TRANSACTIONS OF SECTION E. 819 



replaced by a convex curve of weathering on tlie upper parts of bills, with 

 herbaceous veg-etation in temperate regions, and often entirely replaced by a 

 convex curve in tropical regions, where the sides of the hills are clad with forest 

 growth. 



The dissection of such round-topped hills or moels by stream action is then 

 considered, and it is pointed out that buttress-like lateral peaks will be formed 

 around the resultant central peak. Lateral peaks of this nature have been 

 described by Mr, I. 0. Eussell on Mount Rainier, under the name tahojnas ; he 

 gives roason for their production by glacial denudation in that particuhir case. 



I 



6. On the Pettersson-N^ansen Insulating Water-bottle. 

 By Hugh Robert Mill, D.Sc, LL.D} 



Professor Pettersson has, in conjunction with Professor Nansen, completed a 

 modification of his well-known apparatus for obtaining samples of sea-water 

 without change of temperature, A specimen of the improved water-bottle con- 

 structed by Messrs. Ericsson, of Stockholm and London, was exhibited. The 

 purpose of this apparatus is to enclose a quantity of sea-water at any desired depth, 

 to hold it securely, and to bring it to the surface without any change of tempera- 

 ture exceeding one hundredth of a degree Centigrade. The previous form of insu- 

 lating water-bottle was found by Dr. Nansen in his arctic expedition to be less 

 trustworthy at great depths than in shallow water ; hence the suggestions which 

 resulted in the new apparatus. The insulation, which is the essential feature of the 

 water-bottle, is secured by a series of concentric chambers of non-conductin"- 

 material which are simultaneously filled with water, and so protect the portion, 

 measuring about two litres, which occupies the large central tube. The walls of 

 the inner tubes are so constructed as not to become heated by compression at 

 the greatest depth. This is secured by using metal, which is heated by com- 

 pression, and indiarubber, which is cooled by compi-ession, in such proportions as 

 to ensure constancy of temperature for the whole structure. 



The water-bottle when set is held apart, so that the base, sides, and lid are 

 separated, and the water passes freely through the tubes as the apparatus descends. 

 When the apparatus is being drawn up a propeller (which during the descent 

 revolves freely) engages with a screw and releases a heavy weight, which closes 

 and locks the whole rigidly togethei'. An arrangement is provided for the relief of 

 pressure as the included water expands on being hauled up, The temperature is 

 ascertained by a thermometer, protected againt pressure, enclosed in the central 

 tube, and projecting sufficiently far to be easily read. If preferred, the aperture for 

 the thermometer may be closed by a screw and the thermometer inserted when the 

 water-bottle is brought up. A reversing thermometer to give the temperature of 

 the water independently may be attached to the upper part of the water-bottle, 

 and is set in action at the moment of closing. The whole apparatus weighs about 

 581b., and is used on a wire line and worked by a steam winch. 



During August of this year the improved water-bottle was tested by Professor 

 Nansen on board the Michael Sars in the sea between Iceland and Spitsbergen, and 

 at the greatest depth met with (3,000 metres = 1,670 fathoms) the insulation was 

 perfect. On August 11a sample was taken from 3,000 metres, and when it came 

 up the thermometer read : l°-285 C, after five minutes l°-283, after nine minutes 

 l°-270, and after eleven minutes 1°-210. On August 13 in a sample from 2,000 

 metres the thermometer showed 1°-135, after five minutes 1°-135, after six minutes 

 1°-130, and after eight minutes 1°-110. Professor Nansen considers it essential 

 to use an included thermometer, 



' Published in full in the Geographical Joximal, vol. xvi. (1900), pp, 4G9-471. 



8q2 



