— XLI — REE'ORT 190^-03 APPEND. A 



and furthermore, the C. L. has in July of this year concluded a contract with 

 C. Richter, Berlin, by which the former should have a steady supply of thermo- 

 meters kept ready for its disposal. It may be pointed out, however, that it is 

 always advisable to order instruments through the laboratory, at as long notice 

 as possible. 



III. Control analysis of water samples. 



Samples have been sent to the G. L. from one expedition only, in March, 

 and from two in June. 



It may be pointed out in this connection, that the object of these control 

 analyses — to secure comparability of the results and to detect possible real diffe- 

 rences due to the method of keeping, or of analysis or to other causes — is 

 frustrated if samples are not sent from all the participating expeditions. Specially 

 suitable bottles for the preservation of such samples-, have been procured by the 

 G. L. and sent to the several expeditions. 



IV. Preparation of standard-water. 



A provisional supply of about 100 tubes of standard water (VI, b) has 

 been prepared in July 1903, and has been ' standardised from M. Knudsen's standard 

 water VI a. In this work the G. L. has had the benefit of Decent Knudsen's experience, 

 and for the advice which he has very kindly placed at its disposal, the C. L. here 

 gratefully records its acknowledgment. 



Steps have been taken for the preparation of a primary standard water to 

 be compared directly with Knudsen's water VI. From this primary standard it is 

 intended that the subsequent supplies of standard water should be standardised. 



A detailed report on the preparation and analysis of the water VI b and 

 of the primary standard water, will be made the subject of a subsequent com- 

 munication. 



V. To test and improve the apparatus in use for determining the tempe- 

 rature of the sea. 



In the autumn of 1902 an examination of the properties of the Pettersson- 

 Nansen water-bottle was made. An attempt to find the principles according to 

 which a water-bottle of a certain size must be constructed to give the best pos- 

 sible insulating power, furnished data, which will be made the subject of a more 

 detailed communication. It was found that the insulating power could be increased 

 by combining water-jackets and solid walls of good insulating material. This 

 arrangement would, however, have other inconvenient consequences — especially 

 will the change of temperature of the solid walls, caused by the great decrease 

 of pressure when the bottle is hauled up, cause an error in the temperature 



