PORIFERA. III. 



narrow alae or these may be quite absent; by this fact these aucorae get some resemblance to arcuate 

 chelae and it seems that they also in reality must be nearly related to these. The length of the 

 ancorse varies much, it is 0-025 0-045 mm , an ^ the thickness of the shaft is about 0-002 0-005 mm . 2 - The 

 sigmata are of ordinary shape and more or less contorted; they have a length of 0-06 0-128""" and a 

 thickness of 0-004 0-0065 ram . The microsclera seem to occur through the whole tissue. 



Locality: Station 54, South of Iceland, 63 08' Lat N., 15 40' Long. W., depth 691 fathoms; 

 station 98, in the Denmark Strait, 65 38' Lat. N., 26 27' Long. W., depth 138 fathoms. 



Remarks: As said the ancorae in this species seem to be nearly related to the chelae arcuatse 

 on account of the very narrow and, so far as I could see, sometimes quite wanting alse, but they are 

 however true ancorse. In Part II of this work p. 3 I have stated that I consider the ancorae and 

 chelae arcuatse as nearly related spicula and as having been derived from a common original form of 

 spicule. It is well known, that the chelae arcuatae may be more or less ancora-like, but being at the 

 same time true chelae, and many examples of this are given in the preceding descriptions. As one 

 of the most ancora-like chelae I have examined I may note the chela in Hymedesmia Bowerbanki. 

 The alse do not go here evenly over into the hinder side of the shaft but are somewhat incurved at 

 the shaft, so that in a side view the "tuberculum" is seen quite surrounded by the outer contour of 

 the ala; and so far as I could see the shaft is a little edge-like dilated towards the ends; these 

 spicules are however true chelae. Should there be found in the genus Hymedesmia forms of micro- 

 sclera which were intermediate between ancorae and chelae arcuatae, and this seems not impossible, it 

 would be of great interest, and it would seem to indicate at the same time, that Hymedesmia is a 

 genus of old origin, as already alluded to from other reasons under H. aenigma. 



2. H. conjungens u. sp. 

 PI. XI, Fig. 5. 



Incrusting; surface somewhat hispid. Spicula: megasclera; the skeletal spicules acanthostyli 

 without head-swelling, they are divided into two groups, large, spined nearly in the whole length or 

 only with a small apical part smooth, cr2^o-^.i mm , small, entirely spined, ov/ o-/j'" m ; the dermal 

 spicules tylota with pointed end-swellings, cri6 a 22"""; microsclera two forms, ancora spatulifcrce with 

 three teeth, 0-022 0-042""*, sigmata o-ojj 0-0/5""". 



We have one specimen of this species growing on a living Pecten together with specimens of 

 Melonanchora emphysema, Hymedesmia perforata and procumbens, a Crella, an Eurypon and a Plocamia. 

 It forms an incrustation of an extent of about 15""", and it is very thin, not reaching o'5 mm . The 

 colour (in spirit) is light brownish grey. The surface is somewhat hispid from projecting spicules. The 

 dermal membrane is a thin, transparent film, resting on the skeleton below. Oscula and pores: larger 

 and smaller, simple circular openings are seen in the dermal membrane representing I think both 

 oscula and pores. 



The skeleton. The dermal skeleton consists of bundles of dermal spicules stretching up to the 

 dermal membrane and supporting it, the spicules in the bundles being somewhat penicillately spread 

 outwards; the bundles are generally not large, consisting only of relatively few spicules. There are 



