IF£ MAKE A START I2I 



obliged to pass before it could penetrate farther and reach 

 the bottom of a flat snow-melter, which was placed above 

 it. Then, after having delivered some part of its heat, 

 the air was forced down aoain on the outside of the ring;- 

 shaped vessel by the help of a mantle, or cap, which sur- 

 rounded the whole. Here it parted with its last remain- 

 ing warmth to the outer side of the ring-vessel, and finally 

 escaped, almost entirely cooled, from the lower edge of 

 the mantle. 



For the heating was used a Swedish gas-petroleum 

 lamp, known as the "Primus," in which the heat turns the 

 petroleum into gas before it is consumed. By this means 

 it renders the combustion unusually complete. Numerous 

 experiments made by Professor Torup at his laboratory 

 proved that the cooker in ordinary circumstances yielded 

 90 to 93 per cent, of the heat which the petroleum con- 

 sumed should, by combustion, theoretically evolve. A 

 more satisfactory result, I think, it would be difficult to 

 obtain. The vessels in this cooker were made of German 

 silver, while the lid, outside cap, etc., were of aluminium. 

 Together w^ith two tin mugs, two tin spoons, and a tin 

 ladle, it weighed exactly S pounds 13 ounces, while the 

 lamp, the " Primus," weighed 4^ ounces. 



As fuel, my choice this time fell on petroleum (" snow- 

 flake "). Alcohol, which has generally been used before 

 on Arctic expeditions, has several advantages, and, in 

 particular, is easy to burn. One decided drawback to it, 

 however, is the fact that it does not by any means gen- 



