G-. S. Graham-Smith 
471 
It is stated that house-flies seldom occur in the upper rooms of 
houses and are apt to avoid rooms in which no food is present. During 
October, however, house-flies were found almost every evening settled 
on the walls of a bedroom above the room just mentioned. 106 were 
caught, of which 35 were males and 71 (67 %) females. 
House-flies in the kitchen. 
House-flies were first noticed in the kitchen on June 7, and though 
a few specimens were seen throughout that month, they only com¬ 
menced to enter the room in considerable numbers on July 1. The 
number present daily is recorded on Chart 3. A gradual increase 
in the numbers occurred during July, though at this time many were 
probably passing in and out daily. 
It mil be noticed that in the middle of the season the chart bears 
a strong resemblance to Chart 1, recording the daily fluctuation 
in the numbers of blow-flies confined in an experimental cage. Each 
great rise in the curve probably corresponds, as in Chart 1, to 
the emergence of a new generation, members of' which entered shortly 
after becoming fully developed. 
The most noticeable of these rises occurred on July 30 and 31, 
both days with long periods of bright and hot sunshine, August 10, 
a very hot day, with bright sunshine at intervals, August 26, and 
September 5, 12 and 13 with at least 9 hours of continuous sunshine 
after 8 a.m., September 24, a windy day with little sunshine, and 
October 8, with 8 hours of continuous sunshine. These observations 
do not support the view that the flies enter in order to shelter from 
adverse weather conditions. The flies which entered on the days just 
mentioned usually appeared to leave the room in small numbers daily, 
ten days or more elapsing before the incursion of a fresh swarm. 
On seven days, varying greatly in regard to weather conditions, 
considerable numbers of flies emigrated out of the room; August 3, 
a very wet day, August 13 and 31, very bright, warm, sunny days, 
September 14 a dull day with a strong west wind, September 15 a hot 
and sunny day, September 30 a cold and sunless day and October 18th 
a mild and sunless day. 
During the course of these observations the impression was created 
that the members of each swarm remained indoors for varying periods, 
and that in the periods between the swarms little immigration took 
place. 
