160 SECTIONAL ADDRESSES. 
This, then, is still another important feature of the new anthropology, 
the insistence that research and theory must not be separated but most 
be as closely united as they are in other sciences. The observations of 
the data, the formulation of hypotheses and the testing of these hypotheses 
by further direct observation are all parts of one single process which 
should be carried out as far as possible by the same individual. 
Meanwhile there is one fact that seems to me at times to make the 
position of our science almost tragic. Now that by the gradual develop- 
ment of theory and the improvement of methods of investigation we are 
in a position to make the most important contributions to the science of 
man by the intensive and exact study of the less developed cultures of 
the world, those cultures are being destroyed with appalling rapidity. 
This process of destruction, through the combined action of European 
trade or economic exploitation, government by European officials, and 
missionary activity, is taking place with accelerated pace. During the 
twenty-five years since I first took up this work myself I have seen great 
changes. Tribes in Australia and Melanesia and in North America from 
which we could have obtained most valuable information a quarter of a 
century ago will now afford us little, or in many instances nothing. In 
another quarter of a century the position will be ever so much worse. 
Work that is still possible in all parts of the globe will then be forever 
impossible. Is there any other science, or has there ever been another 
science, faced with such a situation, that, just at the time it is reaching 
maturity, but while through lack of general interest and support it has 
few workers and very scanty funds, a great mass of most important 
material is vanishing year by year without the possibility of making any 
study of more than a minute fraction ? 
It will be through field researches that anthropology makes progress 
towards becoming a real and important science. But intensive studies of 
single cultures or societies are not sufficient in themselves. Such intensive 
studies must themselves be inspired and guided by theory, and theoretical 
sociology must rest on the comparison of different cultures one with 
another, for comparison in this science has very largely to take the place 
of experiment in other sciences. 
The newer anthropology is developing a different conception of the 
comparative method from one that has been current in the past. In the 
older anthropology we weré offered books or monographs in which similar, 
often only superficially similar, customs or beliefs were collected from all 
sorts of cultures all over the world and thrown together. It was this that 
was in fact often thought of as constituting the comparative method. 
Such a procedure may be useful in giving a first survey of some particular 
problem or group of problems, and has been useful in that way in the 
past. But it can never do more than indicate problems, it cannot 
solve them. For that, a more precise and more laborious procedure is 
necessary. 
To understand what precisely the comparative method should be we 
must bear in mind the kind of problems to the solution of which it is 
directed. These are of two kinds, which we can distinguish as synchronic 
