THE PERMEABILITY OF PROTOPLASM 97 



of Trinity College, Dublin. On evaporation of 20 c.c. 

 only 0-012 per cent, of dry residue was left. The secretion, 

 when examined by the thermo-electric method of cry- 

 oscopy, was found to freeze at almost exactly the same 

 temperature as distilled water. This method, it should be 

 remembered, is differential, as one thermocouple is placed 

 in jthe secretion, and the other in distilled water. Below 

 are recorded the values obtained for the freezing-point 

 and electrical conductivity of the secretion. It will be 

 noticed that the [conductivity is less than that of the 

 Dublin tap-water. The latter is extremely pure and soft, 

 and is obtained from the reservoir in the Wicklow Moun - 

 tains. No trace of sugars could be detected in the secre- 

 tion, even when a large volume was concentrated. A 

 trace of a chloride was, however, found. 



When the tip of the leaf was cut off, it was found that 

 the conductivity of the secretion was at first three times 

 as great as before. But after twenty-four hours it had 

 fallen to something less than twice as great, probably 

 because electrolytes from the injured cells had all been 

 removed. At this stage the conductivity was slightly 

 greater than that of Dublin tap -water, so there is no 

 doubt that there is a separation of electrolytes as the 

 secretion passes out from the uninjured leaf. 



In the unopened pitcher of Nepenthes sp. there is, on the 

 other hand, a secretion of liquid which contains a quite 

 considerable quantity of solutes, mainly electrolytes, as 

 may be seen in the table on p. 98. The liquid lodged 

 in the leaf -bases of Dipsacus also contains appreciable 

 quantities of electrolytes. 



Under A e are recorded the values of the depression of 

 freezing-point due to electrolytes calculated from the 

 conductivity, as explained in Chapter IX. Owing to the 

 small values of A, as directly determined, errors of experi- 



7 



