§4] 



TONOTAXIS 



9 



were repelled. The numbers at the heads of the columns ar 



the different values of n in the formula, -—tt: M W % ■ Sine 



1000 



the different solutions were made up on the basis of molecula 

 weights, all solutions of a given concentration contained th 

 same number of molecules. 



TABLE XII 



From this table it appears that, as a rule, solutions of y^ 

 MW fo and over are repelled, while those of xoVo ^^ unde: 

 except in the case of KCN, permit the free migration of th 

 bacteria into the tube. 



In the case of those substances whose isotonic coefficient : 

 4, solutions of j-^\^ MW % and over always repel, and those ( 

 Y^Q-Q in the majority of cases permit free migration. In tt 

 case of those substances whose isotonic coefficient is 2, solutior 

 of over y^|-o MW % repel, and those of under y^t ^suall 

 permit free migration. The solutions at which repulsion jui 

 occurs in the three cases are in the ratio 10 : 7 : 6 ; which ; 

 nearly the same ratio as the reciprocals of the isotonic coe: 

 ficients, which, multiplied by 2 run, 10 : 6.6 : 5. Thus th 

 conclusion seems justified that the repelling action of thes 

 substances is proportional to their isotonic coefficients, and ii 

 therefore, probably osmotic in its nature. 



In a second work, Massakt ('91) has studied this matte 

 with the aid of new methods. A drop of sea water containin 

 bacteria is prepared as before, on a cardboard ring, but, in plac 



