246 NATURAL SCIENCE. Oct.. 



unicellular, and the sexual reproduction of multicellular organisms, are 

 means of producing variations. The process furnishes an inex- 

 haustible supply of fresh combinations of individual variations which 

 are indispensable to the process of selection." Here is a definite 

 scientific hypothesis which can be examined in the light of such facts 

 as we possess. 



Weismann, we must observe, does not affirm that variation is 

 primarily due to conjugation, but agrees with Darwin and most others 

 in thinking that the power to vary lies in all protoplasm independent 

 of conjugation. In support of this, we know that no two cells of the 

 body are quite similar to each other, and many are very unlike, yet 

 these have arisen from the original fertilised ovum by asexual division, 

 without conjugation. By repeated division the fertilised ovum gives 

 rise to cells which at once begin to vary, and finally change into what 

 we call the cells of epithelium, muscle, cartilage, etc. In plant life we 

 constantly come across variations in which sexual union has played 

 no part ; such, for instance, as eccentric leaves and branches which 

 may suddenly arise during the life of a growing tree. 



The use of sexual conjugation is, according to Weismann, to give 

 to natural selection a greater choice, to present to it an increased 

 number of variations. Now, at the outset, this explanation of the 

 utility of sexual conjugation appears to me to be unsatisfactory, 

 because in natural selection alone we have the means of bringing 

 about any degree of variability we require. If some of the individuals 

 of a species are not variable enough, natural selection will preserve 

 the more variable ones, thus increasing the variability of the species ; 

 for variability, like everything else, can be transmitted by heredity. 

 The extraordinary power of selection in modifying variability may be 

 appreciated when we study in contrast wild and domesticated or 

 cultivated animals and plants. In the wild condition most types 

 remain practically unchanged for long periods of time, they have 

 become adapted to their conditions of life, and these as a rule remain 

 pretty uniform. Natural selection will operate largely in destroying 

 any marked variety, and those strains will be perpetuated which 

 evince little tendency to produce variation. When, however, we 

 cultivate or domesticate these types, and when we encourage 

 variations by selecting them in preference to forms of the original type, 

 we preserve variable strains, and the breed under domestication 

 may become very variable. Weismann's theory of the use of 

 conjugation assigns to it a function possessed in marked degree by 

 natural selection, and we are tempted to believe that it will have a 

 function peculiar to itself. 



Let us examine a little more closely the facts of sexual and 

 asexual reproduction, contrasting these with each other. It is 

 generally acknowledged, and it was strongly emphasised by Darwin, 

 that among plants a variation can be propagated with certainty by 

 the asexual method alone. If we try to perpetuate it by seeds it 



