The Problems of Ecology. 197 
descriptive and the experimental, need not be and should not be 
completely separated in practice, though they may be logically 
distinct. So far as possible the descriptive study of plant- 
associations should go hand-in-hand with enquiries into the causation 
of the phenomena involved, but it must be admitted that this ideal 
union is not always a practicable one. In the case of the botanical 
exploration of a new country, it is clearly out of the question for 
the traveller to do much more than collect and record the asso¬ 
ciations he meets with. He is fortunate if he is able to make a few 
hasty observations on temperature, atmospheric humidity and the 
like, while it is obvious that no systematic experimental work can 
be undertaken. Yet his collections and records, if intelligently 
made, are of the greatest u,se in filling up gaps in our knowledge of 
world-vegetation. And the same is true I think even of our own 
country. Here, of course, the resident recorder of the broad 
features of the vegetation has great advantages over the travelling 
recorder. He has time to prepare vegetation-maps of any degree 
of accuracy and detail that seems called for, and he can visit his 
areas over and over again, noting not only seasonal, but progressive 
secular changes in the flora. He has leisure to make systematic 
observations on soil, temperature and rainfall, and to determine 
their relations with the distribution of vegetation. Botanical 
survey work of this kind, such as has been extensively carried 
out on the practically untouched mountain and moorland plant- 
associations of Scotland and the north of England during the last 
few years, is of great value, quite apart from that experimental 
investigation which everyone will admit is required to elucidate 
completely the ecological relations of plant life. Broad problems 
arise which can be solved only by extensive comparative work in 
the field, and until we go into the field and work out the distri¬ 
bution of vegetation systematically, we may not even suspect their 
existence. The experimental investigator confining himself to a 
limited flora will necessarily miss altogether unsuspected combi¬ 
nations of plants which the surveyor and cartographer of large 
areas meets with. 
The importance of map-making as a graphic record of the 
distribution of vegetation is very great. It as necessary to the 
formation of clear mental pictures of the facts of distribution as 
geological maps are in the distribution of rocks and drift-deposits ; 
and such clear mental pictures are essential to the appreciation of 
the problems involved. The making of adequate botanical maps is 
not an easy or a straightforward task, as anyone who has seriously 
tried it knows very well. The science of botanical cartography is 
