THE BEHAVIOR OF BACTERIA 33 



in weak solutions of ether. From the method by which the gatherings 

 are produced, it is, of course, evident that collection in any agent signi- 

 fies merely that the organisms are less repelled by this agent than by 

 the surrounding conditions. All such collections are doubtless to be 

 conceived as brought about by a reversal of the movement on passing 

 from the dilute chemical to water containing none of the chemical. In 

 many cases this has been determined by direct observations ; x in other 

 cases the observations have not been made. 



If the chemical is stronger, the reversal of movement is produced 

 when the bacteria come in contact with it, so that strong chemicals as 

 a rule remain empty. Thus the same chemicals that, when dilute, pro- 

 duce a "positive reaction" cause, when stronger, a negative reaction. 

 All substances in dilute solutions of which Spirillum gathers are avoided 

 if stronger solutions are used. Miyoshi found this to be true also for 

 Chromatium iveissii; and it is indeed a general rule for bacteria. 



Why should the bacteria avoid strong solutions of the very substances 

 that when weak are "attractive"? It is, of course, well known that 

 strong solutions are as a rule injurious; the negative reaction is there- 

 fore distinctly adaptive under these conditions. Even when we can see 

 no use for the positive reaction, as in the case of the collecting of Amylo- 

 bacter in a solution of ether, we find that the reaction becomes negative 

 as soon as the solution becomes injurious. Amylobacter keeps out of 

 stronger solutions of ether. 



Yet the bacteria are no more infallible in detecting injurious sub- 

 stances than are higher organisms. If a poisonous chemical is mixed 

 with a solution in which the bacteria naturally collect, the organisms may 

 continue to enter a drop of the solution, where they are killed. So 

 Pfeffer (1888, p. 628) found that if to an attractive solution of 0.019 P er 

 cent potassium chloride be added 0.0 1 per cent mercuric chloride, Bac- 

 terium termo and Spirillum undula continue to pass into the solution, 

 though they are there immediately killed. Bacterium termo swarms into 

 solutions of morphine (morphium chloride), where after ten minutes to 

 an hour all motion ceases. 



To just what action of the strong solution is the repellent effect, 

 when it occurs, due? Strong solutions may be injurious from two dif- 

 ferent classes of causes. The specific properties of the given chemical 

 may cause injuries when acting intensely, and this might induce the 

 negative reaction. But farther, in any strong solution the osmotic press- 

 ure is high, and this produces injury in organisms by withdrawing the 



1 The reversal of motion under these circumstances has been described especially by 

 Pfeffer (1884), Rothert (1901), and Jennings and Crosby (1901). 



D 



