THE PRESIDENT'S ADDRESS. 73 



of a saw, occurs at each end of the shaft, just where it bends 

 round. 



Now it appears to me quite idle to argue that minute differ- 

 ences in the form of the microscleres, such as I have just described, 

 are of any importance to the sponge in whose soft tissues these 

 microscopic spicules are scattered without order or arrangement. 

 Nevertheless they constitute, as I have already said, constant 

 specific characters, and have undoubtedly arisen by some process 

 of evolution, one form Lading to another just as in the case 

 of any other characters. Such characters are, of course, by no 

 means confined to sponge spicules ; they may be more or less 

 exactly paralleled, for example, in the frustules of Diatoms, the 

 shells of Foraminifera and Kadiolaria, and the calcareous spicules 

 of Holothurians. Natural selection cannot be directly respon- 

 sible for their origin. How, then, are they to be accounted 

 for? 



Before attempting to answer this question let us inquire how 

 a microsclere actually arises in the sponge. It appears that, 

 from an early stage in embryonic development, certain cells, 

 known as scleroblasts, or mother-cells, are set aside for the 

 purpose of spicule-formation. These mother-cells have the power 

 of extracting silica in solution from the sea-water which circu- 

 lates through the sponge, and depositing it in the form of solid 

 opal, and in the particular shape characteristic of each spicule. 

 Each separate microsclere arises thus in the interior of a single 

 mother-cell. Let us examine a little more closely the conditions 

 under which it is deposited. 



The mother -cell is, of course, a nucleated mass of protoplasm, 

 and it appears to be bounded on the outside' by a more or less 

 definite cell -membrane. The spicule, at any rate in the case 

 of sigmata and chelae, appears to be deposited on the inner 

 surface of this membrane, and this fact probably explains why 

 it is curved. If we assume, as seems probable, that the mother- 

 cell continues to grow while the spicule is being deposited, and 

 that the spicule is adherent to the cell-membrane, then we may 

 further suppose that the increasing tension and expansion of the 

 latter may cause the thin siliceous film to split into flukes or 

 teeth. Probably, then, the form of the spicule is largely due to 

 mechanical causes. We cannot, however, explain the minute 

 details of structure so simply as this, for why should the chela 



