221 
remarkable Effects of a Shipwreck. 
This experiment clearly enough confirms the greater dan- 
ger of being wet with fresh than salt water ; but in itself 
points out nothing certain besides, except that it is not to be 
rashly repeated. I mean to try some of these experiments to 
a greater extent on the brute creation, when I have procured 
thermometers better suited to my views. The thermometers I 
employed had not a sufficient mobility for very nice experi- 
ments, and I am well aware that in particular instances this 
may have misled me, though the general results, which is all 
that is of importance in such experiments as these, will, I hope, 
be found just and true. 
Before I conclude, I must offer a few observations on the 
subject that led to these experiments. 
1. It is, I think, already well known among seamen, that 
where there is only the choice of being wet with salt or fresh 
water, it is always safest to prefer the first. In the heavy 
showers of rain, hail, or snow, by which gales of wind are ge- 
nerally accompanied, the men that must be exposed to them, 
ought, like Lieutenant Bligh and his crew, to wring their 
clothes out of salt-water. 
2. In all cases where men are reduced to such distress by 
shipwreck or otherwise, that they can only chuse between the 
alternative of keeping the limbs constantly immerged in the 
sea, or of exposing them to the air while it rains or snows, or 
the sea is at times washing over them, it is safest to prefer 
a constant immersion ; because, in the northern regions where 
the cold becomes dangerous to life, the sea is almost always 
warmer than the air, as the experiments of Sir Charles Doug- 
las shew ; and because there is not only a danger from the in- 
creased cold produced by evaporation, but also from the loss 
mdccxcii. G g 
