OBTAINING WATER FOR BACTERIOLOGICAL ANALYSIS 105 



dota. It is a modification of the apparatus used by Russell (1892) 

 in his investigation of sea water. In the original form, the sam- 

 pler required two lines for manipulation, but was later changed by- 

 Russell so that it required only one. In the survey of Lake Men- 

 dota the Russell sampler was tried, but, the frame being of wood, 



Fig. 1 



it required a large weight which made it cumbersome to handle. 

 The modification of this sampler is simple and overcomes many 

 of the disadvantages of the older type. This apparatus with the 

 exception of an exhausted glass tube is made entirely of heavy 

 sheet brass (see fig. I) . The sterilized, exhausted tube is attached 

 to the holder by means of a rubber covered spring clamp, D (see 

 II, fig. 1,). A brass arm, C, operates between the connecting line, 



