416 REVISION OF THE MONAXONID SPONGES, iii., 



arranged in a radiating bundle placed towards one end of the 

 embryo. 



Bemarks. — Lundbeck has noted the embryonic spicnlation in 

 quite a number of Myxillinae, but in every case observed by him, 

 contrary to what happens in the present species, the basical mega- 

 scleres make their appearance in advance of the auxiliary. In 

 reference to Grayella pyrula and Grayella gelida, Lundbeck(31b, 

 p. 33), says: "It is worthy of notice that the first occurring 

 spicules here are the spined dermal spicules, while elsewhere in the 

 Myxilleae it is the skeletal spicules which occur first." These ex- 

 ceptions, however, are only apparent, since, as I have previously 

 pointed out(18), the dermal spicules of Grayella undoubtedly cor- 

 respond morphologically to the skeletal spicules of normal Myxil- 

 linse, and vice versa. 



L. jacksoniana is probably most nearly related to the species 

 recorded from Port Phillip by Carter(7) as Halichondria isodic- 

 tyalis and by Dendy(13) as Myxilla isodictyalis; but it is hardly 

 likely that the two are identical, since in the case of the latter no 

 mention has been made of the occurrence of strongylote modifica- 

 tions of the skeletal spicules. The original Lissodendoryx isodic- 

 tyalis Carter(5), comes from Puerto Cabello, Venezuela, and pro- 

 bably is not identical with the Port Phillip sponge. 



Familia AXINELLID^. 



Under this, the final family dealt with in the Catalogue, Lenden- 

 feld describes six species, five of which are referred to the genus 

 Axinella, and one to a new genus Spirophorella. Each of these, 

 with the exception of the last-mentioned, is (nominally) represented 

 in the Australian Museum by a specimen duly labelled in Lenden- 

 feld's handwriting, but only in the ease of one, Axinella auran- 

 tiaca, is it possible to reconcile the specimen with the description. 

 It seems quite beyond doubt, however, that the descriptions of two 

 of the species — namely, those designated varieties of A. hispida 

 Montagu — are erroneous, the probability being that each is made 

 up of portions of the descriptions of two entirely different species. 

 For in the diagnosis introductory to these descriptions, we are told 



