THE MOLECULAR FORCES IN PLANTS. 153 



each of these eight solutions into small glass vessels, and in each 

 solution place a small strip, about 1 or 2 mm. long, of the epi- 

 dermis of the under side of the midrib of the leaf of Tradescantia 

 discolor. These epidermal cells, which I have found to serve 

 excellently well for the experiment, contain in the cell-sap a red 

 pigment. The plant is at our disposal at any time of the year, 

 which is a matter of considerable importance ; it is to be grown 

 in the hothouse. We cover the vessels, and leave the strips of 

 epidermis in the fluids for two hours at the usual room tempera- 

 ture, and then examine them under the microscope. What has 

 to be determined is, whether the fluids have induced more or 

 less considerable plasmolysis in the epidermal cells, or whether 

 plasmolysis has not yet set in. The commencement of plas- 

 molysis, with which we are specially concerned, is easily recognised 

 when we use the coloured epidermis of Tradescantia discolor. It 

 is characterised by a just visible retraction of the protoplasm from 

 the membranes of the cells; in what follows we use the term 

 "incipient plasmolysis" with reference to a particular object 

 under examination, when the protoplasm has somewhat contracted 

 in about half the cells of that object. 



We shall find that the solution of 0*1 Potassium nitrate, and 

 of O15 cane-sugar, do not induce plasmolysis, but those of 0'15 

 KN0 3 and of 0*25 cane-sugar bring about very considerable plas- 

 molysis. Incipient plasmolysis is observed with strengths of solu- 

 tion between the two, e.g. in solution of 0'13 Potassium nitrate, 

 and of 0'22 cane-sugar. Two such solutions would therefore 

 attract water with equal energy ; both bring about incipient 

 plasmolysis; their isotonic concentration is the same. The values 

 0'22, 013, stand to one another in the ratio of 1 to O591, and if 

 we take the isotonic coefficient of a molecule of Potassium nitrate 

 as 3, that of a molecule of cane-sugar will be T77. We see, 

 therefore, that a molecule of Potassium nitrate exerts a greater 

 attraction for water than does a molecule of cane-sugar. 



H. de Vries has, by this method, determined the isotonic co- 

 efficients of a long series of different substances which occur in 

 the cell-sap. I shall not, however, here proceed further with his 

 important results, but emphatically commend his work to accurate 

 study. 1 



1 See H. de Vries, in Pringsheim's Jahrbiicher f. wissenschl. Botanik, 

 Ed. 14. 



