ESMOND E. SNELL 433 



all of the compounds in it are of recognized importance for all forms of 

 life so far examined. Each of them, so far as we now know, is contained 

 in the protoplasm of every living organism, be it plant, animal, or 

 microorganism. 



Since these compounds seem to be present generally in all protoplasm, 

 it is clear that those organisms which do not require them as nutrients 

 must synthesize them, and that bacteria which require a preformed 

 supply of one or more of them in the medium must, for some reason, 

 be unable to synthesize those which they require. 



This idea that the growth-factor requirements of microorganisms arose 

 through loss of the ability to synthesize substances of importance in the 

 metabolism of all organisms was given formal expression independently 

 by Lwoff (19,20) and by Knight (13) about fifteen years ago, while 

 knowledge of bacterial nutrition was still fragmentary. These authors 

 envisaged a physiological evolution from the primitive and self-sufficient 

 state of autotrophism toward complete parasitism, brought about by 

 successive losses in the synthetic capacities of microorganisms. Subse- 

 quently Beadle and Tatum (2) showed that by irradiation of Neurospora 

 with X-rays or ultraviolet light, mutants could be produced which 

 showed added nutritional requirements (see also 1,3). The work of 

 Lederberg and Tatum (17,18) and of others (16) showed that similar 

 nutritional mutants could be produced in bacteria, and that the nutritive 

 requirements of such mutants were for the same growth factors as had 

 been previously shown to function in the nutrition of natural populations 

 of bacteria. Through these developments, the theories of Lwoff and of 

 Knight concerning the origin of nutritional requirements in micro- 

 organisms have been given a sound experimental basis. We may now view 

 the nutritional requirements of natural populations of bacteria as having 

 arisen by successive mutations with cumulative effects to produce 

 organisms having requirements of varying complexity. 



Since mutation is a random phenomenon it might be expected that a 

 requirement for growth factors would be found at random among a 

 great variety of bacterial species, and that the identity and number of 

 the growth factors required would vary widely from one species ot 

 bacteria to another, and even from one to another strain within a single 

 species. Such is indeed the case. As one example, the requirements ol 

 various species of the genus Clostridium may be cited. Clostridium 

 butylicum requires only the single growth factor, biotin, for growth in 



