260 PERCY SLADEN TRUST EXPEDITION 



It appears, however, that there is a small group of genera, including Donatia (Tethya), 

 Tethyorhaphis and Xenospongia, of monophyletic origin and closely related to one another, 

 in which the loss of the tetractinellid megascleres is absolutely constant and forms the 

 principal distinguishing feature, instead of occurring sporadically as amongst the Auroras. 

 This character, taken in connection with the characteristic spherasters* and the very 

 strongly developed cortex, seems to form a sufficient justification for retaining the family. 



Hitherto 1 have employed the name " Tethyidse " for this family, on the supposition 

 that the proper name of the typical genus was Tethya and not Donatia. It will be 

 remembered that SoUas expressed himself very emphatically on this point. He says 

 " Tethya lyncurium is a combination that by the accepted laws of nomenclature cannot 

 possibly be disturbed, since the species is the type of the genus so named by Lamarck, 

 and accepted and redefined by 0. Schmidt in 1862" [1888, p. cxxi]. Until recently the 

 name Tethya has been accepted in this sense by nearly every writer, including Vosmaer, 

 Lendenfeld, Thiele, Topsent and Lindgren. 



In 1903, however, Thiele, after consistently using the name Tethya for the genus in 

 question throughout the body of his paper on the " Kieselschwamme von Ternate," 

 remarked in a footnote that by strict application of the laws of nomenclature Tethya must 

 be substituted for Craniella and Donatia for Tethya. In 1905 Baer followed this up 

 by proposing the family name Donatiidse, and Lendenfeld [1903], Topsent [1906 b] and 

 Hentschel [1909] have all fallen into line. 



It must be admitted that there is nothing in Lamarck's original paper [1815] to 

 justify SoUas's confident assertion. The genus Tethya is. there quite unrecognisably 

 defined and Tethya lyncurium, is the fifth species to be described. The first is Tethya 

 ashestella, which, whatever it may be, is certainly not congeneric with T. lyncurium. It 

 is true that T. lyncurium comes before T. cranium, which Lamarck places sixth, but that 

 hardly justifies us in accepting it as the type species of the genus. According to Vosmaer 

 [1887], followed by Lendenfeld [1903], the fourth of Lamarck's Tethyas, T. lacunata, is 

 a Geodia ! As the generic name Geodia was only proposed by Lamarck himself on a later 

 page of the same volume as Tethya, it certainly looks as if neither T, lyncurium nor 

 T. cranium had a right to the name Tethya on the ground of priority of mention. But 

 then there are three other species before T. lacunata. What these may be I do not know, 

 and I doubt if anyone else does, but if the types are still in existence further researches 

 might upset any decision as to priority which might now be made. 



On the whole the wisest course would seem to be to abandon the generic name Tethya 

 altogether and to use Nardo's name Donatia, proposed in 1833, for D. lyncurium and its 

 congeners, and Baer's name Donatiidse for the family. 



Genus Donatia Nardo [1833] 

 = Tethya auctorum. 



Donatiidse of usually more or less spherical form ; megascleres styli ; microscleres 

 large spherasters, with .smaller chiasters or oxyasters or both ; without microrhabds. 

 Many species of Donatia have been described by various authors, but the question 

 * In the aberrant genus Tuberella, however, the microscleres have also completely disappeared. 



