less because of loss of weight. From the mean 
Maturity Index (see pp. 220 to 221) the expected 
mean weight may be calculated. For the male: 
male encounters the observed mean weight for win- 
ners was only 15 grams less than the expected, 
whereas for the losers it was 48 grams less than the 
expected. Similarly for female: male encounters 
the respective reduction in weight of the observed 
from the expected was 24 and 92 grams. In female : 
female encounters the respective losses were only 10 
and 12 grams. Thus, it may be seen that males 
who lose in conflict situations have lost propor- 
tionately more weight than their winning opponents 
of either sex. I am convinced that this greater loss 
in weight is not a result of an inability to obtain 
sufficient food but in some way depends upon the 
greater stress experienced by losers. 
In order to obtain more insight into the relation- 
ships of age and weight to success in conflict the 
data forming table 41 was separated into subgroups 
and plotted in figures 123 to 125. 
Males under 150 grams, which engaged in con- 
flict situations with older, larger rats, were those 
whose mean weight far exceeded the mean weight 
for that age. This is important despite the fact 
that they always lost such encounters, because it 
indicates that an early expression of aggression is 
more likely, provided the individual has a rapid 
rate of growth. The only other category of losers 
whose mean weight exceeded that for the mean of 
the population at that age was for sibs. Most of 
the sibs involved in observed paired conflicts were 
born during 1947. This fact certainly contributes 
to the increased weight of sib losers. However, it 
is also probable that slowly growing sibs infre- 
quently fight among themselves. Females 28, 33, 
37, 42, and 43 of litters 5 and 6 were involved in 
the five female: female sib conflicts. Their Mean 
Maturity Index was 1.64. These same 5 females 
were involved in 12 conflicts with the female sibs 
17, 20, and 25, over whom they always won. The 
Mean Maturity Index for these latter three females 
was 11.47, and they were never observed in conflict 
among themselves, although they often came in 
contact. 
Regardless of the age differential between loser 
and winner, the following general conclusion may 
be made concerning the conflict categories 3 and 4 
shown in these figures: For their respective ages, 
winners have greater than mean weight and losers 
less than mean weight. 
c. Season. I soon came to the conclusion that 
fighting was less intense during the winter months. 
This was a time when rats formed exceptionally 
large harborage aggregations and when sexual 
activity was nearly inhibited. Males between 8 
and 12 months of age were selected as the most 
likely group to reflect a seasonal variation. The 
results of this analysis are shown in figure 126. It 
is quite apparent that during the breeding season 
the intensity of antagonism is something of the 
order of twice that during the winter months. 
The fact that rats assemble in larger harborage 
aggregations during the winter months does not 
prevent them expressing antagonisms following 
departure from their place of rest. Several in- 
stances such as the following were observed : There 
were approximately 20 rats staying in Box 31 and 
its adjoining burrow on February 1, 1949. 
Shortly after dusk they began to emerge and 12 
were seen to proceed to and underneath a group of 
traps piled near Box 36. Occasional fights were 
heard immediately after they had assembled there. 
d. Availability of food. It was the customary 
policy to make available an abundance of food at 
all times. Under these conditions the most fre- 
quently recorded observation was that several rats, 
usually not over 10, were feeding together amicably. 
This held true even though fights were occasionally 
occurring as rats met in the activity recorders at 
the passages into the Food Pen, or when fights were 
occurring more frequently in the vicinity of the 
burrows or harborage boxes. This typical amica- 
bility characterized both the permanent supply of 
laboratory food pellets as well as temporary addi- 
tions of scrap table wastes and similar garbage. 
However, on a few occasions the food was acci- 
dently or purposely permitted to become exhausted. 
These few instances provided important informa- 
tion concerning the role of food to aggression. On 
one night, while movies were being made during 
April 1949, the only food provided was a large pile 
of garbage. Many pieces of the food such as halves 
of oranges or fresh ears of corn were nailed to a 
board. During the early part of the night, while 
the food was in abundance, up to 20 rats fed 
together with little show of antagonism, despite the 
fact each rat was frequently in actual contact with 
others. As the food supply dwindled more and 
more fights at the food source occurred. Evidence 
of a similar situation was observed on November 15, 
1948. When I left town on November 10 the 
hopper contained nearly 50 pounds of food pellets 
191 
