THE PRESIDENT'S ADDRESS. 71 



inhalant canals, where they may perhaps serve to filter the 

 incoming water and guard against parasites, as in the case of 

 the " sigmata " of Esp>erella murrayi ; but in the vast majority 

 of cases it is impossible to assign any value at all to the 

 presence of microscleres. Indeed, the numerous species of 

 horny sponges seem to get on quite as well without these 

 bodies. 



Nevertheless we find that the microscleres, when present, 

 are characterised by very definite and constant forms, and many 

 of them are amongst the most beautiful and wonderful objects 

 that come under the observation of the microscopist. So constant 

 and characteristic are they that they afford by far the most 

 convenient and most reliable data for the classification of the 

 tetraxonid sponges. Particular species, and even particular 

 genera and families of these sponges, are characterised by the 

 presence of highly specialised forms of microscleres, and in the 

 case of species the characteristic form is almost invariable. 



There can be no doubt that the microscleres have undergone 

 an evolution along definite lines, and one species of a genus is 

 commonly distinguished from another by differences in the 

 shape of these spicules, which, though constant, appear at the 

 same time to be utterly trivial as, for example, the difference 

 in the shape of the teeth at the small end of the " isochelae " in 

 Cladorhiza pentacrinus (figs. 23, 23a) and Cladorhiza (?) tridentata 

 (figs. 24, 24a). There may be several kinds of microsclere in 

 the sponge, all characteristic of the species, but a single sponge 

 may contain many thousands, or perhaps millions, of the same 

 kind, all exactly alike in shape and size except for an occasional 

 individual variation such as occurs in all organisms. 



The shape of the microscleres appears to be quite independent 

 of their position in the sponge, and must obviously be attributed 

 to some specific peculiarity of the ovum from which the sponge 

 developed. It is clearly of a blastogenic and not a somatogenic 

 character, and it is usually much more remarkable and quite 

 as constant as that of the megascleres of the same species. 



The microscleres of the tetraxonid sponges may be divided 

 into two categories, termed astrose and sigmatose respectively. 

 The former (figs. 14a 14A) may be derived from the tetraxon 

 archetype by multiplication of the rays due apparently to 

 meristic variation accompanied usually by diminution in size of 



