60 
DR ALFRED E. CAMERON. 
easily reversed. If an area of lour square feet in a forest included a decaying 
tree with its myriads of Scolytidse and other wood-borers, the figures which would 
be obtained would probably surpass any that could be derived from a census of 
any meadow area of the same size. 
In conclusion, it is my privilege to be able to express here my gratitude to the 
following gentlemen who courteously assisted in the identification of much of the 
material, more particularly Diptera, Coleoptera, and Hymenoptera : — Mr F. E. 
Edwards, Natural History Museum, London ; Mr J. E. Collin, Newmarket ; Mr 
C. Morley, Natural History Museum, London ; and Mr J. Ray Hardy, Manchester 
Museum. For the rest, the author found the entomological collections under the 
charge of the last-named of invaluable assistance, and they were constantly 
consulted. 
Summary. 
1. In any given locality the composition of the insect association is determined 
by a complex of factors, which may be classified as follows, viz. physical, physio- 
graphical, topographical, and vegetational. Primarily, it is dependent upon the 
ecological type of the vegetation. 
2. The most accurate index of the various physical factors surrounding animals 
is the “ evaporating power of air.” 
3. The insects of a given association may not be strictly confined to it. It is 
therefore necessary to distinguish between those species which are peculiar or 
proper to an association, and those which owing to a response to varying stimuli, 
play the role of invaders. 
4. In all cases the criterion of an insect’s true surroundings or habitat is the 
place where it breeds. It must be also recognised that as the facies of a habitat 
changes so also will its fauna (phenomenon of succession of organisms). 
5. A census of soil-inhabiting insects is of interest in relation to the problem of 
the food-habits of ground-feeding birds. 
6. A comparison of the soil-insect fauna of two areas which differ radically as 
regards their soil-types and vegetational covering is of intrinsic importance in 
obtaining a solution to the question in what measure the faunistic differences can 
be explained by variations of edaphic factors. 
7. In a restricted way the study of the two grasslands, Glover’s and Alluvial, 
was productive of much information on differences of their soil-insect fauna. In 
this connection the food-habits of the various species rank as of primary con- 
sideration. 
Correction. 
It is necessary here to rectify an error which occurs in the explanation of the 
plates on p. 203 of the paper, “ The Insect Fauna of the Soil,” Jour. Econ. Biol., 
vol. viii, pt. 3. Larva of Rhyphus fenestralis should read larva of Tipulid. 
