148 



SCIENCE. 



[N. S. Vol. XXII. No. 553. 



unpublished data from experiments which he 

 has recently performed. He finds from an 

 average of six different tests in which ferric 

 hydrate was added to the nutrient solution 

 described above, the latter having a concentra- 

 tion of 75 parts per million, that growth is 

 accelerated by this treatment to an extent 

 equivalent to 26.2 per cent., the growth ob- 

 tained in the untreated solution being consid- 

 ered as unity for the comparison. The same 

 nutrient solution with carbon black gave 35 

 per cent, increase in growth on the same basis. 

 The last figure is an average of the results of 

 two experiments. 



Table IV. 

 Data for Experiment XII. 



The data of experiments I. to VI. of the 

 present paper (Table I.) show that the average 

 growth in the nutrient solution of 75 parts 

 per million is 161.6, and for the same solution 

 of a concentration of 155 parts per million 

 the average growth is 188.7. On the average, 

 the latter concentration is the optimum for 

 wheat growth as nearly as this can be approxi- 

 mated from the series, so that the acceleration 

 which it is possible to obtain by increase in 

 concentration is 188.7 — 161.6/161.6, or 16.8 

 per cent. Considering only the four experi- 



ments which showed ail increase in favor of 

 the stronger of these two concentrations (ex- 

 periments L, II., III. and VI.), this average 

 is, of course, much higher, being 27 per cent., 

 or very nearly the same as the increase ob- 

 tained by treating the weaker solution with 

 ferric hydrate, and considerably less than that 

 obtained with carbon. Thus we are con- 

 fronted with a case where two entirely dif- 

 ferent treatments bring about the same effect 

 upon the plant. It is practically proved that 

 the insoluble bodies have their efi^ect here by 

 removing from solution the deleterious excre- 

 tions of the plant roots. The efi^ect of in- 

 crease in concentration may be explained by 

 one or inore of the three following hypotheses : 

 The higher concentration may make the plant 

 more resistant to the poisons; it may actually 

 prevent the excretion of such poisons from the 

 roots; or with higher concentration of salts 

 the poisons themselves may be altered so as to 

 lose their toxic properties. Which of these 

 explanations is correct can not be decided 

 now, but it is at any rate very clear that the 

 acceleration observed has no direct connection 

 with the nutrient value of the medium. 



In soil or sand cultures the effect of con- 

 centration is known to be very different from 

 that in water cultures; for instance, the con- 

 centration best suited to wheat in water cul- 

 ture is about 300 parts per million of nutrient 

 solution, while in sand it lies in the vicinity 

 of 2,500 parts per million. To investigate 

 the question whether the effect of strength of 

 solution in sand is due to physical concentra- 

 tion or to chemical conditions of nutriment, 

 several series of sand cultures were carried 

 out. 



Pure quartz sand was placed in paraffined 

 wire baskets of the form described in Bulletin 

 No. 23 of this bureau, and the hardened par- 

 affin at the bottom was punctured with pin 

 holes to allow free drainage. In experiment 

 X. six wheat plants were grown in these 

 baskets for sixteen days, the sand being 

 flooded daily with nutrient solutions of con- 

 centrations of 15, 75, 750 and 1,550 parts per 

 million of total salts, respectively, while the 

 excess of solution was allowed to drain out 



