SEEIES D.— DEDUCTIONS AND REMARKS. 169 



direction, recross that line for tlie most part (thirteen curves out of twenty) in 

 July, and continue on the north side till the end of the year. Four curves cross 

 a little earlier, in June, and of the remaining three, the two at the extreme north 

 (Newfoundland and Canada) cross in August, and the one at the extreme south, 

 near the southern limit of the system, in May. The time seems to vary somewhat 

 with the latitude and the trending of the adjacent coast. 



Of the four curves on the limit between the equatorial and westerly systems 

 (Plate VIII. Nos. 44, 45, 71, and 80), two do not cross the line of mean direction 

 at all, but lie to the right of it for the vfhole year; and the other two cross it in 

 August, lying to the right before and to the left afterwards. 



Of the easterly winds of the equatorial system, those north of about latitude 

 24° (Plate VIII. Nos. 46, 47, 48, 50, 51, 52, and 82), cross the line of mean direc- 

 tion in April, May, or June, and those farther south (Plate VII. Nos. 84, 85, and 

 87) in October. The latter are at sea, and may possibly be affected by the prox- 

 imity of the coast of South America. 



4. On the eastern side of the Atlantic there is less uniformity in the time of 

 crossing, though (not including the exceptions already named), it is on an average 

 considerably earlier. Out of eighteen curves (Plates VII. and X.), one crosses in 

 February, one in March, four in April, five in May, two in June, one in July, one 

 in September, and three do not cross it at all, but lie to the south for the whole 

 year. One of these three is at St. Petersburg, in Russia, another at Elgin, in the 

 north of Scotland, and the other at sea (No. 77), on the limit between the equato- 

 rial and westerly systems, thus agreeing with its neighbors (Nos. 71 and 80) on the 

 western side of the Atlantic. The curve for the stations in Austria (No. 92) might 

 very properly be added to this list, as it lies south of the line of mean direction 

 over eleven months in the year. 



5. The curvature in India and China is the same as in the westerly system on 

 the west side of the Atlantic, while that in Western Siberia corresj)onds to those of 

 the European stations, so far as it can be said to have any character at all. 



6. The stations east of the Mediterranean Sea are as devoid of law or agreement, 

 in the form of the curves described by their winds, as they were shown to be in 

 regard to the mean direction of their winds. 



Tlieoretica I Considera tions. 



The causes of the peculiarities, in the inflection of the curves we have been con- 

 sidering, are more clearly seen by analyzing them in the manner described in the 

 introduction to the foregoing series (D). By thus detaching the deflecting forces 

 from those which determine the mean annual direction of the wind, the law at once 

 becomes apparent that on the sea-coast, and even for some hundreds of miles from 

 it, both on sea and land, the deflecting forces are directed towards the land, in the 

 warmer parts of the year, and towards the sea, in the colder; a most convincing 

 proof (if any more were needed) of the influence of heat in the production of winds, 

 and that, too, upon an extensive scale. 



