80 THE PRESIDENT'S ADDRESS. 



to direct your attention before concluding. The constancy in 

 the specific form of the microscleres of the Tetraxonida appears 

 to be much greater in the case of the sigmatose than in that of 

 the astrose series, and in the former at any rate seems to point 

 to the different modifications having arisen as mutations rather 

 than as fluctuating variations. This would, I think, be quite 

 in harmony with the views which I have been endeavouring to 

 express. A mutation, however small it may be, is believed 

 to be due to some change, apparently sudden, in the constitution 

 of the germ-plasm, which may then remain without further 

 alteration until another mutation occurs. To say that the 

 change in question is probably of a physico-chemical character 

 seems almost a truism ; but if it is so it seems only natural to 

 suppose that such a modification, transmitted by cell-division 

 to all the mother-cells of a particular kind, may affect in a 

 uniform manner the form of all the microscleres deposited in 

 these mother-cells, just as a change in the character of the 

 reagents employed will affect the form of osmotic growths ex- 

 perimentally produced. If this view be correct, we must suppose 

 also that any adaptive modifications with which the modifications 

 of the microscleres may possibly be correlated must also have 

 arisen as mutations. I see no objection to such a supposition, 

 for mutations, if they occur sufficiently frequently, may be quite 

 as valuable from the point of view of natural selection as small 

 fluctuating variations. 



We do not, of course, know what may be the cause of the 

 modification in the constitution of the germ-plasm that gives rise 

 to a mutation, but there is some reason to believe that it may 

 be due either to the permutations and combinations of ancestral 

 characters which take place in the maturation and fertilisation 

 of the germ-cells, or to the influence of some change of environ- 

 ment upon the germ- plasm. If the characters of sponge spicules 

 are really of the nature of mutations it should be possible to 

 obtain Mendelian results by hybridisation, and I hope that 

 at some time in the future experiments may be made with this 

 object in view. The difficulties in the way of carrying out 

 such experiments would probably, however, be very great, and 

 we should require to know a great deal more than we do about 

 the breeding habits and life-history of sponges before we could 

 hope to bring them to a successful issue. 



