1914.J N. Annandale : Sponges of Lake Baikal. 139 



ever, seem to acknowledge that, whatever criterion is adopted in 

 the separation of families, some or all of them will be of poly- 

 phyletic origin and include genera that resemble one another 

 because of convergent evolution rather than of direct common 

 descent. The precise classification adopted is, therefore^ largely a 

 matter of convenience. If great stress is laid on the microscleres 

 alone there is this difficulty, that in certain genera {e.g. Hom- 

 acodictya) ' the microscleres are very liable to be overlooked or 

 lost altogether, and species assigned not only to the wrong genera 

 but even to the wrong family ; while in many genera microscleres 

 are invariably absent In those genera, however, in which they 

 are present there can be little doubt that they form by far the 

 readiest means of identification and separation in the case of 

 properly preserved specimens, and on the whole it is perhaps 

 most convenient to consider them first in separating the larger 

 divisions, although in their absence other characters must be 

 found. 



In Prof. Dendy's ■' report on the sponges collected by Prof. 

 Herdman in the Gulf of Manaar (1905) there is, on pp. 133 to 

 135, a useful discussion of the composition and position of the 

 families of the suborder vSigmatomonaxonellida. This suborder 

 consists of Monaxon sponges in which the typical microscleres 

 are sigmata, or forms derived therefrom, true asters being absent. 

 The first family assigned by Dendy to the suborder is the 

 Haploscleridae, in which, following Topsent ^ he includes the 

 Homorrhaphidae and Heterorrhaphidae as defined by Ridley and 

 himself * in 1887 and by other authors He assigns to the 

 Haploscleridae those genera in which chelae and anchorae are 

 absent, the skeleton spicules being as a rule amphioxi or amphi- 

 strong3di and the spicule-fibres typicall)^ nonplumose. The marine 

 subfamilies to be considered here belong to this family but have 

 no microscleres. 



In dealing with the Baikal sponges it is necessary to consider 

 the relationship between the Haploscleridae and the Spongillidae, 

 in which all other freshwater sponges must at present be placed. 

 In individual specimens, and even in some cases in species and 

 genera, it is often extremel}^ difficult, if not impossible, to find 

 any definite character that would separate a Spongillid from a 

 Haplosclerid. In both families we find sponges totally devoid of 

 microscleres and having a somewhat amorphous skeleton composed 

 of amphioxi held together by a greater or less amount of chitinoid 

 substance. 



The typical microsclere of the Haploscleridae is a C-shaped 

 spicule (sigma), which may be modified in different ways but 

 never assumes the complicated form of the chela or of the anchora 

 and rarely becomes straight and rod-like. Microscleres are in 



' Lundbeck, ' Ingulf Exp. Vl, pt. i, p. 6, footnote (igo2/. 



2 Herdman's Rep. Pearl Oyster Fisheries (Roy. Soc. J-ondon), III 1 1905). 



3 Mem. Soc. Zoo!. France VII, p. 5 (1894). 



* 'Challenger' Rep. Zool. XX (Monaxonida) (1887 I. 



