86 VARIABILITY 



are transmissible in a state of latency through parents of the 

 opposite sex, 1 and the Mendelian experiments demonstrate that 

 latency occurs in characters other than the sexual. 2 



137. The bacteriological evidence is important and deserves 

 to be considered at length. Every unicellular organism is a germ- 

 cell, just as every germ-cell is in effect a unicellular organism. 

 In the last analysis, therefore, the problems of heredity presented 

 by them are similar. In other words, if we think in terms of the 

 germ-plasm, not in terms of the cell-community, the problems of 

 heredity presented by unicellular organisms and germ-cells are 

 similar. We noted that the question as to whether the acquire- 

 ments of unicellular organisms are, or are not, inheritable is totally 

 distinct from the problem of the transmission of acquirements 

 amongst unicellular types. 3 Its real counterpart amongst the 

 problems of multicellular life is the question whether or not 

 changes caused in the germ-cells by the direct action of their 

 environments are inherited by descendant germ-cells. Obviously 

 this is a very different problem from the question whether the 

 particular changes caused by the environment in the somatic cells 

 are inherited by the descendants of the germ-cells, whether 

 the acquirements of Brown and Jones are inherited by the 

 descendants of Robinson and White. Now unquestionably a 

 unicellular organism or a germ-cell may be injured or benefited 

 by its environment. The problem we have to solve, then, is 

 whether such changes tend to be inherited by the descendant cells. 

 In other words we have to discover whether they necessarily 

 or usually imply alterations (i.e. variations) in the germ-plasm. 



138. Of all unicellular organisms the disease-producing (patho- 

 genic) microbes have been most closely studied, and of all their 

 characters their powers of offence and defence have claimed the 

 most attention. When they enter a living body, a man for 

 instance, a struggle ensues between them and the cells of the 

 host. They, or at least some species of them, secrete toxins, 

 soluble poisons which defend them from the cells, which in turn 

 secrete substances (enzemes) poisonous to the microbes. 4 If only 

 a few microbes enter the body they are destroyed. Thus it has 

 been found experimentally that if less than sixteen thousand 

 virulent anthrax bacilli be introduced into a normal young rabbit, 

 they all perish. A greater number multiply and cause disease 

 in a suspectible individual, presumably because they so poison 

 the cells of the body that the latter are unable to poison them. 



1 See 238. 2 See 273 et seq. 3 See 98 and 431. J See 413. 



