THE METHOD OF ATTACK 13 



that not merely do all the millions of bacilli in a single colony, 

 say of the Bacillus tuberculosis, constitute one individual, but 

 that one individual Bacillus tuberculosis is spread all over the 

 habitable world, and that tuberculosis and not the tubercle 

 bacillus is the entity. This is an impossible position : it formu- 

 lates that the " divisa " are " indivisum." The very idea of 

 individual connotes independent existence, or, if we take a wide 

 survey and include forms like the compound myxomycetes, 

 hydrozoa and polyzoa, potentially independent existence. When 

 a bacillus grows and divides and each half floats away, there 

 is no longer one individual, and whether, as among the bacteria, 

 the division is binary, or, as in man, one of the many billions 

 of cells that constitute the body undergoes a similar binary 

 division, and one of the products floats away and becomes 

 fertilized, in both cases we deal with the development of a new 

 generation. By the same process of reasoning, basing our- 

 selves upon the doctrine of the continuity of the germ plasm, 

 we might with equal logic declare that all living beings constitute 

 one continuous individual ! 



And as regards the advantage in researches upon the bacteria 

 of being freed from the perturbing factor of sex, let me interject 

 that starting with Weismann and his doctrine of amphimixis, 

 but more particularly during the last fifteen years under the 

 influence of the Mendelians with their studies upon cross-breeding, 

 we have been inclined to lay far too much stress upon the part 

 played by sexual conjugation in the production of variation. 

 In nature, in general, the tendency, if not the function, of sexual 

 conjugation is to preserve the mean, not to induce the extreme, 

 is to perpetuate the species rather than favour the variety : 

 sexual conjugation only preserves and intensifies a variation 

 when circumstances favour the segregation of individuals of 

 the two sexes each possessing the variation, or, I would add, 

 when those circumstances actually lead to the appearance of 

 the variation in more than one individual in a particular locality. 

 Even if a property be dominant, with indiscriminate mating 

 and without segregation, such new property is apt to show itself 

 proportionately in fewer and fewer individuals in successive 

 generations. In simpler language, turn a highly-bred mastiff 

 or terrier or beadle loose in a canine population, and his de- 

 scendants descend to plain " yellow dog." It follows, therefore, 



