302 VISITS TO MADAGASCAR. ciup. xi. 



altogether one of the most remarkable that has been discovered 

 in Madagascar. And the extent to which it prevails may be 

 inferred from the native name, ravinala, by which it was de- 

 signated by Sonnerat, its discoverer. Eavinala is literally 

 leaf of the forest, as if it was the leaf by which the forest was 

 characterised, which is the fact where it abounds, though in 

 many parts it is not met with at all. The tree rises from the 

 ground with a thick succulent stem like that of the plantain, 

 or the larger species of strelitzia, to both of which it bears a 

 strong resemblance. It sends out from the centre of the stem 

 long broad leaves like those of the plantain, only less fragile, 

 and rising, not round the stalk, but in two lines on opposite 

 sides, so that as the leaves increase, and the lower ones droop 

 at the end, or extend horizontally, the tree presents the 

 appearance of a large open fan. When the stem rises ten or 

 twelve feet high, the lower part of the outer covering becomes 

 hard and dry, like the bark of the cocoa-nut tree. Many of 

 the trees in this region were at least thirty feet from the 

 ground to the lowest leaves. I frequently counted from 

 twenty to twenty-four leaves on a single tree, the stalk of 

 each leaf being six or eight feet long, and the broad leaf itself 

 four or six feet more. 



The whole of these twenty-four bright green gigantic leaves, 

 spread out like a fan at the top of a trunk thirty feet high, 

 presented a spectacle as impressive as it was to me rare and 

 beautiful ; and in this part of the country they were the most 

 conspicuous objects for miles together, and were it not that 

 these vast bright green shining leaves are slit on each side by 

 the winds, and so flutter in smaller portions with the passing 

 breeze, the prevalence of this tree would impart a degree of 

 almost inconceivable magnificence to the vegetation of the 

 country. 



In the fan-like head of the traveller's tree, there were 

 generally three or four branches of seed pods. The parts of 



