36 EEPOET OF THE SECEETAEY. 



attached to tlie publications of our Institution and to the labors of one 

 of its collaborators. Few persons not having experience in the matter 

 could imagine the amount of labor required in the reduction and dis- 

 cussion of physical observations. In the case of the observations re- 

 ported to the Smithsonian Institution nearly three millions of figures 

 have to be gone over. But this is not all. Various hypotheses have to 

 be provisionally assumed, and the deductions from them tested by com- 

 parison with the actual results of observations, while many special re- 

 sults have to be deduced in accordance with previously established for- 

 mulas. The result of the discussion of the rain-fall, besides being given 

 in tables, is illustrated by curves and rain-charts. The printing of the 

 tables is necessarily a slow operation, requiring special care in the cor- 

 rection of the proof. The printing of the charts has been facilitated by 

 exhibiting the relative fall of rain in each part of the country by dif- 

 ferent depths of shading in the original engraving, the distinction being 

 made more obvious by a second printing in a single color. The rain- 

 charts are three in number ; one exhibits the relative fall of rain for the 

 whole year ; another for the three months of summer ; and a third dur- 

 ing the months of winter.' For agricultural purposes' the rain -fall in the 

 summer is most important, but data are given in the tables to ascertain 

 it for every month of the year. 



The distribution of rain is very materially affected by the prevailing 

 direction of the wind, and this again is modified by the declination of 

 the sun, a fact which must be evident when we consider that the motive 

 power of the great currents of the aerial ocean is the greater heat of 

 the equatorial regions derived from the perpendicular rays of the sun, 

 which, expanding the air, causes it to ascend and flow over in each 

 direction toward the poles. The medial line along which this expansion 

 takes place must move north and south with the sun in his varying de- 

 clination. If the earth were covered entirely with water, and were at 

 rest, the currents of the air would be comparatively simple ; but, since 

 the earth is in a rapid rotatory motion eastward, the currents which flow 

 at the surface toward the medial line move on one side from the north- 

 east and on the other from the southeast, thus forming what are called 

 the "trade-winds," while the same currents continued upward and 

 northward on one side, and upward and southward on the other, curv- 

 ing eastward, form the great stream of western return-trades which in 

 the northern hemisphere, in summer, continually flow over the United 

 States, at a high elevation, and which waft the higher clouds eastward, 

 giving a similar direction to the principal storms of every season. In 

 midsummer, when the medial line we have referred to is carried north- 

 ward by the northern declination of the sun, the upper current reaches 

 the earth beyond the fiftieth parallel of latitude, and precipitates the 

 vapor which it brings from the Pacific Ocean in the form of rain on the 

 western coast of America, to the depth, at Sitka, of ninety inches. As 

 the sun declines to the south the rain belt gradually descends along the 



