THE KETURN OF DAYLIGHT. 251 



Referring to my remarks on December 14th, in rela- 

 tion to Weypreclit saying, '^ Beginning at a certain 

 thickness the ice is ahnost free from salt," it may be 

 as Avell to state here the result of our examination of 

 the ice in which we are drifting. A piece of floe ice, 

 formed from direct freezing, and three feet nine inches 

 thick, was selected for examination. The following are 

 Dr. Ambler's figures for the result of his test (Parke's 

 test). (See Appendix E.) 



Number of grains of salt per gallon of sea- 

 water equal 2045. 



Number of grains of salt in cube out from 



upper five inches of our block .... 548.06 



Number of grains of salt in cube, cut from 



lower five inches of our block .... 347.25 



Number of grains of salt permissible in po- 

 table water 10. 



From which it will be seen that the "certain" thickness 

 has not been attainable by us, for we cannot find a sin- 

 gle piece of floe from which we can get potable water, 

 and since it seems never to snow up here we have to 

 distill every drop of water we drink. If, as Weyprecht 

 says, the salt is all crystallized out during the winter 

 and washed off during the summer, the upper layers 

 of old ice remaining ought to be fresh ; but in our ex- 

 perience they were as salt in September last as the 

 new floes are salt now. We may be having phenom- 

 enal ice, but I hardly think so. If all, or nearly all, 

 the salt resulting from the freezing of sea-water comes 

 to the surface as efflorescence, and is w^ashed off into 

 the sea during the following summer; and if the ex- 

 posed upper ice then melts by the action of the sun's 

 rays and is in its turn frozen in the fall, squeezing out 

 again a small residuum of salt, I can understand that 



