Jan, 1906. | THE SAVANNAHS OF GUYANA. 163 
be the reason that we, in the above-mentioned book of 
Professor Schimper, who undoubtedly collected the most 
reliable data for his book, do not find what we might expect 
for Guyana? It is possible that no other reliable data are 
available than those of Schomburgk, which we saw before 
are vague and unsatisfactory; and surely it is extremely 
difficult to give a correct description of this characteristic, 
tropical association after a short, rather hasty, exploration, 
perhaps only in one direction. How can one have a true 
impression of a vegetation which, in every corner, in all 
directions, keeps both a secret and surprise for us? Nowhere 
else is the vegetation so varied as in the savannahs: more 
than half the number of the plant orders in the Guyanas— 
and there are many—are represented here; every moment 
we may find a new or scarcely known species, convincing us 
that, only after a patient and exact exploration, it will be 
possible to give a satisfactory description of the savannah. 
For it is difficult to say what the typical character of this 
association is. Decidedly xerophilous plants are growing 
alongside pronounced hygrophilous plants, and amongst) 
these, plants of a dubious or tropophilous character. But 
how to explain that? Well, we have to go back to the 
formation of the soil and the climatic influences. As seen 
above, the soil was mainly coarse quartz sand in the upper 
layer, with loose garden soil and clay underneath. This soil 
is manifestly permeable, but the layers are not everywhere of 
the same thickness, and here and there the upper layer 
alone or both the upper and middle layer disappear, leaving 
the under layer at the surface. It is evident that we have 
three distinct conditions before us, which will affect the vege- 
tation. Then again, if we compare the data of the rainfall 
on Placer de Jong, situated in the savannah region, with the 
data of the rainfall in Paramaribo, we find for Placer de 
Jong an average of 21564 mm., and for Paramaribo an 
average of 2226-9 mm.; the highest and lowest in one year 
for Paramaribo respectively 2754:5 and 1240:1 mm.; for 
Placer de Jong 2719°5 and 1372°6 mm. It is notorious also 
that the temperature over the whole country is fairly equal 
and constant, and the relative humidity on the savannah 
a little less than in the forests; but the open savannah soil 
has the full benefit of the immense amount of dew that 
