the skin, while Bowerbank describes the surface as smooth. 



The species in hand has a thin dcriiinl mciiihrauc that cannot, however, be torn off; no parti- 

 cular dermal skeleton is found; but if a thin section is cut off parallel to the surface, this section 

 shows an irregular network, which is poly.spicular, but here and there also formed of .single spicules. 

 Spicules and bundles of spicules project through the skin, so that the surface is not smooth. The 

 pores are round, and are situated in the meshes of the dermal membrane; they are measured of a size 

 of o-oo8— o-i48""". 



The skeleton, as far as I have been able to examine it, has a somewhat halichondroid structure 

 consisting of loose, little marked fibres, among which, however, in numy places, a unispicular reticu- 

 lation is found. The ends of the spicules are united b>- a distinct, but clear mass of spongin. 



Spicula: a.. Mega sc /era are oxea, more or less curved in the middle, gradually tapering to the 

 point, which is most frequenth' somewhat marked off, and the ver\- outermost point is oftenest more 

 or less shortly pointed; this structure, however, is not always pronounced; the length varies between 

 0-32 — 0'388'"'^, and the thickness is 0-009 — 0-012'""'. Shorter and finer oxea are only seen in ver}' small 

 numbers, b. Microsclera : \) Toxa; with regard to form the>' resemble the bows of the preceding 

 species, but they are smaller and especialh' thinner; the cur\-e in tlie middle is sometimes so sharp, 

 that the branches form a right angle; as in arcoferiis the bending at the ends consists onh- of a little 

 point inclined backwards. The length of the bows is between 0-12—0-064""", and the thickness is from 

 0-002""" down to o-ooi""". 2) Sigmata; these are rather small, of an average length of 0-021'"'", and a 

 thickness of o-ooi"'"'. — Besides these sigmata a ver\- few much larger ones are found, of a length of 

 about 0-078""" and a thickness of 0-005""". These sigmata are often of a somewhat monstrous form, 

 with one or both ends rounded or showing other irregularities. As has been said, they are onh* found 

 quite singlv, but they are constantly found, so that they cannot be taken to be extraneous. They are 

 not seen by transitions in sizes to be connected with the small sigmata. — In a specimen from Egedes- 

 minde, the spicules of which were upon the whole of the largest of the sizes given above, the sig- 

 mata were a little larger than those mentioned abo\-e, of a length of 0-028"'™ and a thickness of 0-002"'"'. 



Locality: Adelvig on the north-western coast of Iceland, depth 6—15 fathoms, some fragments 

 which have apparently belonged to one specimen (the author); Egedesminde, a few small fragments 

 (Bergendahl). 



Geogr.dJsir. Gelliiis ai/g/ilatt/s is common in the English Channel (Topsent), and has further 

 been taken at the Azores (Challenger, Topsent) in tlie latter place on depths of up to 450 fathoms. 



3. G. luridus n. sp. 

 PI. VI, Figs. 5—8, PL XIII, Fig. I a--c, Figs. 2— S. 



Free; ohloiig pyrifojnn, more or less irregular, soweiiiiics ro/iiidisk or lobate. The surface 

 smooth. Outermost a hark cojisistii/g of a thinner or thicker layer 7i'i/h close-packed spicules lying 

 ill all directions, but parallel to the sitrface. The skeleton consists of a rather irregular, mostly uni- 

 spicular netzoork; it is crossed by concentric lamella' of a similar structure as the dermal layer. Spi- 



