THE PERMEABILITY OF PROTOPLASM 99 



halophytes in sodium chloride solutions for considerable 

 periods. 



In preliminary experiments leaves were immersed in 

 sea-water, and after some hours sections were cut with 

 a dry razor and mounted in g silver nitrate. These were 

 then exposed to the light of an electric lamp for twenty 

 minutes at a distance of one foot. Leaves so treated 

 always showed a much more intense reaction for chloride 

 than similar leaves examined as a control. The treated 

 leaves were in every case thoroughly rinsed before testing. 

 The chloride appeared in greatest quantity in the upper 

 and lower epidermis and in the palisade cells. This ob- 

 servation shows that the quantitative results, to be quoted 

 shortly, are not explicable by the entrance of salt solution 

 into the intercellular spaces. 



The leaves used for the quantitative determination of 

 chloride were thoroughly washed and dried at 100 C. for 

 an hour and a half. They were subsequently incinerated 

 at a temperature sufficiently low to avoid any appreciable 

 volatilization of sodium chloride. Titration was then 

 carried out in the usual way with silver nitrate after neu- 

 tralization of the slightly alkaline ash with nitric acid. 

 The same individual plants were used throughout all the 

 experiments, and the leaves were kept in darkness, between 

 the successive weighings for water determinations. In 

 very case the leaves were cut freshly from the plant and 

 their petioles sealed with paraffin wax. At three-hour 

 intervals they were removed from the salt -solutions and 

 weighed after drying quickly. By this means both the 

 variations in total weight of the leaves were studied, as 

 well as the final change in chloride content. 



The subjoined table shows the amount of sodium chloride 

 in the ash, calculated as a percentage of the dry weight of 

 the leaves. Three experiments with each type of leaf are 



