CHAP. XIII TROPICAL GLACIATIONS 243 



It is true that, in the Himalaya, ice action once operated 

 several thousand feet below its present limit, but there is no 

 proof that it reached the plains ; ^ moreover, this occurred in the 

 midst of a continent and 30 degrees from the Equator. 



Some evidence has indeed been adduced from the Cape of 

 a recent glaciation, but so far it is inconclusive, and the facts 

 opposed to it are very weighty. For example, one of the most 

 easily recognisable rocks in South Africa is the volcanic " Pipe 

 Amygdaloid " of the Stormberg series. If glacial action had 

 occurred since this was formed, boulders of it must have been 

 distributed across the country. Their absence seems incom- 

 patible with the existence of an ice sheet in Cape Colony since 

 the Cretaceous period. It is therefore less surprising that by 

 the Europeans who have visited Kilima Njaro (more than a 

 hundred in all) no extension of its glaciers has been recorded. 

 It may be objected that negative evidence in the case of 

 Equatorial Africa is not reliable ; but that of the Andes shows 

 conclusively that there has not been any universal extension of 

 the glaciers in the tropics. The fact that the glaciers at the 

 Andes are now at their maximum extension is well established : 

 D'Orbigny indeed argued from it that the Andes are still 

 undergoing elevation. Mr. Whymper has informed me that 

 only twice in Ecuador did he see any trace of glaciation below 

 the level of the existing glaciers. In the more important case 

 he found some decayed rocJies ino2it07inces below his second 

 camp on Chimborazo ; but they were so little below the 

 " Glacier de Debris," that a mere local variation in wind would 

 account for the slight extension of the ice that made them. 



Thus the negative evidence of Equatorial and Southern 

 Africa and the analogy of the Andes seemed to forbid any 

 appeal to glacial theories for an explanation of Central African 

 anomalies in biological distribution. When, therefore, above the 

 forests of Kenya, I found an old moraine several thousand feet 

 below the level of the existing glaciers, I felt that I had found 

 not only a valuable clue to the causes of the spread of the 

 Alpine flora, but a chance of determining the date of its 

 extinction on the lowlands and of the change in the African 

 climate. This evidence has been considered at some length in 



' C. A. M'Mahon, Records Geol. Surv. India, vol. xiv. (1881), p. 310 ; vol. xv. 

 (1882), p. 49. See 2\s,o"Man. Geol. India, ed. 2, p. 484. 



