ATMOSPHERE IN RELATION TO HUMAN LIFE AND HEALTH. 347 



increased a mile or so southward and perhaps a few miles northward, 

 but would be diminished over the northern half of Sussex, and prob - 

 ably in Surrey, to an appreciable degree. 



Similarly, a wall of 400 feet in height between Yes Tor and Hartland 

 Point, in Devonshire, would increase precipitation along a band par- 

 allel with the wall, but would give a drier climate to the more easterly 

 portions of the county, and probably also to Somersetshire. In Eng- 

 land, not only does the greatest quantity of rain reach us from the 

 southwesterly quarter, but the clouds are lowest in the rains from that 

 quarter, so that the greatest effect of a barrier is produced on rains 

 coming from south and southwest. 



The method of construction is a question for engineers. Would it 

 be possible to construct a screen several hundred feet high, of iron, as 

 used in the large gasometers which we see in the neighborhood of our 

 large towns ? Or is masonry necessary in order to withstand the extreme 

 possible pressure of strong winds? 



The desirability of forming any such artificial barrier would, of 

 course, depend on the calculated probable benefit to be conferred on 

 any county or district, and it would very likely be only in rare cases 

 that the increased geniality of climate would repay the outlay. Pos- 

 sibly it is only worth considering in the case of very wet climates, or 

 of places where little rain falls and more is needed. In England, sup- 

 posing for a moment that its erection is desirable, the line to be taken 

 for a wall must be such that there would be very little disturbance of 

 natural features of interest or beauty; in fact, it should either be across 

 barren moors or wastes, or else parallel to the cliffs on a desolate coast. 

 The line above suggested from Yes Tor, near Okehampton, toward 

 Hartland Point, appears in all respects a favorable one for the pur- 

 pose, as the country to be crossed is dreary and almost uninhabited. 

 The wall would have an additional advantage of permitting trees to be 

 planted on its northeast side in a broad belt, so as to make the begin- 

 ning of a forest, where the winds are now too severe for vegetation. 

 Another favorable stretch of country lies along the ridge of the South 

 Downs between Swanage and Bridport. A high barrier here would 

 give to a large part of Dorsetshire and southeast Wiltshire a climate 

 not unlike that of Bournemouth, which owes its dryness to the hilly 

 promontory of the Isle of Purbeck. 



Portsdown Hill, which runs east and west for nearly 7 miles, and is 

 over 400 feet high, would be another highly favorable ridge for an 

 experimental wall, say 400 feet in height. The practicability of works 

 of this kind can hardly be questioned when we hear of structures like 

 the reservoir embankment at Bombay, a stone barrier 118 feet thick, 

 over 100 feet high, and 2 miles long. A less amount of material would 

 have gone toward a wind wall 30 feet thick at the base, 300 feet high, 

 and 3 or 4 miles long. 



A wall 300 or 400 feet in height and 5 or 6 miles in length, extending 



