48 



ions must be excluded. Hence the conclusion seems unavoidable that 

 the carbonic acid (HCO3) ions produce the stimulating effect, improb- 

 able as this would appear. To what agency should be ascribed the 

 characteristic toxic action (so different in kind from that of sodium 

 sulphate and sodium chloride) of stronger solutions of sodium bicar- 

 bonate, in which no free hydroxyl could be detected, is a question to 

 which no answer can at present be given. ^ 



None of the other salts with which experiments were made in pure 

 solution were shown to stimulate elongation of the roots, although 

 the possibility is not excluded that solutions still more dilute than 

 those employed will give i^ositive results. Magnesium sulphate was 

 found to be indifferent (neither toxic nor stimulating) at 0.0003125 

 normal, magnesium chloride at 0.000625 normal, sodium sulphate 

 (nearl}^) at 0.002 normal, and sodium chloride^ (approximately) at 

 0.005 normal. 



These observations accord with the well-known principle that many 

 violent poisons, if given in sufficiently minute doses, serve as benefi- 

 cial stimuli. Familiar examples are the action of arsenic, mercury, 



'In experiments with sodium carbonate and sodium bicarbonate as to their 

 effect upon animals, Loeb encountered a very similar anomaly. The stimulating 

 effect of various hydrates upon the absorption of water by a muscle immersed in 

 a sodium chloride solution was shown to be clearly due to the hydroxyl ions, 

 being eciual in amount when equivalent solutions of hydrates were used, irre- 

 spective of the character of the basic ions [see Pfliiger's Archiv, 69, 10 (1898)]. 

 The similar effect produced by carbonates of sodium and potassium was ascribed 

 to the same factor, the hydroxyl ions (1. c, p. 20). On the other hand, the effect of 

 sodium carbonate (NaaCOa) in stimulating skeleton formation in the pluteus of a 

 sea urchin appears to be due to the carbonic acid (HCO3) ions, since sodium in 

 other forms, as well as hydroxyl in the form of potassium hydrate, gave negative 

 results [Am. Journ. Physiology, 443, (1900)]. 



- Pfeffer [Pflanzenphysiologie, Ed. 2, 1, 425] observes that possibly chlorides {e.g., 

 sodium chloride), like so many other substances, act in dilute solutions as chem- 

 ical stimuli. Storp [Biedermann's Centralbl. , 13, 76 ( 1884)] obtained a stimulating 

 effect upon the germination of seeds by immersing them in a 0.01 per cent solution 

 of sodium chloride. Jarius [Landw. Versuchsst., 32, 149 (1886)] found that 

 even a 0.4 per cent solution of sodium chloride stimulated the germination of seeds 

 of Vs^heat, rye, rape, maize, bears, and vetches. Jones and Orton (Bui. Vermont 

 Agrlc. Exp. Station No. 56, p. 12) observed, as a consequence of the application of 

 sodium chloride to land in order to exterminate the weed known as Orange Hawk- 

 weed {Hieraciam anrantiacum) ,& marked stimulating effect upon the growth of 

 grass in the same field. Peligot [Comptes rendus, 73, 1078 (1871)] suggests that 

 the stimulatmg effect upon field crops which is sometimes obtained with sodium 

 chloride may be due to its facilitating the decomposition of calcium phosphate 

 and thus increasing the amount of phosphoric acid at the disposal of the plant. 

 Kellner [Landw. Versuchsst., 32, 365 (1886)] attributes to a similar liberation of 

 phosphoric acid the stimulating effects of iron sulphate upon plant growth recorded 

 by Koenig and by Griffiths (see p. 49). Reveil [De Taction des poisons sur les 

 plantes,p. 41 (1865)] found that sodium hypochlorite in a solution of 0.1 per cent 

 stimulates germination and growth, but is injurious, especially to herbaceous 

 plants, when applied in greater concentration. 



