240 PHYSIOLOGY OF NUTRITION. 



the roots. The plants are now removed from the solutions, each 

 is washed with distilled water (which is then added to the residual 

 fluid in the corresponding vessel), dried with blotting-paper, and 

 weighed. We thus ascertain on the one hand the quantity of 

 water which the plants have lost in the form of vapour, and on 

 the other the amount of water which has gone to form a vital part 

 of the plants. There is a source of error in our experiment, due 

 to the fact that the cotton wool does not close the vessels air- 

 tight, and so a small quantity of water can escape into the air 

 without the help of the plants. It is, however, easy to determine 

 the magnitude of the error, and eliminate it from the result by 

 filling with 100 c.c. of salt solution a few glasses not provided 

 with plants, but merely closed with a cork and cotton wool, and 

 then determining the loss of weight which these undergo during 

 the experiment. Similarly there is, of course, no difficulty at all in 

 taking into consideration the increase in dry weight which the 

 plants undergo during the period of vegetation. Finally, we 

 determine the weight of salt contained in the fluid remaining in 

 the glasses, by boiling to dryriess and weighing the residue. Thus 

 all the data for calculation are before us. From these we find 

 that bean plants absorb from 0'2oO per cent, solutions of Potassium 

 nitrate relatively much water and but little salt, the fluid left in 

 the glasses being therefore more concentrated than the solution 

 originally supplied to the plants (de Saussure's law l ). In contact 

 with O'OoO per cent, or 0'025 per cent, solutions of Potassium 

 nitrate, on the other hand, the plants absorb a comparatively con- 

 centrated solution, the fluid remaining in the vessels being more 

 dilute than the solution originally provided. 2 At all events, then, 

 we have the interesting fact that the roots of plants absorb 

 solutions placed at their disposal not necessarily in the form in 

 which they are supplied, but with a particular quantity of water 

 they take up, according to circumstances, sometimes a smaller, 

 sometimes a larger quantity of salt. In conclusion, we may obtain 

 an approximately correct result if we grow bean seedlings, as 

 above, in glass vessels holding 100 c.c. of Potassium nitrate solu- 

 tions of different strengths, and without further weighings simply 

 determine the amount of salt in the fluids left in the vessels after- 

 half of the original solution has been absorbed. 



1 See de Saussure, Recherches sur la vegetation, 1804, p. 247. 



2 See W. Wolf, Venuchsstationen, Bd. 6 and 7. 



