PORIFERA. I. 



103 



the canals. The two other forms, on the contrary, are found scattered in the tissue of the sponge, in 

 the dermal membrane, and in the membranes lining the canals. 



As will be seen, the three mentioned forms of diancistra are well characterized and distinctly 

 separated; this fact, however, has not hitherto been acknowledged, the smaller forms having been 

 regarded as developmental phases of the largest one. Thus Ridley and Dendy (Challeng. Report, XX, 

 ]\Ionaxonida 61) sa)- of the small diancistra of the H. esperioides established by them that they seem 

 to be young forms of the large diancistra. I have had occasion to examine H. esperioides, and it was 

 seen that the small diancistra of this species are forms having great resemblance to the small diancistra 

 of the present species; accordingly they are an independent form without any connection with the 

 large diancistra, neither could they, by a growth by mere apposition, get a form like that of the large 

 ones'). Topsent, in the work quoted above p. 87, says of the small diancistra in the species deter- 

 mined by him as H./ohnsoniVtO^-. . . . petits diancistres greles que Ton pent prendre pour des spicules 

 jeunes on frappes d'un arret de developjjement. > Both these interpretations are untenable, as the small 

 diancistra, as is shown in the preceding description, are independent forms without any connection 

 with the development of the large diancistron; as was to be expected, this development takes place 

 in a quite different way, and they grow only by apposition. 2. Toxa; these are comparatively fine; 

 they have a strong curve in the middle, while the ends are long and evenly recurved, and most fre- 

 quently the outermost piece of the ends forms a point which is slightly recurved with a rather sharp 

 bend. Within the bounds of this description they may be somewhat varying, being higher or flatter, 

 or somewhat irregular, and they are also often somewhat contort. The length, which is, to a certain 

 degree, dependent on the degree of curving, may vary between o'oS and o-i57""", and sometimes, in 

 more flat bows, it may be a little greater; the thickness in the middle is 0-0017— 0'002™". The bows 

 are found scattered in the tissue, but are not numerotis; the}- are not found in the dermal membrane. 



If I have not thought to be justified in determining the present species as H. Jolmsmii Bow., 

 the reason is partly that we have no detailed descriiJtion of this species. Bowerbank, as is well 

 known, in 1864 in the first volume of Mon. of Brit. Sponges, only describes the large diancistron of a 

 species from Madeira without giving this species a name in the text, only mentioning that it is re- 

 lated to the genus Hymedesmia\ in the explanation of the figures he calls it Hyincdrsinia Jo/insoni 

 M. S. He gives two figures of it, from which is seen that it has large diancistra and st\-li. In 1870 O. 

 Schmidt, in Spong. d. atlant. Gebiet, enumerates a species, which he identifies with that of Bower- 

 bank, and calls Desiiiacclla Johnsoni, but he gives no distinct description. Topsent, in the work 

 quoted before, enumerates a species which he determines as Hainacaiifha Johiisoni Bow. ; of the micro- 

 sclera in this species he says that large diancistra are alwa\-s found, small fine diancistra almost 

 always, sometimes sigmata, sometimes bows, and rarely (in one easel rhaphides. Topsent, however, 

 gives no measurings of the different forms. Now it is rather certain that Topsent has had before 



I) Ridley and Dend}-, PI. XVII, fig. 2 d, figure one of these small diancistra, and say in the explanation of the plate: 

 young diancistron, showing tlie hooks still united to the shaft by a thin web<, which fact is also shown in the figure; this 

 interpretation, however, is wrong, the structure being the same as in the small diancistra of the present species; the fact is 

 that the edge issuing from the shaft is verj- broad, and comes close to the recurred end, but it is always distinctly separated 

 from this end. In H. esperioides only two forms of diancistra are found, corresponding to the smallest and the largest of 

 the present species. 



