iS HARBOUR CONSTRUCTION. 



move. If inclined to rise, the surface is convex; if to fall, 

 concave. This arises from the friction or adhesion of the mercury 

 to the sides of the tube. As the tendency of the mercury is 

 usually a more sensitive indication than the actual level at 

 which it stands, this point is worthy of notice. 



The significance of different kinds of clouds depends much 

 upon their surroundings, so the value of conclusions arrived at 

 from their study will naturally be in proportion to the experience 

 and local knowledge of the observer. 



Different forms of cloud have, however, from long observation, 

 been popularly associated with certain kinds of weather. For 

 instance, cirrus cloud, of the kind known as " mares' tails," " sea- 

 grass," etc., is a pretty certain indication of wind, especially if its 

 movement be rapid. 



This cloud always occupies a high position, so its drift is 

 indicative of the direction and velocity of the upper wind currents, 

 and deserves careful attention. 



Hard cumulus to seaward, showing over a dark haze, often 

 presages a gale. This must, however, be distinguished from the 

 rocky detached summer cumulus, which in many parts is 

 associated with fine weather; as are also soft-looking, fleecy 

 clouds, or cirro-cumulus. 



Abercromby, in speaking of cloud prognostics, says that what 

 is known as " festooned cumulus " is usually followed by a 

 severe gale in Orkney, whereas similar clouds in Lancashire are 

 only considered a sign of rain. 



Small hard stratus clouds, such as are frequently seen at sunset, 

 are indicative of fine weather, while soft-looking, heavy stratus 

 usually degenerates into nimbus or rain-clouds. Stratus clouds, 

 however, from the conditions necessary for their formation and 

 existence, seldom indicate wind. 



It will be seen that, as clouds have not the same significance in 

 all localities, or under all conditions, it is not possible to lay down 

 any laws regarding them which shall be universally applicable. 

 The inference to be drawn from various forms and movements of 

 clouds, as well as from other symptoms in any locality, may, 

 however, soon be learnt by a little careful observation, and by 

 noting the sequence of weather associated with them. Such a 

 knowledge, coupled with the readings of the barometer and 

 thermometer, will be valuable aids to forecasting to those who 

 may happen to be in remote places, where they have not the 



