SPONGES.-H ALLM AN IN . 



143 



colonensis,^ lotrochota cocciJiea^ and of other species which 

 might be mentioned, are auxihary megascleres, and that they 

 therefore differ from normal species of their respectixe genera 

 in the absence of principal megascleres ; and since we are en- 

 abled to detect the relationships of these species only by reason 

 of the marked peculiarities of their microscleres, it is extremely 

 probable that a number of the species included in certain My- 

 caline genera {e.g., Desmacidon, Amphilectus, Esperiopsis and 

 Batsella) are similarly derived from various Myxillinje which 

 are lacking- in striking- microscleric characters. Further, one 

 feels scarcely any hesitation in asserting that the dermal spi- 

 cules of Pseudoclatliyia, CreUa and GraycUa, and the scattered 

 spined diactinal spicules of Histodennelln, are accessory or, 

 at any rate, undifferentiated basical megascleres ; whilst the 

 peculiar forms assumed by the accessory spicules in certain 

 species of Acar)ius, e.g., by the "cladotylostyles" of .1 . tortilis^ 

 and the cladotylota of .4. tenuis'^ lead one to suspect that the 

 microtylota of Microtylotella gihitheri belong to the same cate- 

 gory, and that this genus ought therefore to be given a place 

 in the vicinity of Acarnus. Also, it is not altogether improb- 

 able that the spined forcipes of Leptohasis and Forcepia are 

 derivatives of accessory megascleres ; the larger forcipes of 

 L. arcuata^ are especially suggestive of such a derivation, 

 and it is worthy of notice also that in the genera CreUa, Gray- 

 ella and Histodermella, which, like Forcepia and Leptohasis, 

 possess chelag arcuatae, the scattered accessory spicules are 

 frequently curved and diactinal. If the megascleres other than 

 the accessory are of a single kind only, it is not always pos- 

 sible to decide with certainty whether they are principal or 

 auxiliary spicules ; as a rule, however, reasons can be found, 

 depending upon their form, in support of their identification 

 with one, rather than with the other, of these categories. 

 Principal megascleres, in nearly all cases in which their identity 

 is certain, are sub-conical or more or less fusiform, somewhat 

 curved styli, which are either quite smooth or are provided 

 with spines over a greater or less portion of their length ex- 

 tending from the basal extremity upwards. The auxiliary 

 megascleres, on the other hand, are typically straight and of 

 fairly uniform diameter, and are rarely spined except at their 

 extremities ; in comparison with the principal spicules they are 

 of relatively slender proportions, and in most genera are typi- 

 cally diactinal in the fully developed state, with usually tornote, 



1 Carter— Ann. Mag. Nat. Hist. (5), xv., 1885, p. 110. For this species 



Hentschel (1911) has recently proposed a new name. F. MichaeUeni. 



2 Carter— Ann. Mag. Nat. Hist. (5). xviii., 1886. p. 378; Dendy— Proc. Roy. 



Soc. Vict., viii (n.s.), 1896, p. 23. 



3 Topsent— Resiiltats Camp. Scient. Pr. de Monaco, Fasc. xxv., 1904, 



p. 171, pi. xiv., fig. 8. 



4 Dendy— Proc. Roy. Soc. Vict., viii., (n.s.), 1896. p. 50. 



5 Topsent— Op- cit., p. 183, pi xv., fig. 13b. 



