298 rirrTEii ssos . os waticu asii i en. 



We must now try to get an idea of the bearing of the 

 above results ' on the phenoinena, which really take placo in 

 the arctic sea. On plate 21 we will find representations of the 

 raost prominent kinds of polar ice. I have ventured to estab- 

 lish the following types of ice, partly from their manner of 

 formation, partly from the resemblance in their chemical com- 

 position shown by titration. We may understand |he red line 

 III approximately to designate the volumes of fresh water-ice, 

 as for example the glacier-ice. The changes in the volume of 

 such ice are most uniform and regular, and therefore the ice- 

 blocks and icebergs arriving from the ice fiords of Greenland 

 will conserve their form and size longer than any other kind 

 of ice, in spite of the abrupt changes of atmospheric tempera- 

 ture, which the icebergs must be exposed to, incomparably 

 more than all other kinds of ice. This ice will survive the 

 mouldering process, which incessantly frets upon such ice as 

 contains a minimal quantity of salt, even at low temperatures, 

 and will succumb only to the infiuence of a warmer atmos- 

 phere or to the corrosion of the warm water of the Atlantic 

 upon its l^ase. But we may see from the abruptly sloping 

 branch of curve III, that just in the vicinity of the melting 

 point destruction rapidly approachos the solid framework of 

 the icc-colosses; an extremcly littlo variation iu temperature 

 will cause expansions or contractions in the mäss, which may 

 suffice to carve out and split tho gigantic structures into 

 fantastical shapes. 



The Curve IV will give us a representation of the vol- 

 umes of old hay-ice and of the freshly formed ice of brackish 

 water [see page 308 & 309]. Also the ice of the torosses or 

 hummocks, which has been long exposed to the air, Ijolongs to 

 this tyi)e (see the titrations on page 308 by Nordenskiöld, 

 Palander, Almqvist a. O. on samplcs of toross-ice), Accord- 

 ing to the analytical determination on page 307 the constitu- 

 tion of sample V accords well with that of the neiv hay-ice, 

 which successively formed during the winter at the coast of 

 Pitlekaj. The titrations by Nordenskiöld a. O. indicate a 

 percentage of chlorine of about O.n.. in such ice. We may 

 therefore consider V to represent the type of ice, which in 

 the winter covers the bays and fiords of the arctic sea. In 

 the sample VI we meet with an exact counterfeit ^ of that 



* Owing to the system of observiition, adopted in the experimental 

 series V, B and VI, B, the arrangement of tlio numhcrs ennnierated in the 

 , tables is soincwhat different from the usual. 



'^ According to its manner of formation (see page 2'Jl). 



