PORIFERA. IIT. 



on the bottom, as large and small stones, shells of every kind, corals, Bryozoa and hydroids, worm- 

 tnbes, other sponges etc. The shells and other objects are generally dead, but the sponges incrust 

 also not rarely living Molluscs and Brachiopods. Often several different species grow together on the 

 same object, and when their outer apperance, as generally, is the same, it may be a matter of no 

 little difficulty to make out the borders for the different species. 



The genus Hymedesmia in its present conception was established by Topsent in 1892 (Arch, 

 de Zool. exp. et gen. 2. X, XXII) under the name of Leptosia, but Thiele has correctly shown (Ab- 

 handl. Seuckenberg. naturf. Gesellsch. XXV, 1903, 955), that the genus must bear the name Hymedesmia 

 Bow. as Topsent enumerates H. zetlandica as one of its species, and this species is the type of 

 Bowerbank's Hymedesmia. 



Topsent placed the species in his subfamily Dendoricinae, but I prefer to include it, together 

 with the nearly related Hymenancora n. g. and Leptolabis Tops. 1 ), in the subfamily Ectyoninac; the 

 whole construction of the skeleton points towards the Ectyoninae, the acanthostyli are always much 

 varying in size and often divided into two different groups; among the Mycalinae we have no genus 

 with a similar skeleton; there is also no doubt that Hymedesmia is nearly related to the Ectyonine 

 genera Stylostichon and Plumohalichondria. 



To the genus Hymedesmia I refer only such species which are quite incrusting and in which 

 the base of the main skeleton is in one plane, that is to say that all the skeletal styles have their heads 

 based on the substratum. This character is then the main distinguishing character in contrast to Stylo- 

 stichon, in which genus the main skeleton forms columns, and if we wish to have a sharp distinction it 

 is necessary to make sure, that the main skeleton va. Hymedesmia must be quite basally arranged; if the 

 dividing line is not drawn in this way, we get no sharp distinction, and I am also inclined to think, that 

 this division is somewhat natural. The group of genera which may here be taken into consideration 

 are: Hymedesmia, Plumohalichondria, Eurypon (= Hymeraphia) and Microciona. These genera I 

 understand in the following way: Hymedesmia and Stylostichon are nearly related, but are distinguished 

 by the character mentioned; a character which is often found in Hymedesmia and which may also be 

 found in Stylostichon is the above mentioned crowding of the chelae in the dermal membrane, so that 

 a more or less distinct layer is formed; Plumohalichondria stands a little more remote, distinguished 

 by having smooth, diactinal spicules in the fibres; Eurypon and Microciona are still more remote, and 

 nearer towards the Clathria-V&s. forms; they are, so far as I know them, characterised in contrast to the 

 three first-named genera by a greater difference between the skeletal and the accessory spicules and 

 by (generally) monactinal dermal spicules, the chelae are (always?) palmate chelae, and toxa are very 

 often present; I think these characters will prove valid in most cases; the two genera are distinguished 

 from one another by the character that Eurypon has a basally arranged main skeleton, while Micro- 

 ciona has columns. Some single species of Eurypon without microscleres may perhaps be difficult to 

 distinguish from some species of Hymedesmia likewise without microscleres, but as a rule I think that 

 the mentioned characters from the megascleres will be sufficient. 



The genus Hymedesmia is rather large, and it was therefore of some importance, if it could 

 be divided; Topsent has already separated the species with forcipes, forming for them the genus 



') I think Leptosastra Tops, and perhaps also Dragmatyle Tops are likewise to be placed in the lu-tyoninae. 

 The Injfolf-Rxpi-.lition. VI. 3. 6 



