!54 PORIFERA. II. 



or similar forms, sometimes with spined ends; microsclera arc chelce arcuatcr of one or more forms, and 

 often sigmata. 



The genus Lissodendoryx was established by Top sent, first in 1892 (Resultats des campagu. 

 scient. du Prince de Monaco, Fasc. II, 97), as a subgenus of Dendoryx (= Myxil/aj, and later, in 1894 

 (Mem.de la Soc. zool. de Fr. VII, 35) raised to an independent genus. The typical species was L. lepto- 

 derma Tops. The species first described among those belonging to Lissodendoryx seems, as stated by 

 Top sent, to be L. isodictyalis Cart, described by the author under Halichondria. Later, Top sent 

 (Rev. Suisse de Zool. IV, 1897, 457) has described a species from Amboina, which he identifies with 

 isodictyalis Cart, and at the same time he identifies his own leptodcrma with this species; according 

 to this, isodictyalis Cart, would become the type. Thiele (Studien iiber pazif. Spong., Zoologica, Heft 

 24, II, 1899, 18, Taf. V, Fig. 10) thinks, however, that Topsent's species from Amboina is not identical 

 with isodictyalis Cart, and gives it the name of similis ; at the same time he says that he cannot 

 decide with which of the two species leptodcrma may possibly be identical. Accordingly leptodcrma 

 must still for the present be regarded as the type of the genus. 



The character by which Topseut distinguished the genus from his Dendoryx, was exclusively 

 the smooth styli in contradistinction to the spined ones in Dendoryx. This character is a very unten- 

 able one, as species are found with scarcely perceptible spinulation, and Topsent himself admits also 

 this fact. Instances of such species are firma Lambe, with styli with scarcely perceptible spinulation, 

 and the species fragilis Frstdt. and iudistincta Frstdt, to be treated hereafter, which have slightly 

 spinulous to quite smooth styli. An even transition is in reality found from species with strongly 

 spined styli to such ones where the styli are quite smooth. According to this the genus therefore 

 would scarcely be maintainable. There is, however, another character which sharply separates the 

 species of the genus Dendoryx (or Myxilla. as it ought now to be called) into two groups, whether 

 they have smooth or spined styli; the fact being that in some species ancora; occur, in others chelae 

 arcuatse (see the account of these spicules in the introduction). This character is a quite sharp one 

 without transitions of any kind, and in my opinion great stress may be laid on it for systematic 

 purposes. As L. leptoderma Tops, is a species with arcuate chelae, the genus containing species with 

 these chelae will accordingly get the name Lissodendoryx, while in the genus Myxilla, with the typical 

 species rosacea Lieberk. only species with ancorse remain. 



The name of Lissodendoryx is not exactly a good one by the new limitation of the genus, but 

 it cannot be rejected. Otherwise the fact seems to be that most Myxilla-sipecies have spined styli, 

 and most Lissodendoryx-species smooth ones. 



1. L. lobosa n. sp. 

 PI. V, Fig. 5. PI. XV, Fig. 4 a— c. 



Erect, irregularly lobed-branched. The dermal membrane very thin. The skeleton a somewhat 

 diffuse and irregular reticulation, chiefly of polyspicular fibres with single transverse spicules between 

 them. Spicula: Mcgasclera : the skeletal spicules acanthostyli with rather dense spinulation o-ji — 0-369""", 

 the dermal spicules tylota 0-23 — 0-29"""; microsclera of one form, chelce arctiatm 0-038 — 0-044'"'". 



