38 
SPICES 
CHAP. 
advisable to plant vanilla in hollowed ground or trenches 
6 in. deep, so as to keep more easily the roots out of reach of 
the burning sun, or the soil must be porous enough to prevent 
it being water-logged. This is not done in Seychelles, where 
vanilla is grown on the surface of the ground and weeds or 
twigs accumulated in heaps to protect the roots. This kind of 
protection is insufficient, because as a rule planters protect the 
roots only once a year, when the time comes for pruning the 
vines and clipping the props (March to June). I wish to 
emphasise the sad consequences of this exposure of the roots, 
because it accounts to some extent for the irregularity of 
growth and flowering in any one plantation. These physical 
conditions of growth are of great importance for delicate plants 
like orchids, and even when the soil is not porous I think that 
an attempt should be made to plant in trenches filled first with 
a good layer of broken stones and then with 6 in. at least 
of fibrous soil. It was thought that improvement in the 
growth of the vine also could be obtained by manuring them. 
Up to now the results obtained are very promising, although 
the manuring was started after the rains in ISTovember last, or 
a few months ago. 
After several investigations I came to the conclusion that 
a careful examination of the roots while the plant was under 
treatment with manure would establish which kind of manure 
would be welcomed by the vine and which not. Eegarding 
phosphates I have employed phosphatic guano from Flat Island, 
which contains 72 per cent of tribasic phosphate and only 0*40 
nitrogen (7 kilos per basin). These rocks were broken in 
small pieces like macadam and placed at mid-distance between 
two rows of vines placed 4 ft. apart. The coral was used 
in just the same way (5 kilos per basin). Kainit was also 
employed in the solid state (1 kilo per basin), but as this salt 
is hygroscopic, small quantities were arranged in a layer of 
coco-nut husks placed also at mid-distance between the vines. 
The kainit would thus dissolve more slowly and be absorbed by 
the coco-nut husks. After three months it was found to be still 
incompletely dissolved under these conditions. 
Mtrogen is employed in the soluble state as nitrate of soda. 
The vines are watered once a week with weak solutions of 8 to 
16 grams in 100 litres of water (50 litres per basin). As far 
as these experiments go, it is interesting to record the root 
development which immediately followed the applications of 
these measures. As already stated, the roots were, previous to 
the application, all found within 18 in. of the walls of the 
basins, but three weeks after the application numerous roots 
