74 the president's address. 



of one species have always three flukes and that of another 

 always more ? Why should the two ends in some cases be equal 

 and in others unequal 1 Why should the teeth at the small end 

 sometimes be shaped as in fig. 23 and sometimes as in fig. 24 ? 

 and why should some be roughened with spines and others not ? 

 We must, I think, assume that these minute differences are 

 dependent upon minute differences in the constitution of the 

 protoplasm of which the mother-cell is composed. It may be 

 a question of the chemical and physical composition of the 

 cytoplasm in which the spicule is actually deposited, or it may 

 be that the nucleus exerts some direct controlling influence 

 upon the form of the spicule, of the nature of which we know 

 nothing. 



At any rate we can hardly be wrong in attributing specific 

 differences of spicule-form to corresponding differences in the 

 constitution of the mother-cells by which they are secreted. The 

 remarkable thing is that such differences should be so constant, 

 not only throughout hundreds of thousands of mother-cells in 

 the same sponge, but throughout the mother-cells of all the 

 individuals of the same species. We can only suppose, as I said 

 before, that this constancy depends upon some constant peculiarity 

 of the germ-plasm from which all the cells of the individual and 

 all the individuals of the species originate. Obviously the ferti- 

 lised ovum must contain within itself the potentiality of pro- 

 ducing, amongst other things, all the different kinds of spicules 

 which may happen to characterise the particular species to which 

 it belongs. As development goes on differential divisions must 

 take place whereby all the different kinds of cells of which the 

 adult sponge is composed are segregated, and each mother-cell 

 must ultimately retain the power to secrete only one particular 

 kind of spicule. Now there is strong reason for believing that 

 differential cell-division is effected always by the complex process 

 of mitosis or karyokinesis, which concerns chiefly the chromosomes 

 of the nucleus, and hence I think we may pretty safely conclude 

 that specific differences in the form of the microscleres must 

 depend upon differences in the constitution of the nuclei of the 

 mother-cells, or, in other words, that the nuclei of the mother- 

 cells determine to a large extent the form of the microscleres. 



There appear, in short, to be three secondary factors concerned 

 in the production of any particular form of microsclere : (1) the 



