1902] DEFECTIVE INSULATION 251 



bunker, and the latter, being in direct communication with the 

 engine-room and thence with the open air, is always consider- 

 ably below freezing-point. As a consequence of this we get 

 very cold draughts in the wardroom, and a thermometer placed 

 on the deck anywhere but near the stove falls to 32° or 34°. 

 A week or two ago it was so bad that I was obliged to sit in 

 my cabin with my feet in a box of hay, an efficient but incon- 

 venient foot warmer. 



' Before the gale in May, when we had no snow about us, 

 the ship was getting very badly iced up inside, but after that 

 gale we were able to improve matters, and now they are a good 

 deal better. At the end of April the temperature in my cabin 

 averaged about 40° during the daytime and 33° during the 

 night, a condition under which one was not tempted to dawdle 

 over the processes of dressing and undressing ; now the tem- 

 perature keeps up to nearly 50°, except near the floor, where it 

 is much colder. The course of improvement was accompanied 

 by much thawing, and for some time we had a general dripping, 

 which was much worse than the ice and infinitely more ruinous 

 to our effects, amongst which mildew is already making rapid 

 strides. In this way, as in others, we have had to buy our 

 experience, and since May we have been fighting the evil by 

 banking up snow without and by nailing up quantities of felt 

 within. 



' The most difficult place to fight is the galley-space, because 

 here it is impossible to avoid the volumes of steam given ofif by 

 the cooking ; directly this steam strikes against the cold sides 

 of the compartment it condenses, and during cooking-hours this 

 space is very much like a shower bath. We have improved 

 matters a little by trying to guide the steam up through the 

 skylight, but the place is still very bad. 



' Our stoves have also been a source of trouble to us, and 

 are likely to continue to be so. They are of the slow-combus- 

 tion type, designed to burn anthracite coal, and though it was 

 claimed that they would be equally efficient with our steaming- 

 coal, we find that to burn it at all we must greatly increase the 

 draught, and consequently we do not achieve the economy of 



