RAINFALL 19 



when the sea surface is much warmer than the air ; but 

 sea fogs cannot always be accounted for in this way. 



Rainfall. There is no department of marine meteor- 

 ology which has been so neglected as the measurement 

 of rainfall, and we have only a few fragmentary obser- 

 vations to go upon in comparing the rainfall of the land 

 with that of the sea. In this instance, again, the large 

 steamers of to-day offer facilities for accurate obser- 

 vations which were formerly unknown, though fair 

 results have been obtained even on a vessel so small 

 as the " Discovery," of Captain Scott's first Antarctic 

 expedition. The rain-gauge for use at sea must be 

 swung on gimbals, and the lower part heavily loaded 

 so as to keep it as horizontal as a compass bowl. 

 It should be placed in the middle of the upper deck, 

 as far as possible from masts, funnels, or stays, and 

 from the sides of the ship. The two chief difficulties 

 to be overcome in measuring rainfall at sea are ascend- 

 ing eddies of air and the breaking of spray into the 

 gauge. The eddies may be neutralized to some extent 

 by preventing the wind from striking directly on the 

 gauge or the box in which it swings ; this may be done 

 by having a wooden frame fixed to the deck, with walls 

 equal to the height of the gauge, and of such a size as 

 to be at least as far from the gauge as its own height 

 at every point. The amount of sea water which finds 

 its way into the gauge can easily be calculated by 

 determining the salinity of the contents of the gauge 

 by titration, for which all appliances would be found 

 on any vessel fitted out for marine investigations. If 

 it wore found practicable to instal a rain-gauge on an 

 ordinary steamer, it would probably be as much as 

 could be done to record, every time the gauge was 

 emptied, whether the contents tasted quite fresh, or 

 were slightly or very salt. In the first case no cor- 



