THE president's ADDRESS. 235 



exact parallel that I know of, however, occurs in quite a different 

 group of animals — the Hoplonemertines. These remarkable 

 worms possess a long, eversible proboscis, armed at its tip 

 with a single calcareous stylet, shaped like a nail. The head of 

 the nail is fixed in a kind of handle, formed as a secretion by 

 certain glandular cells. But the stylet does not develop in this 

 position ; a supply of them is formed — apparently by the secre- 

 tive activity of special mother-cells — and kept in reserve in little 

 sacs which lie at the side, not far from the handle, and it seems 

 that wlien one stylet is broken or worn out it can be replaced 

 by another taken from one of these armouries. 



But even if we assumed, which I am far from doing, that the 

 removal of the discorhabds from their places of origin and their 

 arrangement at the surface was due to some intelligent and pur- 

 posive action on the part of the sponge, we should hardly be any 

 nearer to an explanation, for the sponge is apparently devoid 

 of any organs through which its intelligence might operate in 

 bringing about such a result. The mystery will be cleared up 

 only when we are in possession of adequate information as to the 

 various changes through which the sponge passes in its develop- 

 ment as a whole. We cannot solve such problems by the con- 

 templation of completed results, and the truth of this statement 

 is very well illustrated by the development of the discorhabds 

 themselves, to the consideration of which we must now 

 pass on. 



It is generally believed that the siliceous sponge-spicule, 

 whether microsclere or megasclere, originates within a mother- 

 cell or scleroblast. In some cases it is supposed that the entire 

 growth takes place within such a cell, but where very large 

 spicules are formed it is suggested that the later additions of 

 hydrated silica, or opal, are contributed by accessory silico- 

 blasts. It has also long been recognised that the secretion of 

 silica takes place around an axial thread of different but un- 

 known composition, the position of which is frequently indicated 

 by the presence of a more or less distinct axial canal in the fully 

 developed spicule. The mode of origin of the axial thread, or 

 " protorhabd " as I now propose to term it, is entirely unknown, 

 but it seems generally to be assumed that it arises in the same 



