HYDROGRAPHY IN PASSAMAQVODDY BAY 301 



SESSIONAL PAPER No. 38a 



A necessary condition to make a determination of the salinity of sea-water is to 

 secure a sample of water collected at a certain date in a certain place, at the surface 

 or at a known depth, which is guaranteed free from mixture with different water and 

 which has in no way evaporated. 



The surface water can be collected in a bucket and hauled up. The glass bottle 

 in which the water is to be preserved for analysis is rinsed with a portion of the 

 sample, then filled, well stoppered and it can be kept as long as the bottle is almost 

 completely filled and hermetically closed. 



From July 14 to July 25 my samples were kept in Imperial pint bottles; after 

 the latter date I used citrate of magnesia bottles. I took the temperature of the sur- 

 face from the water in the bucket by means of a Centigrade thermometer graduated in 

 tenths of a degree and whose accuracy I had verified beforehand. 



To collect samples from below the surface, a great number of methods have been 

 invented. At first, an ordinary stoppered bottle was sent down to a certain depth by 

 means of a weight, and, at the desired depth, the bottle was opened and filled Avitb 

 water by pulling a cord attached to the stopper. In drawing it up, very little water 

 from the surface layers could mix with the saxiiple. 



The Petterson-Nansen bottle, which we used for collecting our samples from July 

 14 to July 25, can isolate a sample of water at any depth. This bottle is sent down 

 open, the lid being suspended in the upper part of the frame and held by a spring. 

 We used the reversing thermometer attached to the frame of the bottle. We left the 

 bottle at the desired depth for five minutes so that the thermometer could have time to 

 accurately mark the temperature of the water in situ. A messenger was then sent 

 down along the wire ; this messenger unhooks the lid ; the weight, which hangs below 

 the apparatus, clasps the whole thing together and closes the bottle. This is composed 

 of a series of metallic cylinders to insulate the water and a thermometer can be placed 

 on the inside; this thermometer, which, however, is but slightly affected by varying- 

 temperatures as the bottle is pulled up, was not used in our determinations.' 



When we used the Petterson-Nansen bottle, the depth was taken in fathoms, as 

 the meter-wheel had not arrived at the station, but, in my tables, the fathoms are 

 expressed in metres. 



From July 25, we used the Nansen reversing bottles for collecting our samples 

 and the meter-wheel or determining the depth. The Nansen bottle has attached to 

 it a thermometer which is tipped over with the bottle by means of a messenger. We 

 allowed this bottle to remain at least three minutes in the water before pulling it up 

 for a reading. A number of these bottles can be fastened along the line; a messenger 

 is hooked below each bottle, except the lowest one ; this messenger is released when the 

 bottle is tipped over by means of a messenger sent from above; the result is that the 

 next bottle is reversed ; this releases another messenger and so on. By this apparatus, 

 a number of samples can be taken at the same time at different depths and the bottles 

 are not so heavy and clumsy as the Petterson-Nansen bottle. 



The samples of water collected must afterwards be analysed. In such analysis the 

 halogens are titrated with silver nitrate and the results given as grams of chlorine per 

 thousand grams of water. 



We have seen that there are many substances in sea-water, and, though the pro- 

 portion of salts varies from one place to another, the relative proportion of the diiferent 

 elements is about the same everywhere; thus, when thie quantity of chlorine has been 

 accurately determined, we have the proportion of total salts in the sample examined. 

 Mohr's method is used for the determination of chlorine. If a neutral or slightly 

 alkaline solution of a chloride, bromide or iodide, in which there is a little potassium 

 chromate comes in contact with a neutral solution of silver nitrate a white precipitate 

 is formed as long as there is a trace of halide in solution. Thus, in sea-water, the 

 bromine and small amount of iodine present are precipitated along with the chlorine, 

 but the whole is calculated in grams of chlorine per thousand grams of water. As soon 



