24 BOTANICAL GAZETTE [JANUARY 
ends were within a few millimeters of the root-tips, there being a 
tube for each root. Every few hours the position of the tubes of 
potassium nitrate was changed in order to follow the descent of 
the root-tips. This method will be recognized as an adaptation 
of that method first employed for other purposes by Pfeffer.’ 
Altogether eighty-seven seedlings of Helianthus annuus L. and 
seventeen seedlings of Raphanus sativus L. were employed, for 
periods ranging from twenty-four to forty-eight hours at 20° to 
24° C., but without showing a response. The positive curves 
and the negative curves were no more numerous than one might 
observe in these roots growing in water. 
Thinking that the amount of potassium nitrate might have 
been too small to act as a stimulus, glass vials holding 20° were 
used instead. The mouths of the vials were closed by plugs 
of cotton previously wet with the potassium nitrate solution. 
Here also the roots showed no curves that could be ascribed to 
the chemical. 
The absence of response in the foregoing experiments might 
be ascribed to any of several conditions. It might be that the 
roots used were not chemotropic, though others might be. It 
might be that the sunflower and the radish were chemotropic to 
some chemicals, but not to the potassium nitrate; though the 
thought suggested itself at the outset that, if roots are chemo- 
tropic they are likely to respond to chemicals which form their 
necessary food. It might be that the potassium nitrate was not 
present in large enough proportion. It might be that the differ- 
ence in chemical composition was not great enough on opposite 
-sides of the root. It might be, finally, that, while the seedling 
had a full supply of all kinds of stored food, it would show 
itself indifferent, but would respond, when it needed mineral 
food from without. Manifestly the thing to do was to satisfy 
these conditions as fully as possible. 
Il. THE ONE-SIDED APPLICATION OF A CHEMICAL BY: DIFFUSION 
THROUGH A MEMBRANE. 
With a view to testing the roots of plants when the food 
stored in the seed had been exhausted, water cultures were 
* Ueber chemotactische Bewegungen, etc. Untersuch. Bot. Inst. Tiibingen 1 : 367. 
