THE INSECT ASSOCIATION OF A LOCAL ENVIRONMENTAL COMPLEX. 
59 
on the supposition that this species, besides being coprophilous, may also feed on 
humus and decaying vegetable matter. 
Of Curculionidse, species of Sitones were more numerous in the Alluvial Pasture, 
but the only two species of Hypera occurred in Glover’s Meadow (cf samples Nos. 
2, 4). Larvae of Otiorhynchus sulcalus were about equally distributed in both fields, 
but never very numerous. 
Perhaps with such small numbers aS our census represents, no really strict com- 
parison should be drawn between the faunistic composition of the two areas. The 
physical-factor differences of the two were very restricted and confined mostly to 
those of soil, water-content, altitude, and exposure, to which is added the fact of the 
Alluvial Pasture being also grazed. Impartially, the facts as revealed by the census, 
in broad outlines, tend rather to indicate that, in this country at least, the soil-insect 
fauna of grassland in any given locality is not likely to vary to any great extent. Of 
one thing the author is convinced, and that is, large numbers of specimens must be 
treated of in order to throw minor differences due to variable local conditions into 
bold relief. 
In addition to insects alive and dead, one encounters in the soil representatives 
of Araneida and Acarina, Annelida and Mollusca, not to mention fruits and seeds, 
all of which may form food for ground-feeding birds. In many cases they are 
definitely associated with insects, either preying on or being preyed upon by the 
latter. Spiders are quite prevalent, and mites of the genera Gamasus, Trombidium, 
and Smaris are not infrequent, as are likewise various species of worms. Generally, 
in those samples which bore a covering of moss on their surface, two species of 
shelled Mollusca, Cochlicopa lubrica and Vitrea nitidula, were quite common, as 
well as specimens of the slug Avion circumscriptus, and innumerable slug eggs. 
Earthworms of various species and their cocoons, as one would naturally expect, 
were abundant in almost every sample. 
Soon after the work of the soil-insect census had been commenced, the author’s 
attention was directed to a paper by M‘Atee,* in which, from an enumeration of 
all the insects and other invertebrates, besides seeds and fruits, present in four 
square feet of forest floor near Washington, U.S.A., it was calculated that, for the 
particular locality specified, there were in each acre 1,216,880 animals belonging 
variously to Insecta, Arachnida, and other Arthropoda, Annelida, and Gastropoda, 
and 2,107,810 seeds and fruits. Truly formidable figures ! which almost pale into 
insignificance in the light of this same author’s calculation, on a similar basis, of 
the numbers in one acre of meadow land for the same locality, viz. 13,654,710 
animals and 33,822,745 seeds ! 
Of course, these numbers only apply to the class of soil surface indicated and 
to the particular locality stated. As a basis for general conclusions they avail 
nothing. It would, further, be absurd to deny that these figures might not be 
* M'Atee, W. L., “ Census of Four Square Feet,” Science, N.S., 1907, vol. xxvi, pp. 447-449. 
