340 



least possible influence upon the alkalinity of the water. A 

 little air should be left above the water in the bottle, when it 

 is not immediately analysed. 



The taking of samples from the surface presents no diffi- 

 culties whatever. Of course the samples, especially of fresh- 

 water, should not be unnecessarily 

 exposed to the air. 



For the taking of samples from 

 shallow depths up to 100 metres 

 I have found the following method 

 very convenient: The bottle is 

 stoppered with rubber and provided 

 with the two tubes (presently to 

 be described) which are applied 

 by the actual determination of ten- 

 sion. The taps are kept open, and 

 the tubes are connected above as 

 seen in the adjoined figure. The 

 short piece of rubber-tubing fitting 

 on to the vertical tube Ъ must be 

 thick-walled and should not be 

 pressed down loo hard. The whole 

 apparatus is lowered to the appointed 

 depth and then the lead p is 

 allowed to drop. It is easily seen 

 from the figure that the bumping of 

 the lead will cause the disconnection 

 of the tubes. The water will rush 

 in through a while the air escapes through 6, and after a 

 couple of minutes the sample may be hauled up and analysed. 

 If it is read immediately the thermometer in the bottle will show 

 with an accuracy of about Vio degree the temperature at the 

 depth examined. 



At greater depths than 100 m. a metallic water-bottle that 



Fig. 2. 



