81 



decrease of 3* per 1,000 feet altitude which agree best with South-east African 

 observations we may get for each point, in the absence of more direct data, an ap- 

 proximation to the temperature, which cannot be very inaccurate in a small stretch of 

 country like Natal. In this manner, the isothermal lines shown on Map IV., have been 

 deduced from the configuration of the ground shown on Map II. 



Map II. shows the average height of the land, in Natal. It was constructed by 

 plotting all the reliable observations of altitude that could be got. The heights of many 

 points were given accurately by the railway surveys, and by Major Morris' trigono- 

 metrical survey ; the heights of ther points were found by correcting the barometrical 

 observations available, including those of Major Grantham and Dr. Sutherland, in 

 proportion to the error shown by observations of the same series, at points of which the 

 height has since been fixed by survey. From a knowledge of the topography, the average 

 heights of the different portions were then deduced. They are shown on the map by 

 contour lines giving the mean elevation every thousand feet. 



The Durban observations of temperature for 1884-1888, give a mean value of 69'7 P 

 Fahr. reduced to 70*4* at sea level. The Maritzburg observations taken from 1858 to 

 1866 by Dr. Mann, give a mean of 64*6 reduced to 70'8 <? at sea level at Durban. A 

 mean of 70'6 has been taken to represent, for Natal, the mean temperature at sea level 

 in the latitude of Durban.* 



The distribution of temperature, derived from this value, is sketched in Map IV* 

 The isothermal lines here represent actual, not reduced temperatures as in Map III. above. 

 But for places of known height, it would always be preferable to reckon the mean 

 temperature independently, with the aid of the data given. Comparing Maps I. and IV., 

 it may be seen that the distribution of forests nearly follows that of temperature, the coast 

 region being limited by the isotherm for 67 and the Midland region by that for 59. 



The isothermal lines for 70, shown on Map III., cross the countries which have a 

 mean annual temperature equal to that of Natal, and where corresponding climates will 

 be found at corresponding elevations ; it is to these countries that we must chiefly look 

 to, to find trees suitable for Natal. In the northern hemisphere, the curve for 70" cuts 

 both coasts of North America near the 30th parallel, bending inland towards the central 

 plateau ; then crosses the Atlantic, follows the coast regions of Morocco, Algeria, and 

 Tunis j thence across the Mediterranean to Syria ; inflecting northward, it crosses Central 

 Asia along the 37th parallel, re-descending to 29 Lat. to leave the coast of China and 

 pass south of Japan. In the southern hemisphere, the corresponding curve, leaving 

 Natal, touches the southern point of Madagascar, and cuts the centre of the Australian 

 continent at 24 Lat. ; in South America, the West coast is crossed at 14 Lat., and the 

 curve then descends along the Andes, to leave the East coast near 30 Lat. 



* It is to be regretted that longer series of meteorological observations should not exist in Natal. The 

 latter Maritzburg (Botanic Gardens) observations have been taken without standard instruments or methods, 

 so that they are useless for purposes of comparison ; the exposure of the thermometers is defective, and would 

 probably affect the readings to a considerable extent. 



