144 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, xvi., 



appear to be a thinning away of the toxin, for the wandering- 

 colony generally becomes heaped up at the margin of the toxic 

 zone and then spreads round and eventually encircles the colony. 

 It may be that the thickening of the edge is caused by the toxin 

 becoming so weak as to be able to exercise a stimulating action 

 like other weak poisons."^ The Actinomyces-colonies are gener- 

 ally toxic to such wandering moulds and bacteria. 



As soil contains so much sand and inert matter, it did not 

 appear to be a suitable medium for experiments with these pre- 

 sumable toxin formers, and it was considered that a fluid medium 

 would be better. With regard to the nutrient added to the 

 water, Bottomley's work with auximones had suggested the idea 

 that there might be a relation between them and decayed soil- 

 toxins. Bottomley used moist peat-moss as a culture-material, 

 but, as this could not be obtained, the first experiments were 

 made with washed sphagnum-moss suspended in water. 



The moss was picked, dried, cut up, washed, and again dried. 

 Ten grams were put into a flask with 600 c.c. of tap-water and 

 sterilised. 



Two bacteria, T.P.2 and T.P.4, were selected as being possible 

 toxin-producers. They were not identified at first, for, in ex- 

 periments such as these, one determines the value of the bacte- 

 rium first and identifies it afterwards, if its activity justifies the 

 identification.! The bacteria were seeded into the flasks of sus- 



* Journ. Roy. Soc. N. S. Wales, 1916, p.77. 

 t Bac. T.P.2 is a short, motile rod measuring 0"5 : 2yut. Gelatine is 

 liquefied slowl3% the colonies being round, white, and slightly raised with a 

 ciliate edge. On agar, the growth is ivory-white, rough and cohesive. 

 Long, needle-shaped crystals are quickly formed in agar containing sodium 

 phosphate, and these are fairly characteristic of the organism. In some 

 media, slime is formed from dextrose, althougli none is produced on 

 Lipman-Brown agar. In bouillon, the groM th produces turbidity, a film 

 and a sediment; nitrates are not reduced. Milk is coagulated and slowlj' 

 peptonised. On potato, the growth is yellow-brown. The indol test is 

 positive, the Gram test negative, and the bacillus produces neither gas or 

 acid from dextrose, saccharose, mannite, or glycerine. Nitrate is not pro- 

 duced from ammonia salts. It appears to have Bac. imincHis as its closest 

 ally. 



