DISTEIBUTION : FUETHEK EESULTS 81 
These results, however, I must beg you to keep to your- 
self, as we are not permitted to communicate Botanical 
Information (does it deserve the name ?) except through 
the Lords Commissioners ! 
He perceives also that the distribution and abundance of 
vegetation in this region depends not on the height of the 
mean temperature, but on the amount of moisture in the air 
and the equable level of heat and cold, free from extremes. 
To establish this accurately would prevent critics from 
repeating that ' nothing of importance had been done towards 
investigating the causes of difference in Geographical distribu- 
tion since the publication of Humboldt's work.' 
To his Father 
March 7, 1843. 
I long to see your new work on Ferns ; perhaps you will 
do something to their Geographical distribution, which seems 
most dependent on a uniform and moist temperature such 
as Islands enjoy. All the Magellan species that inhabit 
the Falklands, there become harsh and coriaceous, from the 
vicissitudes of temperature, and of the hygrometric state 
of the air to which they are exposed. . . . The Hygro- 
meter I consider of more importance than the Barometer 
in all ordinary cases, that is, where the Islands are not 
large and the mountains not high. ... I have lately 
been examining some of my hygrometer observations and 
find that the difference between the vegetations of the 
Falklands and the Fuegia may be well accounted for. When 
the results are placed in a tabular form it is quite surprising 
to see to what vicissitudes of temperature and moisture the 
Falkland plants are exposed. Now the mean temperature 
of the Falklands is the highest, but its plants are exposed 
to dry winds, great heat of the sun's rays unimpeded by 
any vapour when it is calm, and great cold at night, whilst 
those of Fuegia are not so, and enjoy perpetual moisture, 
and are very sensitive to extremes of temperature, as also 
to dryness. 
His original intention had been to write a Flora Antarctica, 
where his work would be on a fairly little exploited field. As 
