REPLACEMENT OF THE ANCIENT E. AFRICAN FOREST 495 



from the Cape Province to Elgon and beyond, in the higher rainfall 

 areas; but in parts of the Cong^o basin and elsewhere in the west 

 it covers the face of the country. A suggestive gradation occurs 

 from this condition, through the forests of British East Africa, 

 which, if isolated, are still large, down to the diminutive dense 

 forests of Southern Rhodesia. Where continuous they are re- 

 garded, I believe, as primary. Are they then to be looked on 

 as secondary in the other areas? If so, how. with such 

 laboriously-moving constituent members as, for instance. 1Vid~ 

 dringtonia, have they reached their jMXsent widely isolated posi- 

 tions? If not, whv are they no longer continuous througlnmt? 

 It is the last problem, chiefly, which I ])ropose to discuss in this 

 paper. 



3. Present Distridu'tion of High Forest in the 

 Melsetter-Beira Section. 



Chirinda is our largest |)atch on the British side of the bor- 

 der, but, as I have said, it is by no means the only high forest in 

 the country. Small i)atches are scattered along the Portuguese 

 border and in Portuguese territory south of the Lusitu River, 

 and wooded kloofs often tend to reproduce this type of forest. 

 Further north, the higher mountains, including Chimanimani, 

 boast many forest patches of the " mountain " type, and forest 

 patches crop out along the high Anglo-Portuguese boundary so 

 far north as Umtali and, I am told, beyond. I am not ac(|uainted 

 with the last-named patches, but Mr. L. Cripps told me, when in 

 Chirinda some years ago, that he recognised there species of big 

 trees which also occur in forest on his farm near Umtali. I am 

 not aware that forest of this kind occurs much more than 25 miles 

 west of the border, and, if that be the case, it may be regarded as 

 falling entirely within the area of greater rainfall ; but, travelling 

 30 or 40 miles east, one comes on an extensive forest patch, 

 larger than Chirinda, on the eastern slope of the Sitatonga Hills,* 

 and, far nearer sea-level and relatively nearer the coast, the well- 

 known patches on the Beira Railway and one south of the Buzi 

 Ri\'er which contains, as well as Khaya iiyasica, other trees that 

 in growth, foliage and bark I was unable to- distinguish from 

 certain trees of Chirinda, though I was unfortunately unable to 

 secure flowers. It is true that further dense forests — the great 

 dwarf rubber forests of the lower veld — cover immense areas, 

 and mav well be taken into considerati(^n in connection with the 

 di,stribution of our forest animals (they are the only i>lace in 

 which I have seen Papilio opliidoccphahis, also occurring in Chir- 

 inda, really abundant, Intt they seem to me to belong to a very 

 different type from the kind of forest I am discussing here, and 

 ])robably to be, in part, secondary. At the same time, I am 

 informed that in portions of these forests not visited by myself 



* I include the Sitatonga Forest with a slight reservation. I passed 

 through in May. 1900. when 1 had not yet taken up botany. ;md wlien. 

 also, I had not yet seen Chirinda, so tliat my memory of it may he 

 inaccurate. I believe it to be correctly included here. 



