OP ARTS AND SCIENCES. 337 



instrumental observation that a fully developed thunder-storm needs, 

 may be inferred from the following description of the phenomena 

 accompanying such a storm, while passing through the storm-belt at 

 right angles to its length. The earliest forerunner of the storm is the 

 high cirro-stratus overflow, reaching far ahead of the rain : as this 

 rises above the western horizon, the outlines of massive gray " thunder- 

 heads," or cumuli, appear beneath it; they reach to a great height, 

 and are seen towering aloft as they approach. Towards the base they 

 are very dark ; but often at the lower front margin of these heavy 

 clouds there appears a ragged fringe of gray squall-cloud, much 

 agitated, and below this is the falling rain. All these clouds should 

 be examined to detect their motion and growth. As the storm comes 

 nearer, and the thunder gi'ows louder, the wind squall may sometimes 

 be recognized in the distance by the dust that it raises ; it strikes 

 suddenly, often with destructive force, and quickly passes on as the 

 rain begins. This squall is not an invariable accompaniment of thunder- 

 storms, and it is yet to be determined why it is sometimes present, 

 sometimes absent. The temperature falls rapidly as the rain comes ; 

 for the temperature of the rain, and especially of the hail that often 

 accompanies it, is much lower than that of the air. Much is hidden 

 now by the falling rain, but, as it passes, the lower clouds appear and 

 rapidly break away, showing brighter clouds or clear sky and sunshine 

 in the west. The high cirro-stratus is then seen on the rear of the 

 storm : it often shows the curious festoon-clouds on its lower surface. 

 As the sky clears, the temperature returns towards its normal value. 



The altitude, dimensions, and motion of all these many parts manifestly 

 offer plenty of material for systematic study ; and all weather services 

 agree that many observers are needed in such work. The average 

 distance between the stations must be a small fraction of the dimen- 

 sions of the storm, or else features of importance may escape notice. 



Plan of Observations. — In laying out the task for the volunteer 

 assistants, no predetermined hours for occasional synchronous observa- 

 tions can be found that serve well in this work; times set before hand 

 would too often fail to catch the swiftly passing storms. Records must 

 be made chiefly at times announced by tlie storms themselves. It 

 would take the greater part of an observer's attention for the two or 

 three hours of a storm's passage to detect and record all its noteworthy 

 peculiarities, and this is much more time than most volunteer observers 

 can give. But the interest, time, opportunity, and ability of observers 

 are so varied, that what is easily accomplished by one is far out of 

 reach of another. "We have therefore given detailed instructions to the 

 VOL. XXI. (n. s. xin.) 22 



