16 



BJ0RN HELLAND-HANSEN 



[REP. OF THE "MICHAEL SARS" NORTH 



13. The Water-Bottles. 



For collecting water-samples from beneath the surface 

 Ekman's reversing water-bottle, Pettersson-Nansen's in- 

 sulating water-bottle, and Nansen's automatic water- 

 bottle were used. 



Ekman's water-bottle was used in by far the greater 

 number of cases from the beginning of the cruise until 

 July 27th (Stat. 96). The exceptions at Stats. 10, 10 A, 

 and 91 have been mentioned in the preceding section. 

 Descriptions and illustrations of the instrument will be 

 found in papers by Ekman [1905] and Helland-Hansen 

 [1912], and we shall here only mention the following. 

 The water-bottle consists of a brass cylinder pivoted on 

 an axis in a metallic frame. The lids at the ends of the 

 cylinder are closed or opened as the cylinder turns, by 

 two pairs of eccentric rods. When closed the cylinder 

 is caught and held by a hook. During the "Michael 

 Sars" Expedition it sometimes happened that the hook 

 did not catch, because the eccentric rods had become 

 somewhat bent, or the rubber plates on the lids were a 

 little out of position. A note was always made in the 

 journals when this happened. By the titrations made at 

 Gibraltar it was proved that the water-samples in such 

 cases were quite worthless, the salinity found being that 

 of the water near the surface. Afterwards no samples 

 were drawn when the water-bottle was not perfectly 

 closed. We had two sizes of the Ekman water-bottle. The 

 larger one was almost always used until it was lost 

 by an accident at Stat. 88. 



The smaller one was used for some few observations 

 before this station, and for Stats. 88-96. At Stat. 96 

 the tube for the reversing thermometer was lost. The 

 water-bottle was, therefore, not used for the later obser- 

 vations. The Ekman water-bottles were constructed for 

 use in series with two or more on the line simultane- 

 ously. On the "Michael Sars" Expedition, however, we 

 used only one at a time, except at some few stations, 

 because it was found that two apparatus together did not 

 work quite satisfactorily in our case. 



A description of the Pettersson-Nansen water-bottle 

 will be found e. g. in publications by Nansen [1901], 

 Ekman [1905] and Helland-Hansen [1912]. It was used 

 at Stats. 10, 10 A, 91, 97—116, (cf. sect. 12). 



Hansen's automatic water-bottle [Ekman 1905, Hel- 

 land-Hansen 1912] was only used at Stat. 13 when the 

 ship lay still. The depths of observation were measured 

 by the meter-wheel, and compared with the indications 

 of the bathymeter in the water-bottle. As this bathymeter 

 proved not to work satisfactorily the water-bottle could 

 not be used when the ship was moving. 



Ekman's and Pettersson-Nansen's water-bottles were 

 operated by means of messengers. From a number of 

 experiments it was found that our messengers fell along 

 the line at a speed of about 4 m. per second. In the 

 journals the time when the messenger was sent off was 

 always noted. 



14. The Water-Samples and Their 



Treatment (Titration). 



The water-samples were mostly collected in glass- 

 bottles with a patent stopper. During the latter part of 

 the investigations we had no more bottles of this kind 

 left, and had to use ordinary medicine bottles of 159 to 

 250 cc. volume with cork-stoppers for some of the 

 surface-samples collected in the last days of July, and for 

 all samples collected in August (/. a. Stats. 97—116). 

 Shortly after such a bottle was filled, the stopper and the 

 neck of the bottle were carefully covered with paraffin 

 wax in order to prevent evaporation from the sample. 



The water-samples were titrated in the way described 

 in "The Ocean Waters" [Helland-Hansen, 1912]. 



Bulb-burettes were used. Some of them were 

 specially made for the Expedition with a division from 

 18-5 to 22 corresponding to the high values of chlorinity 

 found in many of the waters to be examined. The unit 

 of division embraced 2 cc. and the readings could easily 

 be exact within 0-005 units. Titrations of the samples 

 from the first part of the cruise, until Stat. 17, were made 

 ashore in Gibraltar when the "Michael Sars" visited that 

 harbour. The samples from the subsequent stations up 

 to Stat. 34 were examined in La Luz (Gran Canaria) 

 18-21 May, 1910. All other water-samples were titrated 

 in Bergen during the following autumn-and winter-months. 

 By far the larger number of samples from serial 

 observations at stations and a number of the surface- 

 samples were examined twice or more, some of them even 

 5 or 6 times. All samples were titrated by the author of 

 the present paper; the second titrations were to a great 

 extent made by Mr. 1. C. Grondahl. The mean error 

 for the sub-surface samples hardly exceeds 0-01 "/oo of 

 salinity. 



15. Time. Depths of Observations. 



The direct results of the serial observations will be 

 found in Table II (pp. 21*— 33*) where the first column 

 contains the time when the water-bottle was actually closed 

 or the thermometer reversed at the depth of observation. 

 The time is given as local mean time (L. M. T.), being 

 converted from the ships's time (S. T.) which has been 



