PHYSICAL OCEANOGRAPHY 



235 



bottle ; when the water-bottle is reversed the bar is withdrawn, 

 and the messenger is let go. The next water-bottle is knocked 

 over, releasing in its turn the following messenger, and so on. 

 It is indispensable with this, as with all other water-bottles, that 

 when closed it should be absolutely water-tight, otherwise water 

 might get in from the higher layers and vitiate the sample.^ 



The water-sample, when brought on board, may be dealt 

 with at once, and its salinity, etc., determined, but it is generally 

 the best plan to store the samples for examination in a shore 

 laboratory. In this case the samples must be preserved Preservation 

 absolutely air-tight, so that they shall not suffer any change g^nT^^^^s for 

 in the interval. As a rule, the water may be kept in good glass examination 

 bottles with lever stoppers, like those used in soda-water bottles. °" ^^°''^- 

 Cork stoppers will not do, unless capped with paraffin or wax, 

 as it is difficult to avoid some degree of evaporation which 

 would invalidate the results. 



The chemical composition of sea-water has been very care- Chemical 

 fully determined. Wellnigh all known elements are found in o?'^ea°^at°en 

 solution in the sea, but most of them in such small quantities 

 as to be detected only by the most delicate methods. A 

 kilogram of sea-water contains about 35 grams of solid sub- 

 stances altogether ; the quantity varies slightly in different 

 places, but on an average there are about 35 weight-units of 

 solids in 1000 weight-units of sea-water (35 per thousand). 

 According to the results of Dittmar's analyses of the " Challenger" 

 water-samples there are on an average in 1000 grams of sea- 

 water : — 



^ The highest perfection must be exacted with regard to this point. It formerly frequently 

 occurred that the instruments leaked a little ; as the knowledge of the sea has grown, many 



