Chapter III — 27 — Collecting Samples at Sea 



of a steamer. This procedure is open to question because some of the 

 sedentary organisms which grow in and around the external orifice of the 

 sea cock will probably enter the sample. Most submerged surfaces, in- 

 cluding ships' bottoms and water conduits, are covered with a profuse 

 growth of plant and animal life along with large numbers of bacteria. 



Metal cylinders for collecting samples : — Some of the pioneer work 

 has been done on water samples collected with water bottles designed by 

 hydrographers for collecting water for chemical analyses. Essentially 

 Ekman, Nansen-Knudsen, Green-Bigelow, Sigsbee, and similar water 

 bottles (Committee on Oceanography, 1932) are open metal cylinders 

 through which the water flows while the device is being lowered to the 

 desired depth. When a messenger is dropped down the hydrographic wire 

 by which the bottles are lowered, the ends of the cylinders are closed by 

 various kinds of valves thus entrapping a sample of water. 



In general, bacteriological results obtained with samples collected with 

 such apparatus are of questionable validity. In the first place, they may 

 contain contaminating organisms. Secondly, most of the metals used in 

 the construction of hydrographic samplers have a bacteriostatic or bac- 

 tericidal effect in sea water (page 31). Although glass bottles were used 

 for collecting some of their surface samples, Fischer (1894a), Gazert 

 (19066), and many other workers used metal cylinders for sub-surface 

 samples. This must be taken into consideration in the appraisal of their 

 quantitative results. 



Otto and Neumann (1904) collected samples for bacteriological anal- 

 ysis in a metal cylinder, the ends of which could be closed with rubber 

 gaskets by dropping a messenger at the desired depth. A nickel-plated 

 cylinder with cocks operated by a messenger was used by Bertel (191 1). 



None of the cylinders which go down open can be relied upon to bring 

 up samples which are uncontaminated by adventitious bacteria. Even 

 if the cylinders are first sterilized, which is not always practical, they may 

 be contaminated by the water passing through them while being lowered. 



Mathews (1913) sought to obviate this difficulty by filling a glass- 

 lined cylinder with 95 per cent alcohol to provide for the sterility of the 

 apparatus as well as for the equalization of pressure inside and outside of 

 the closed cylinder. One messenger was dropped to open the ends of the 

 cylinder, which permitted the disinfectant to be flushed out by the water. 

 Then, after allowing sufficient time for the disinfectant to diffuse away, a 

 second messenger was dropped to entrap a water sample at the desired 

 depth. Drew (1914) used this apparatus with expressed confidence for 

 collecting samples for bacteriological analysis to depths of 800 meters. 

 Others have used it or modifications with the cylinder filled with phenol 

 solution or other disinfectants. 



Besides being somewhat complicated and inconvenient to use, there is 

 always the possibility of the disinfectant diffusing out prematurely 

 through a faulty valve. With a slightly compressible disinfectant such as 

 95 per cent alcohol, sea water would be forced into the cylinder through 

 nearly any kind of valve feasible for use on such an apparatus. On the 

 other hand, all of the disinfectant may not be washed out of the cylinder 

 during the time the valves are opened and closed. 



Young et al. (1931) fitted a phosphor-bronze cylinder with a brass 

 piston which aspirates a water sample when the piston is pulled out by a 

 messenger-activated mechanism. A similar device made of a loo-ml. 



