Vol. 8, 1922 
BIOLOGY: PEARL AND PARKER 
217 
Bilski {loc. cit.) uses essentially the same equation to describe the influence 
of density of population and growth in tadpoles. Certainly the subject 
deserves further study. Possibly our present results offer a clue to a 
part at least of the solution of the problem of the practically world-wide 
decline in the birth-rate which had been going on for more than a quarter 
of a century before the outbreak of the war. 
Discussion. — How can the result be explained? The extreme closeness 
of the observations to the theoretical curve indicates clearly that we are 
dealing here with no haphazard performance, but with a highly lawful phe- 
nomenon. The first point that occurs to one is that since we are dealing 
with imagoes only, the result may perhaps be explained by larval and pupal 
elimination due to overcrowding in these stages. In the bottle say with 20 
females there were laid, on this view, something like 20 times as many eggs 
as in the bottle containing one female, these eggs developed into 20 times 
as many larvae, there was not room or food for so many larvae, therefore 
many of them were selectively eliminated, and the survivors which got 
through to the imago stage indicated only a low reproduction rate, which 
represented however only the capacity of 80 cc. of banana-agar or 23.76 
cm. 2 of food surface to nourish larvae, and nothing else. 
Plausible as this hypothesis is — the more so because it would at once 
indicate why our rate of reproduction variable follows Farr's Law, since 
it would mean Farr's Law applying to the larvae — it cannot possibly 
account for the facts. To demonstrate this we have only to examine with 
^ome care the absolute figures given in the fifth column of Table 1. In 
the first place let it be recalled that the greatest drop in rate of reproduction 
(imagoes per 9 per day) occurs as we pass from a mean density of 2 flies 
per bottle to one of 3.4 flies per bottle (1 and 2 pairs originally). The 
next greatest drop in the rate is from the bottles of density 3.40 to those 
of density 5.50. Now at all these low densities one in fact cannot possibly 
speak of larval crowding at all. This anyone knows who has experiment- 
ally bred Drosophila. A half-pint bottle with 80 cc. of banana-agar will 
support many more than all the larvae one or two or three females will 
produce in 8 days. This is indicated by the absolute figures in the fifth 
column of Table 1. The bottles containing 9 mated pairs at the start 
produced 2117 offspring imagoes. Therefore the food was sufficient to 
allow at least that many larvae to come through successfully. But in 
the bottles with 1, 2, or 3 mated females the absolute number of larvae 
was much less than this. Now the drop from the one pair to the two 
pair bottles in rate of reproduction was 8.8. To prove that this cannot 
be explained by larval crowding and elimination we have only to multiply 
21.4 the rate per 9 day in the 1 pair bottles, by 89 the number of 9 days 
in the 2-pair bottles. This will give the number of imagoes which would 
have been produced in the 2-pair bottles had the reproduction rate of the 
